0

Honorary Chair
Arif Özaydın, Rector of Gaziantep University, TURKEY
Türkay Dereli, Rector of Hasan Kalyoncu University, TURKEY
Maciej Żukowski, Vice-Rector of Poznan University of Economics and Business, POLAND

Co-chairs
Eren Özceylan, Co-Chair, Gaziantep University, TURKEY
Mehmet Tanyaş, Co-Chair, Logistics Association (LODER), TURKEY

Organizing Committee
Eren Özceylan, Co-Chair, Gaziantep University, TURKEY
Mehmet Tanyaş, Co-Chair, Logistics Association (LODER), TURKEY
İbrahim Halil Korkmaz, Gaziantep University, TURKEY
Aslı Özpolat, Gaziantep University, TURKEY
Ayca Özceylan, Gaziantep University, TURKEY
Mazlum Çelik, Hasan Kalyoncu University, TURKEY
Zehra Vildan Serin, Hasan Kalyoncu University, TURKEY
İbrahim Akben, Hasan Kalyoncu University, TURKEY
Ahmet Çetindaş, Hasan Kalyoncu University, TURKEY
Atiye Tümenbatur, Logistics Association (LODER), TURKEY
Gülçin Büyüközkan, Logistics Association (LODER), TURKEY
Umut R. Tuzkaya, Logistics Association (LODER), TURKEY
Maciej Szymczak, Poznan University of Economics and Business, POLAND
Sylwia Konecka, Poznan University of Economics and Business, POLAND
Zbigniew Bentyn, Poznan University of Economics and Business, POLAND

Technical Committee
İbrahim Miraç Eligüzel, Gaziantep University, TURKEY
İlker İ. Avşar, Gaziantep University, TURKEY
Can Özcan, Hasan Kalyoncu University, TURKEY
Oya Öztürk, Hasan Kalyoncu University, TURKEY

Scientific Committee
Adil Baykasoğlu, Dokuz Eylül University, TURKEY
Ahmet Çetindaş, Hasan Kalyoncu University, TURKEY

1
Scientific Committee (Cont.)

Albert Veenstra, Eindhoven Technical University, NETHERLAND

Aleksandra Laskowska-Rutkowska, Lazarski University, POLAND
Alexandre Dolgui, IMT Atlantique, FRANCE
Alice Smith, Auburn University, USA
Aliye Ayça Supçiller, Pamukkale University, TURKEY
Anna MAryniak, Poznan University of Economics and Business, POLAND
Anna Lupicka, Poznan University of Economics and Business, POLAND
Artur Swierczek, University of Economics in Katowice, POLAND
Aslı Özpolat, Gaziantep University, TURKEY
Aşkıner Güngör, Pamukkale University, TURKEY
Atiye Tümenbatur, Maltepe University, TURKEY
Aybek Korugan, Boğaziçi University, TURKEY
Ayca Özceylan, Gaziantep University, TURKEY
Aydın Sipahioğlu, Osmangazi University, TURKEY
A. Zafer Acar, Piri Reis University, TURKEY
Bahar Yetiş Kara, Bilkent University, TURKEY
Bandar Alkhayyal, Prince Sultan University, SAUDI ARABIA
Banu Yetkin Ekren, Yaşar University, TURKEY
Batuhan Kocaoğlu, Piri Reis University, TURKEY
Bela Illes, University of Miskolc, HUNGARY
Berna Dengiz, Başkent University, TURKEY
Bilge Bilgen, Dokuz Eylül University, TURKEY
Birdogan Baki, Karadeniz Technical University, TURKEY
Bülent Çatay, Sabancı University, TURKEY
Can Berk Kalaycı, Pamukkale University, TURKEY
Ceyda Oğuz, Koç University, TURKEY
Christoph Glock, Technical University of Darmstadt, GERMANY
Cihan Çetinkaya, Adana A.T. Science and Technology University, TURKEY
Çağrı Koç, Social Sciences University of Ankara, TURKEY
Danuta Kirsperska-Moron, Karol Adamiecki University, POLAND
Dimitrios V. Lyridis, National Technical University of Athens, GREECE
Elif Kongar, University of Bridgeport, USA
Elif Özgörmüş, Pamukkale University, TURKEY
Emel Aktaş, Cranfield University, ENGLAND

2
Scientific Committee (Cont.)

Ender Gürgen, Mersin University, TURKEY

Erdal Nebol, Yeditepe University, TURKEY
Eren Özceylan, Gaziantep University, TURKEY
Erkan Çelik, Munzur University, TURKEY
Esther Alvarez, University of Deusto, SPAIN
Fabrizio Dallari, Università C. Cattaneo, ITALY
Fikri Karaesmen, Koç University, TURKEY
Filiz Çayırağası, Gaziantep University, TURKEY
Frank Straube, Technical University of Berlin, GERMANY
Frank Witlox, Ghent University, BELGIUM
Fulya Altıparmak, Gazi Universitesi, TURKEY
Füsun Ülengin, Sabancı University, TURKEY
Gábor Nagy, University of Kent, UK
Gülçin Büyüközkan, Logistics Association (LODER), TURKEY
Gülfem Tuzkaya, Marmara University, TURKEY
Güneş Erdoğan, University of Bath, UK
Hacer Güner Gören, Pamukkale University, TURKEY
Hakan Keskin, Nişantaşı University, TURKEY
Haldun Sural, Middle East Technical University, TURKEY
Hans Otto Guenther, Pusan National University, GERMANY
Hasan Akyer, Pamukkale University, TURKEY
Hasan Kıvanc Aksoy, Osmangazi University, TURKEY
Hüseyin Tunç, Social Sciences University of Ankara, TURKEY
İbrahim Akben, Hasan Kalyoncu University, TURKEY
İbrahim Halil Korkmaz, Gaziantep University, TURKEY
İmdat Kara, Başkent University, TURKEY
Kannan Govindan, University of Southern Denmark, DENMARK
Kenan Karagül, Pamukkale University, TURKEY
Kishore Pochampally, Tufts University, USA
Krzysztof Witkowski, University of Zielona Gora, POLAND
Leyla Demir, Pamukkale University, TURKEY
Maciej Szymcak, Poznan University of Economics and Business, POLAND
Mehmet Ali Ilgın, Celal Bayar University, TURKEY
Mehmet Şakir Ersoy, Beykoz University, TURKEY

3
Scientific Committee (Cont.)

Mehmet Tanyaş, Logistics Association (LODER), TURKEY

Michael Grabinski, Neu-Ulm University, GERMANY
Michael Kay, North Carolina State University, USA
Muhammet Gül, Munzur University, TURKEY
Murat Baskak, İstanbul Technical University, TURKEY
Mustafa Köksal, İstanbul Commerce University, TURKEY
Nezih Altay, DePaul University, USA
Niaz Ahmed Wassan, University of Kent, UK
Nursel Öztürk, Uludağ University, TURKEY
Olcay Polat, Pamukkale University, TURKEY
Orhan Feyzioğlu, Galatasaray University, TURKEY
Ömer Baybars Tek, Yaşar University, TURKEY
Özalp Vayvay, Marmara University, TURKEY
Özay Özaydın, Doğuş University, TURKEY
Özcan Mutlu, Pamukkale University, TURKEY
Özgür Toy, Yaşar University, TURKEY
Öznur Yurt, University of Roehampton, ENGLAND
Paolo Toth, University of Bologna, ITALY
Pedro Amorim, Universidade do Porto, PORTUGAL
Piotr Banaszyk, Poznan University of Economics and Business, POLAND
Rene De Koster, Erasmus University, NETHERLAND
Reza Lotfi, Yazd University and Behineh Gostar Sanaye Arman Co., IRAN
RobertOgulin, University of the Sunshine Coast, AUSTRALIA
Russel King, North Carolina State University, USA
Semih Coşkun, Pamukkale University, TURKEY
Semra Tunalı, İzmir University of Economics, TURKEY
Serap İncaz, Nişantaşı University, TURKEY
Serpil Erol, Gazi University, TURKEY
Soner Esmer, Dokuz Eylül University, TURKEY
Surendra M. Gupta, Northeastern University, USA
Süleyman Mete, Gaziantep University, TURKEY
Sylwia Konecka, Poznan University of Economics and Business, POLAND
Şemsettin Çiğdem, Gaziantep University, TURKEY
Şeyda Serdar Asan, İstanbul Technical University, TURKEY

4
Scientific Committee (Cont.)

Şule Önsel Ekici, Doğuş University, TURKEY

Tolga Bektaş, Liverpool University, UK
Turan Paksoy, Necmettin Erbakan University, TURKEY
Umut R. Tuzkaya, Logistics Association (LODER), TURKEY
Waldemar Budner, Poznan University of Economics and Business, POLAND
Yavuz Günalay, Bahçeşehir University, TURKEY
Yıldırım Omurtag, Robert Morris University, USA
Zbigniew Bentyn, Poznan University of Economics and Business, POLAND
Zehra Vildan Serin, Hasan Kalyoncu University, TURKEY
Zeynel Abidin Çil, İzmir Democracy University, TURKEY

5
Program at a Glance (First day)

6
Program at a Glance (Second day)

7
Table of Contents
Title of the Full Papers
Pages
Authors
MULTI-CRITERIA DECISION MAKING APPROACHES FOR WAREHOUSE LAYOUT AND
12 – 21 DESING: A LITERATURE SURVEY
Şakire Nesli Demircioğlu, Eren Özceylan
DETERMINATION OF RAILWAY NETWORK BY HIERARCHICAL MAXIMAL COVERING
22 – 32 PROBLEM IN TURKEY
Metehan Atay, Serap Ulusam Seçkiner, Eren Özceylan
A COMPARATIVE STUDY FOR THE MAXIMUM FLOW PROBLEM ARISING AT ROAD
33 – 47 NETWORKS IN KOTA KINABALU
Şeyme Safadi, Alptekin Durmuşoğlu, Eren Özceylan
IMPROVING SERVICES WITH AUGMENTED REALITY IN AIRPORTS: FACILITATING
48 – 54 AIRPORT NAVIGATION AND IMPROVING CUSTOMER SERVICE
Sercan Demir, Ersin Aktas, Turan Paksoy
A NEW ERA IN MARKETING: FACIAL RECOGNITION PAYMENT IN RETAIL INDUSTRY
55 – 64
Sercan Demir, Ersin Aktas, Turan Paksoy
SUSTAINING A RESILIENCE OF COMPETITIVE ADVANTAGE DURING COVID-19 IN
65 – 88 MARITIME SUPPLY CHAINS: FROM PERSPECTIVE OF RESOURCE-BASED THEORY
Kazım Yeni, Sedat Baştuğ
IMPACT OF COVID-19 PANDEMIC ON AIR TRANSPORTATION INDUSTRY
89 – 98
Ersin Aktas, Sercan Demir, Turan Paksoy
THE IMPACT OF LOGISTICS INNOVATIVENESS ON PERCEIVED ORGANIZATIONAL
99 – 119 PERFORMANCE: APPLICATION IN THE LOGISTICS INDUSTRY
Karahan Kara
USE OF CHECK-IN SELF SERVICE TECHNOLOGIES BEFORE AND DURING THE COVID-
120 – 134 19 PANDEMIC: EXAMPLE OF ESENBOGA AIRPORT
Mehmet Mustafa Özalp, Özgür Akarsu
SUBORBITAL SPACE TOURISM
135 – 137
Enes Demiralay, Engin Hasan Çopur, Turan Paksoy
CHANGES IN SUPPLY CHAINS DURING THE COVID-19 PANDEMIC
138 – 142
Ahmet Çetindaş
A SCOPING REVIEW OF THE LOT SIZING PROBLEM WITH MULTIPLE ITEMS, MULTIPLE
PERIODS AND MULTIPLE SUPPLIERS
143 – 153
Matheus Allgaier, Luiz Felipe Scavarda, Andréa Regina Nunes de Carvalho, Fernanda
Ruffo Tercero

8
 EXPLORING THE SUSTAINABLE SUPPLY CHAIN INTEGRATION: A CASE STUDY OF
154 – 162 DOGUS GROUP
Yasemin Ülker, Banu Demirel
BLOCKCHAIN’S POTENTIAL FOR WASTE MANAGEMENT IN TEXTILE SECTOR
163 – 176
Elif Dursun, Yasemin Ülker, Yavuz Günalay
THE USE OF AUTONOMOUS TECHNOLOGIES IN AİRPORT CHECK-IN PASSENGER FLOW
177 – 183 PRE AND POST COVID-19 PROCESS: THE CASE OF ANKARA ESENBOGA AIRPORT
Erkam Emin Ayvaz , Savaş Selahattin Ateş
THE NEXUS BETWEEN THE LOGISTICS SECTOR AND CLIMATE CHANGE: AN EMPIRICAL
184 – 190 STUDY FOR NEWLY INDUSTRIALIZED COUNTRIES
Aslı Özpolat
MAKING THE SHIFT TO A NEW PARADIGM OF LOGISTICS MANAGEMENT: SPACE
191 – 196 LOGISTICS
Enes Demiralay, Engin Hasan Çopur, Turan Paksoy
BLOCKCHAIN APPLICATIONS FOR TRACEABILITY AND FOOD SAFETY IN AGRI-FOOD
197 – 213 SUPPLY CHAINS
Şebnem İndap, Mehmet Tanyaş
ASSESSMENT OF FACTORS AFFECTING SUPPLIER SELECTION DURING CORONAVIRUS
214 – 220 PANDEMIC FOR A COSMETIC AND CLEANING PRODUCTS MANUFACTURER
Pınar Kocabey Çiftçi
AN APPLICATION TO FORECASTING OF A HOSPITAL’S ENERGY CONSUMPTION AND
221 – 235 REDUCING IMBALANCES COSTS
Aylin Ozan, Seren Özmehmet Taşan
A MIXED INTEGER PROGRAMMING MODEL FOR SUPPLY CHAIN NETWORK DESIGN
236 – 246 BASED ON NET PRESENT VALUE CONCEPT
Rahmi Baki
FACTORS AFFECTING GREEN LOGISTICS POLICIES AND THEIR EFFECTS ON
LOGISTICS PERFORMANCE INDEX: COMPARISON OF THE EUROPEAN UNION AND
247 – 257 TURKEY
Abdullah Oktay Dündar, Abdurrahim Furkan Mercimek
THE IMPACT OF COVID-19 ON E-COMMERCE: A STUDY OF SUPPLY CHAIN NETWORK
258 – 263 AND CARGO TRANSPORTATION
Muhammet Mustafa Akkan, Tufan Yayla
EUROPE’S NEW TRADE DOOR OPENNING TO ASIA: NAKHCHIVAN CORRIDOR
264 – 267
Tufan Yayla, Sinan Çizmecioğlu

9
 GREEN LOGISTICS PRACTICES IN AVIATION INDUSTRY: A RESEARCH ON ANTALYA
268 – 275 AIRPORT
Ayşegül Toy, Abdullah Oktay Dündar
BIG DATA TECHNOLOGIES: APPLICATIONS AND EVALUATIONS IN THE LOGISTICS
276 – 293 SECTOR
Faruk Taşkiran
CRITERIA PRIORITIZATION FOR ALLOCATION OF ELECTRIC VEHICLE CHARGING
STATIONS BY GROUP DECISION MAKING BASED INTUITIONISTIC FUZZY AHP
294 – 304 APPROACH
Mehmet Hanifi Aydın, Mücahid Günay
EVALUATION OF SUPPLIER SELECTION CRITERIA IN THE MARBLE SECTOR THROUGH
305 – 322 THE DEMATEL METHOD
Tuğrul Bayat, Önder İnce, Betül Kocatürk
A TIME-VARIANT CAPACITATED FACILITY LOCATION AND NETWORK DESIGN PROBLEM
323 – 331 FOR WASTEWATER TREATMENT SYSTEMS
Seyda Alperen Pehlivan, Duygun Fatih Demirel, Sıdıka Melek Başak
LEAN TRANSPORTATION MANAGEMENT THROUGH AN APPLICATION OF VEHICLE
332 – 339 ROUTING PLANNING: A CASE STUDY
Nattapong Kongprasert, Chichakorn Duriyarapee, Watcharain Larptrakool
AN INTEGRATED SUPPLIER SELECTION MODEL CONSIDERING RISK CRITERIA IN THE
340 – 347 TIRE INDUSTRY
Elif Sena Kutlu, Merve Er
ANALYZING THE OHS RISKS EMERGED IN TRANSPORTATION OF MEDICAL MATERIALS
348 – 358 IN THE COVID-19 PANDEMIC
Aylin Adem, Burcu Yılmaz Kaya, Metin Dağdeviren
RESTRICTED DYNAMIC PROGRAMMING APPROACH FOR SOLVING STOCHASTIC LOT-
359 – 364 SIZING PROBLEMS
Mustafa Çimen
THE ROLE AND IMPORTANCE OF OUTSOURCING IN LOGISTICS ENTERPRISES
365 – 381
Şuayıp Özdemir, Tuğrul Bayat, Betül Kocatürk
MULTI-CRITERIA ASSESSMENT FRAMEWORK FOR FREIGHT VILLAGES BASED ON
382 – 390 OPERATIONAL EFFICIENCY CRITERIA
Burcu Yılmaz Kaya, Aylin Adem, Metin Dağdeviren
IMPROVEMENT OF THE PRODUCTION EFFICENCY WITH LEAN MANUFACTURING
TECNIQUES
391 – 400
Eray Türkmen, Elif Tuyan, Orhun Gökdemir, Majed Alkhatib, Kübra Nur Şahin, İsmet
Söylemez

10
 SUPPLY CHAIN RISK MANAGEMENT: A LITERATURE REVIEW ON DEFINITION, THEORY,
401 – 415 AND THE ROLE OF SIMULATION AND EXISTING OPTIMIZATION METHODS
Gizem Topcu, Oğuzhan Çağıl, Yasutay Susoy, Banu Çalış Uslu
NEW META-HEURISTIC APPROACH FOR SOLVING THE VRPTW IN SUPPLY CHAIN
416 – 427 MANAGEMENT
Nesrine Bidani, Hela Moalla Frikha, Adnan Yassine
ANALYSIS OF THREAT IN GLOBAL SUPPLY CHAINS CREATING LOGISTICS BARRIERS
428 – 437 DURING A GLOBAL PANDEMIC
Zbigniew Bentyn, Sylwia Konecka
APPROACH OF A B2B DISTRIBUTION NETWORK OPTIMIZATION AS VEHICLE ROUTING
438 – 446 PROBLEM SOLVING AS A VRP LIMIT TIME IN LAST CUSTOMER
David Camacho Fonseka
ARTIFICIAL INTELLIGENCE AND SUSTAINABLE SUPPLY CHAIN FINANCE IN
447 – 454 OMNICHANNEL LOGISTICS
Monika Bal, Kinga Pawlicka
THE EFFECT OF DIGITALIZATION LEVEL AND SMART LOGISTICS PRACTISES ON
455 – 467 LOGISTICS PERFORMANCE
Gökhan Akandere

11
 MULTI-CRITERIA DECISION MAKING APPROACHES FOR
WAREHOUSE LAYOUT AND DESING: A LITERATURE SURVEY
Şakire Nesli DEMİRCİOĞLU1, Eren ÖZCEYLAN2

Abstract ⎯ Warehouse layout and design is a major input in terms of the total costs of the enterprise, as well
as playing a key role in increasing the level of competition of the enterprise. Decision problems are
encountered at the tactical and operational level in warehouse management, which is of high importance.
With the literature review studies, it is emphasized that Multi-Criteria Decision Making (MCDM) applications
should be handled as highly qualified or multi-purpose problems and that the problems should be compatible
with real life problems. It has uncertain conditions because the real-life problems are taken into account and
the criteria contain contradictory features. In this case, it becomes a difficult task to choose the best location
for the product / raw materials. For this reason, the aim of this study is to examine the MCDM applications in
the field of warehouse layout and design comprehensively. This literature study has been carried out with
relevant keywords on different international databases. In the literature review, it has been determined that
the most used techniques in the field of warehouse layout and design, which are carried out by applying the
MCDM methods, are the Analytical Hierarchy Process (AHP), ELECTRE TRI and PROMETHEE, and it has
been found that MCDM methods are used as an integrated manner in order to deal with the advantages to be
obtained by using more than one MCDM method in its field. In addition, the suggestions were presented to
researchers for further studies.

Keywords ⎯ Literature Survey, Multi-Criteria Decision Making, Warehouse Design, Warehouse Layout

INTRODUCTION
Over the past two decades manufacturing has become more complex because of competition, great varieties
and technological advances. Nowadays, the basis of competition has changed, so ensuring not only cost
effectiveness but also customization (Kumar et al., 2016). While businesses are competing in the market with
the price, quality and technological infrastructure of their final products, they may also face the risk of not
being on the shelf with the sales and stock management they apply (Sever, 2006).
Warehouses generally are refered to the storage of raw materials and all products and is used in work in
process (Yerlikaya, 2020). In his study, Yerlikaya (2020) states that warehousing activities are the activities in
which the most time and money are spent among all the activities of the enterprise. Warehouse facility design
and layout are considered as important logistics activities, in addition, warehouse selection is among the
important steps in order to realize the most efficient activities (Amin et al., 2019).
Warehouse activity is an element of fundamental importance in the integration of the logistics channel, and it
is regarded as an extension of production (Silva et al., 2015). It aims to achieve an efficient use of space that
facilitates the separation of applications and minimizes the cost of order picking (Gu et al., 2007). Different
criteria should be considered for product placement and classification in the warehouse; these criteria are size,
weight, volume, demand, cost, distance etc. Considering all these criteria at the same time and placing the
products in the most appropriate place is a difficult activity (Gu et al., 2007).
The criteria used in the fields of facility layout, design and facility location selection are determined according
to the type and content of the problem. The explanation related to these studies is explained as in Table 1.
Using Table 1., it can be seen which criteria are included in which study.

1 Şakire Nesli DEMİRCİOĞLU, Res. Assist., Hasan Kalyoncu University, Faculty of Engineering, Department of Industrial
Engineering, Gaziantep, Turkey, [email protected]
2 Eren ÖZCEYLAN, Assoc. Prof., Gaziantep University, Faculty of Engineering, Department of Industrial Engineering, Gaziantep,

Turkey, [email protected]

12
The criteria in Table 1 were encoded with letters and entered in the table; A Distance, B Adjacency Score, C
Shape Ratio, D Flexibility, E Accessibility, F Maintenance, G Closeness Gap Value Cluster, H Process
Capacity, I Performance Measures, J Productive Area Utilization, K Quality, L Human Issues Cluster, M
Throughput Time, N Product Indicators, O Work in Process, P Machine Utilization, R Material Handling
Vehicle Utilization and Handling Cost, S Productivity, T Layout Reconfigurability, U Cost, V Reliability, W
Flow.
Multi-Criteria Decision Making (MCDM) method has been used to make complex decisions, and allows for
the consideration of multiple criteria, both measurable and also non-measurable, combining quantitative and
qualitative criteria (Ilbahar et al., 2018). For all scientific researches, the decision making process is a planned
path of solving problems (Sehra et al., 2016). MCDM methods can gather historical data and expert opinions
(Pohekar and Ramachandran, 2004). The master purpose of MCDM is to evaluate and maybe choose the best
one from several alternatives based on criteria (Afshari et al., 2016).

Table 1. Criteria used in MCDM processes (adapted from Besbes et al., 2018)

Shahin and Poormostafa (2011)

Jafaryeganeh et al (2020)

Maniya and Bhatt (2011)

Singh and Singh (2011)
Abdul-Hamid al (1999)

Yang and Hung (2007)

Yamg and Kuo (2003)

Lateef, Ateekh (2013)

Hawari et al (2014)
Shokri et al (2013)

Wang et al (2016)
Yang et al (2000)

Ertay et al (2006)

Kuo et al (2008)
Shang (1993)

Abdi (2005)

Abdi (2009)
Authors

A x x x x x x x x
B x x x x x x x x
C x x x x x x x x
D x x x x x x x x x x x x x
E x x x x x x x
F x x x x x x x
G x
H x
I x
J x
K x x x
L x x x x
M x
N x
O x x
P x
R x
S x x
T x x
U x x x
V x
W x

13
 Amin et al (2019) revealed the areas of use, advantages and disadvantages of some of the MCDM
processes in their studies, and in this study, the most frequently used processes are given in Table 2.

Table 2. Explanation about the some MCDM processes (Amin et al., 2019)

Some MCDM
Application Areas Advantages Disadvantages
processes
Location and
It takes some uncertainties into
ELECTRE transportation problems Results are difficult to interpret.
accounts
etc.
Location selection, Suitable for low number of
Not suitable for too many
AHP supply chain strategy alternatives and criteria, easy to
alternatives and criteria.
etc. understand.
They also reveal the
Application limitations
ANP relationship between the The solution takes too much time.
are unknown.
criteria.
Not suitable for a small number of
Supply chain and Suitable for a large number of
TOPSIS alternatives or a large number of
logistics, energy etc. alternatives and criteria.
criteria.
It can be used when
Grey Theory It needs accurate information. It’s not give an optimal solution.
there is missing data.
The field of application There is no exact method for
PROMETHEE It is easy to use.
is quite wide. determining weights.
Simple Additive Business management, It has simple method steps. Sometimes its solution may not be
Weighting (SAW) Financial problems etc. understandable.

METHODOLOGY
The main goal of this paper is to review the literature concerning the application of MCDM methods in the
field of warehouse layout. The Web of Science (WoS) Core Collection Database was chosen to identify papers
in the period of 1977–2021.
In the publication research, a large number of studies that did not specialize in the warehouse were found, in
this case, the framework of the research was expanded a little more and citations were made about the facility
layout.
Scanning via Elsevier; (29.07.2021) 47.980 papers with the keyword of “Facility Layout”, Find articles with
these terms: 306 publications with “Layout” and “MCDM” search.
Scanning via Science Core Collection; (29.07.2021) Between 1977 year and 2021 year, 12.487 papers with the
keyword of (((((ALL=(MCDM)) AND ALL=(MCDM WAREHOUSE LAYOUT)) OR ALL=(WAREHOUSE
DESIGN)) OR ALL=(INVENTORY REPLACEMENT)) OR ALL=(SHELF AASIGNMENT)) OR
ALL=(STORAGE ALLOCATION)

14
 PRIMARY REVIEW RESULT
A summary content as in Table 3 was prepared for the analysis of similar studies. The main reason for
including this table in the study is to sample the publications that do not work directly on warehouse layout
and design.
In addition to the studies in Table 3, more important studies are included under the Detailed Review Result
title.

Table 3. Summary information about some of the analyzed studies

Specialization in
Author and Year Methods Inputs Warehouse
Layout/Management
Criteria: Stock Turnover
Analytical
Rate, Transport Distance
Colak et al (2016) Hierarchy Process 
Quantity Alternatives: 47 raw
(AHP)
materials
AHP, Data 9 different layout alternatives
Arunyanart and
Envelopment for 7 departments of the
Pruekthaisong (2018)
Analysis (DEA) company
Criteria: Cost, Volume and
Height Utilization, Ease of
Loading, Stock Turnover
Rate
Indap (2018) AHP 
Shelf Systems: Back to Back
Shelving, Narrow Aisle
Racking System, Automated
storage and retrieval systems
Criteria: weight, area,
ELECTRE TRI,
Micale et al (2019) demand, profitability and 
TOPSIS
number of customers
Criteria: Consumption
Amount, Availability in
Fuzzy Analytical
Demircioglu ve Finished Product, Access
Network Process 
Ozceylan (2021) Status, Storage Conditions,
(Fuzzy ANP)
Analysis, Delivery Adequacy
Alternatives: 9 raw materials
Fontana and Calvante
ELECTRE TRI Characteristics of products 
(2013)
6 criteria determined for 18
business areas of the business
(Criteria: Closeness gap
Hawari et al (2014) ANP value, Expansion flexibility,
Routing flexibility, Volume
flexibility, Productive area
utilization, Human issues)
The Fuzzy Full
Consistency 78 products, 4 criteria
Method, The Fuzzy (Criteria: Quantity, Unit
Vukasovic et al (2021) 
Evaluation based price, Annual Procurement
on Distance from Costs and Demand)
Average Solution
Alsyouf et al (2011) SAW 12 criteria, 3 locations

15
 33 different product brands, 4
criteria (Criteria: Ordering
Nitkratoke and
Fuzzy AHP Cost, Selling volume, 
Aengchuan (2019)
Perishability, Opportunity
Cost)
50 products, 4 criteria
(Criteria: Popularity,
Fontana et al (2020) ELECTRE TRI 
Maximum inventory, Profit,
Sensitivity)

DETAILED REVIEW RESULT

In this section, the literature review is given in more detail and specific to the studies.
Each method of MCDM follows different procedures in determining criterion weights, normalization
techniques and alternative ranking (Jahan and Edwards, 2015). Hybrid evaluation method; The first step is to
uncover the criteria and alternatives. The second step is to perform tradeoffs between criteria using pairwise
comparisons. The third step is to calculate the criteria priority vector, normalize the relevant weights, and
calculate the consistency ratio.
If the consistency rate is below 10%, the logical consistency of the binary matrix is considered to be sufficient
(Ribeiro vd., 2011). Silva et al. (2015) proposed a multi-criteria decision model for assigning and ranking
products to shelf locations in a warehouse. Fontana and Nepomuceno (2016) proposed a multi-criteria decision
model to realize the product classification and solve the storage location assignment problem in a multi-
layered warehouse.
In his study, Yerlikaya (2020) determined 3 criteria, since assigning products to storage areas creates many
time problems, especially when there is uncertainty in product demands. These 3 criteria are demand,
profitability and sensitivity. In the mentioned study, a decision matrix was created with a fuzzy number for 3
criteria. The decision matrix is expressed with linguistic values. Net flow values were also listed with the
fuzzy PROMETHEE method. According to the order of importance of the obtained product, the storage areas
were assigned. So, it is proposed to rank the products with the Fuzzy PROMETHEE method under qualitative
criteria (demand, sensitivity, profitability) for warehouse systems where the demand is uncertain and assign
them to the best storage locations according to this rank.
Jafaryeganeh et al (2020) have proven that it is possible to solve a multi-objective design problem based on
different objectives as an MCDM problem. An oil tanker has been included in the review because it has a
similar use to a warehouse. This design problem has objectives, constraints, and design variables. Purposes;
minimization of pollution prevention parameter, cargo capacity (maximizing available space) and structural
safety parameter (maximizing); constraints, regulations and other limitations. The importance of the criteria
was determined by the objective weights of the Shannon entropy technique. Alternative rankings were
evaluated by MCDM methods.
In Amin et al (2019) studies, MCDM process was used to select the best warehouse according to all criteria
determined among 5 warehouses. The criteria that guide the selection were obtained by means of a
questionnaire.
Besbes et al (2018) could not find a general model to cover all companies in their study; because each
company differs on the basis of the sector in which it operates, in the product variety, in the context of
available technologies and economic conditions. Using AHP and TOPSIS, the following conclusion was
reached; distance, adjacency, space, work flow and material handling cost.

16
In the studies of Fontana and Nepomuceno (2016), products are categorized in many companies, the only
problem is to assign them to the warehouse shelves. However, in some warehouses, products should be
categorized, their weights should be determined according to the criteria determined according to the groups
they are in, and shelf assignment processes should be completed. In this study, 4 criteria were determined, and
it is desired to evaluate them simultaneously. In other words, while providing one criterion, the other criterion
should not be compromised. For this reason, the Electre TRI method was chosen. These criteria are as follows;
product weight, area size, demand size, product's contribution to the firm.
Silva et al (2015), Storage of products which increases the speed of delivery and the competitiveness of
enterprises. The aim of this study is a multicriteria decision support model for ranking products and assigning
them to warehouse storage locations. Warehouse activity is an element of fundamental importance in the
integration of the logistics channel, and it is regarded as an extension of production. It aims to achieve an
efficient use of space that facilitates the separation of applications and minimizes the cost of order picking (Gu
et al, 2007). Different criteria should be considered for product placement and classification in the warehouse;
these criteria are size, weight, volume, demand, cost, distance etc. It is a difficult activity to consider all these
criteria at the same time and to place the products in the most appropriate place. In this study, two of the
MCDM methods were used as hybrids. The areas determined for storage with the SMARTER and
lexicographic methods were evaluated as an alternative. The criteria are listed as follows: Number of
Customers, Inventory Turnover Rate, Area Size.
In the studies of Kumar et al (2016), facility layout design selection is examined by taking the opinions of
experts. An application was made in a company with a Flexible Manufacturing Systems infrastructure. There
are 5 criteria, 5 alternatives and 4 experts.

Table 4. Criteria for FLDs

Subjective Criteria Objective Criteria
 Size and Shape of the Departments (C1)
 Distance between Facilities (C2)
 Quality of the Products (C3) Total Investment Cost (C5)
 Lighting, Ventilation and Identification
Colours used (C4)

In this paper there is an attempt to propose a FLD selection algorithm that is based on a combination of
hierarchical structure analysis and Fuzzy Set Theory. A measure called ‘Fuzzy Facilities Layout Index (FFLI)’
is proposed in this paper that handles fuzziness or vagueness inherent in the evaluation process and to provide
a standard for selecting the most appropriate FL of the alternatives without losing sight of the importance of
various criteria in FL selection process. The criteria: Many potential FL attributes Availability of Skilled
workforce, Size & Shape of the Departments, Distance between facilities, Quality of production, various cost
components involved, Lighting & Ventilation etc are considered for the selection of a right kind of FL for a
particular type of production process.
In the real-world, attributes so selected to help decision making regarding FL design selection can be
categorised into: (1) Subjective issues (These factors have qualitative definition and based solely on an
individual’s (expert) perception and ratings) such as availability of skilled workforce, size & shape of the
departments, distance between facilities, quality of production etc;
(2) Objective issues (These factors are defined in real quantitative terms) such as Investment cost, MH costs,
Operating Costs, Improvement Costs etc. The basis for that can be found in (Tompkins and White, 1984). Also
they introduced a method whereby the selection criteria regarding facilities were classified into 3 main
categories: (1) Critical Factors, (2) Subjective Factors, (3) Objective Factors.

17
 CONCLUSION
Operational performance or by managerial performance may not be the right choice for the warehouse
manager. Thus, the proposed model was able to balance these approaches and provide a more significant
overall performance.
This paper proposes a multi-criteria assignment approach to solve the storage location assignment problem
(SLAP), in class-based storage (CBS) policy, to improve the warehouse operations, as well as inventory
management. This model considers the ELECTRE III method, a well-known multi-criteria decision aiding
(MCDA) method, to construct a medium-sized valued outranking relation, and a multi-objective evolutionary
algorithm (MOEA) to exploit the outranking relation to derive a recommendation. The model compares the
classes to define their allocation in the warehouse, and it finds a SLAP solution that can be used for inventory
management, balancing the operational and tactical factors, allowing considering warehouse manager
preferences, client requirements and stock keeping unit (SKU) characteristics simultaneously. The results of
the simulated case showed the robustness of the proposed model for improving the order picking system
performance.
The benefits from this FLD on the issues of Size and Shape of the Departments, Distance between Facilities,
Quality of the Products manufactured, Lighting, Ventilation and Identification colours used outweigh the cost
as it is evident from the ratings given by the experts.
If progress is to be made by obtaining opinions from experts, more than one person with different expertise
should be interviewed. Otherwise, its reliability will be low due to its dependence on the opinion of experts.

18
 REFERENCES
Abdi, R. (2005). Selection of layout configuration for reconfigurable manufacturing system using the AHP. ISAHP,
Honolulu.
Abdi, R. (2009). Layout configuration selection for reconfigurable manufacturing systems using the fuzzy AHP. Int. J.
Manuf. Technol. Manag. 17(1/2), 149-165, Doi: 10.1504/IJMTM.2009.023783.
Abdul-Hamid Y. T, Kochhar A. K, Khan M. K. (1999). An analytic hierarchy process ap-proach to the choice of
manufacturing plant layout. Proceedings of Institute of Mechanical Engineers, Part B, 213, 397–406.
Afshari, A., Vatanparast, M. and Cockalo, D. (2016). Application of Multi Criteria Decision Making to Urban Planning-
A Review, Journal of Engineering Management and Competitiveness (JEMC), 6(1), 46-53.
Alsyouf, I., Al-Araidah, O., Tates, M. and Ciganovic, R. (2011). A multi-criteria decision-making framework for
assessing the quality and cost of facility layout alternatives: a case study, Proc. IMechE Vol. 000 Part B: J. Engineering
Manufacture, Doi: 10.1177/0954405411408664.
Amin, M., Das, A., Roy, S. and Shikdar, M. (2019). Warehouse Selection Problem Solution by Using Proper
MCDM Process, International Journal of Science and Qualitative Analysis, 5(2), 43-51.
Arunyanart, S. ve Pruekthaisong, S. (2018). Selection of multi-criteria plant layout design by combining AHP and
DEA methodologies. MATEC Web of Conferences, 1-5. France: EDP Sciences.
Besbes M., Affonso R.C., Zolghadri M., Masmoudi F. and Haddar M. (2018) Multi -criteria Decision-Making
Approaches for Facility Layout (FL) Evaluation and Selection: A Survey. In: Haddar M., Chaari F., Benamara A.,
Chouchane M., Karra C., Aifaoui N. (eds) Design and Modeling of Mechanical Systems—III. CMSM 2017.
Lecture Notes in Mechanical Engineering. Springer, Cham. Doi: 10.1007/978 -3-319-66697-6_59.
Çolak, M., Keskin, G., Günel, G. ve Akkaya, D. (2016). A Model For Efficient Layout of a Raw Material Warehouse in a
Chemical Company, Beykent University Journal of Science and Engineering, 9(2), 55-76.
Demircioglu, S. N. and Ozceylan, E. (2021). Assigning Raw Materials to Shelf With Fuzzy Analytical Network
Process and Mathematical Model Integration: An Application in The Food Industry, Journal of Industrial
Engineering, 32(2), 321-340.
Ertay, T., Ruan, D., Tuzkaya, U. R. (2006). Integrating data envelopment analysis and analytic hierarchy for the facility
layout design in manufacturing systems, Information Sciences, 176(3), 237-262, Doi: 10.1016/j.ins.2004.12.001.
Fontana, M. and Cavalcante, C. (2013). Electre Tri Method Used to Storage Location Assignment Into Categories,
Pesquisa Operacional, 33(2), 283-303, Doi: 10.1590/S0101-74382013000200009.
Fontana, M. and Nepomuceno, V. (2016). Multi-criteria approach for products classification and their storage location
assignment, The International Journal of Advanced Manufacturing Technology, 88, 3205-3216.
Fontana, M., Lopez, J., Cavalcante, C. and Noriega, J. (2020), Multi-Criteria Assignment Model to Solve The Storage
Location Assignment Problem, Revista Investigation Operational, 41(7), 1019-1029.
Gergin, R. and Peker, I. (2019). Literature review on success factors and methods used in warehouse location selection,
Journal of Pamukkale University Engineering Sciences, 25(9), 1062-1070.
Gu, J., Goetschalckx, M. and McGinnis, L. (2007). Research On Warehouse Operation: A Comprehensive Review,
European Journal of Operational Research, 177(1), 1–21.
Hawari, T., Mumani, A. and Momani, A. (2014). Application of the Analytic Network Process to facility location
selection, Journal of Manufacturing Systems, 33, 488-497, Doi: 10.1016/j.jmsy.2014.04.006.
Ilbahar, E., Karaşan, A., Cebi, S. and Kahraman, C., (2018). A novel approach to risk assessment for occupational health
and safety using Pythagorean fuzzy AHP & fuzzy inference system, Safety Science, 103, 124–136.
Indap, S. (2018). Application of the Analytic Hierarchy Process in the Selection of Storage Rack Systems For E-
Commerce Clothing Industry. Journal of Management, Marketing and Logistics, 5(4), 255-266.

19
Jafaryeganeh, H., Ventura, M and Soares, C. (2020). Application of multi-criteria decision making methods for selection
of ship internal layout design from a Pareto optimal set, Ocean Engineering, 202, 1-14, Doi:
10.1016/j.oceaneng.2020.107151.
Jahan, A. and Edwards, K. (2015). A state of the art survey on the influence of normalization techniques in ranking:
improving the materials selection process in engineering design, Materials and Design, 65, 335-342.
Kumar, G., Sarkar, B and Sanyal, S. (2016). Evaluating Facilities Layout Designs using Fuzzy-MCDM method,
International Journal od Scientific and Engineering Research, 7(8), 2093-2106.
Kuo Y., Yang T, Huang G.W (2008) The use of grey relational analysis in solving multiple attribute decision-making
problems, Computers & Industrial Engineering, 55(1), 80–93.doi: http://dx.doi.org/10.1016/j.cie.2007.12.002.
Lateef-Ur R, Ateekh-Ur R (2013) Manufacturing configuration selection using multicriteria decision tool, Int. J. Adv.
Manuf. Tech., 65(5–8), 625–639, Doi: 10.1007/s00170-012-4201-5.
Maniya K. D, Bhatt M. G. (2011). An alternative multiple attribute decision making meth-odology for solving optimal
facility layout design selection problems, Computers & In-dustrial Engineering, 61, 542–549, Doi:
10.1016/j.cie.2011.04.009.
Micale, R., Fata, L. ve Scalia, L. (2019). A combined interval-valued ELECTRE TRI and TOPSIS approach for solving
the storage location assignment problem, Computers&Industrial Engineering, 135(9), 199-210.
Nitkratoke, S. and Aengchuan, P. (2019). FAHP in Multi-Criteria Inventory Classification for Storage Layout,
International Conference on Advanced Research in Applied Science and Engineering, Amsterdam: Netherlands.
Pohekar, S. and Ramachandran, M. (2004). Application of Multi-Criteria Decision Making To Sustainable Energy
Planning-A Review, Renewable and Sustainable Energy Reviews, 8(4), 365-381.
Ribeiro, R., Moreira, A., Broek, V. and Pimentel, A. (2011). Hybrid assessment method for software engineering
decisions, Decision Support Systems, 51(1), 208-219.
Sehra, S., Brar, Y. and Kaur, N. (2016). Applications of Multi-criteria Decision Making in Software Engineering,
International Journal of Advanced Computer Science and Applications (IJACSA), 7(7), 472-477.
Sever, M. (2006). Application of Enterprise Mobility in Warehouse Management: A Case Study, (Master Thesis),
İstanbul University, Institute of Social Sciences, İstanbul. Address: http://nek.istanbul.edu.tr:4444/ekos/TEZ/42062.pdf
Shahin, A., Poormostafa, M. (2011). Facility layout simulation and optimization. An integra-tion of advanced quality and
decision making tools and techniques. Modern Applied Science, 5(4), 95, Doi: 10.5539/mas.v5n4p95.
Shang, S. (1993). Multi-criteria facility layout problem: An integrated approach. European Journal of Operational
Research, 66(3), 291–304, Doi: 10.1016/0377-2217(93)90218-C.
Shokri, H., Ashjari, B., Saberi, M., Yoon, JH. (2013) An integrated AHP. VIKOR methodology for facility layout design,
Ind. Eng. Manag. Sys., 12(4), 389–405, Doi: 10.7232/iems.2013.12.4.389.
Singh, S.P and Singh, V.K (2011) Three-level AHP-based heuristic approach for a multi-objective facility layout
problem, International Journal of Production Research, 49(4), 1105–1125, Doi: 10.1080/00207540903536148
Silva, D., Vasconcelos, N. and Cavalcante C. (2015). Multicriteria Decision Model to Support the Assignment of Storage
Location of Products in a Warehouse, Hindawi Mathematical Problems in Engineering, Doi: 10.1155/2015/481950.
Tompkins, J. and White, J. (1984). Facility Planning, John Wiley&Sons Inc.
Vukasovic, D., Gligovic, D., Terzic, S., Stevic, Z. and Macura, P. (2021). A Novel Fuzzy MCDM Model for Inventory
Management In Order To Increase Business Efficinecy, Technological and Economic Development of Economy, 27(2),
386-401, Doi: 10.3846/tede.2021.14427.
Wang P, Zhu Z, Wang Y (2016) A novel hybrid MCDM model combining the SAW, TOPSIS and GRA methods based
on experimental design. Information Sciences, 345(1): 27-45.doi:http://dx.doi.org/10.1016/j.ins.2016.01.076.
Yang, T., Su, C.T, Hsu, Y.R. (2000) Systematic layout planning: A study on semiconductor wafer fabrication facilities,
International Journal of Operations and Production Management, 20(11), 1360–1372, Doi:
10.1108/01443570010348299.
Yang, T., Kuo, C.A. (2003) A hierarchical AHP/DEA methodology for the facilities layout design problem. European
Journal of Operational Research, 147(1), 128-136, Doi: 10.1016/S0377-2217(02)00251-5.

20
Yang, T., Hung. CC. (2007) Multiple-attribute decision making methods for plant layout de-sign problem. Robotics and
Computer Integrated Manufacturing, 23, 126–137, Doi: 10.1016/j.rcim.2005.12.002.
Yerlikaya, M. (2020). Storage Location Assignment with Fuzzy PROMETHEE Method in Warehouse Systems with
Uncertain Demand, Journal of the Institute Electronics and Computer, 2, 142-150.

21
 DETERMINATION OF RAILWAY NETWORK BY HIERARCHICAL
MAXIMAL COVERING PROBLEM IN TURKEY

Metehan ATAY1, Serap ULUSAM SEÇKİNER2, Eren ÖZCEYLAN3

ABSTRACT

As the world population growth increases, public transportation becomes one of the most
important development challenges. Among these, High Speed Rail (HSR) is an efficient
alternative for rapid mass transits that has been developed in many countries around the
world. HSR network is a solution of mass transportation, since it offers a type of transport
that fast, convenience, reliable, save, ecological, and environment friendly, which connected
cities across countries. In this paper, it is studied to determine optimal station locations on a
HSR using a hierarchical maximal covering model in Turkey. The objective is to maximize the
number of passengers traveled by HSR, under the limitations of the number of stations for
both main and intermediate, and coverage distance. Turkish conventional railway line is
presented as HSR network under case scenarios. While the model is constructed, 43 cities are
considered in the network. The number of real passengers travelled during year is used as
HSR passengers. Optimum result reached in a reasonable time under several conditions and
considered scenarios.

Keywords ⎯ High Speed Rail, Hub Location, Maximal Covering, Railway Network

1
Metehan ATAY, Istanbul Arel University, Faculty of Engineering and Architecture, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]
2
Serap ULUSAM SEÇKİNER, Gaziantep University, Engineering Faculty, Department of Industrial Engineering, Gaziantep, Turkey,
[email protected]
3
Eren ÖZCEYLAN, Gaziantep University, Engineering Faculty, Department of Industrial Engineering, Gaziantep, Turkey,
[email protected]

22
 1. Introduction
As the world population growth increases, public transportation becomes one of the most
important development challenges. High Speed Rail (HSR) is an efficient alternative for rapid
mass transits that has been developed in many countries around the world. HSR network is a
solution of mass transportation, since it offers a type of transport that fast, convenience,
reliable, save, ecological, and environment friendly, which connected cities across countries
or between continents. However, the cost of building HSR is very expensive. So a design
process on the routes, station locations, number of trips, which according customers’ demand
is necessary.

Although the main purpose of transportation methods is to benefit society, it is expected to

contribute to the environment and economy. Some of these can be exemplified as the increase
in the use of public transportation methods decreasing greenhouse gas emissions and making
it more profitable in terms of GDP. However, due to some acceptable reasons such as being
fast and comfortable, There is a systematic formation based on high speed infrastructure in
acquiring a high speed railway line. It should be noted that the sequence of lines to be
revealed is directly related to the success and integrity of the system. Infrastructures with
special conditions in which high-speed trains can be operated are a requirement in catching
high speeds. Because of all these, a single definition or standard of high-speed railways has
not been reached yet. Due to the complex structure in question, there is no single standard for
high-speed railways. However, it is possible to mention a few definitions that offer joint
evaluation opportunities under various conditions. There are evaluation criteria such as
infrastructure, speed and service parameters for a line to be considered as HSR class. It is
known that the speed is limited to 110 km/h due to noise-related problems in populated areas
and 160-180 km/h for safety and carrying capacity reasons in places with long bridges and
tunnels.

From this point of view, it is clearly understood that the speed factor, which is one of the most
important and distinguishing features of the transportation type in question, can be varied
according to the relevant operating conditions or requirements. On the other hand, some laws
and measures taken to prevent COVID-19 have stopped many industrial and transportation
activities. The cessation of many production and transportation activities has led to many
changes in both environmental and economic terms. Air pollution rate and greenhouse gas
emissions decreased by 6,4% compared to the previous year with the measures taken due to
COVID-19 and the cessation of industrial transportation (Tollefson J., 2021). In addition, the
French government has announced that it will remove short-haul airline flights to be operated
by train (Loh, C., 2021). It is known that the main reasons for this situation are the unbearable
costs and the encouragement of railway transportation, which is relatively less costly to the
airline. This situation is made clear that the crisis situation in Europe will bring more
restrictions on industrial passenger transport by states.

Similarly, flight time between many airports in Turkey is very short and it is a bit more
expensive than train transportation. In addition, there are some time losses in terms of the total
time consumed by the procedures to be completed for transportation to airports and boarding.

23
Study carried out by Baumeister and Leung can be given as example in terms of explaining
the situation for Finland. Table 1 explaining the comparative airline and train access and
travel times presented by the study is given below.

Table 1: Real travel times of Aircraft and NHSR (Baumeister and Leung, 2021)

Mode Aircraft NHSR

Transfer to Airport by car 8-22 min -
Transfer to Rovaniemi/Oulu by bus - 115-230 min
Check in at airport 30 min -
Security control, transfer to gate 45 min -
Early Arrival at station for boarding - 10 min
Scheduled flight/travel time 35-95 min 91-512 min
Baggage claim and transfer to ring line 40 min -
Ring line travel 31 min -

In this paper, we proposed optimal station locations on a HSR using a maximal hub covering
location model. The objective is to maximize the number of passengers travel by HSR, under
the limitations of the number of stations for both main and intermediate, and coverage
distance. Case study of the Turkish railway line was presented. Number of real passengers
travelled during year was used to estimate expected number of HSR passengers. Results can
be found in reasonable time. According to TURKSTAT, change in the number of passengers
transported by population is shown in Figure 1.

3,50%
3,00%
2,50%
2,00%
1,50%
1,00%
0,50%
0,00%
2008

2015

2019
2004
2005
2006
2007

2009
2010
2011
2012
2013
2014

2016
2017
2018

Figure 1 : % Portion of Total Amount of Railway Passenger by Population (TURKSTAT,

2020)

In this paper, rest of the study was constructed under three main headings as literature review,
implementation results and conclusion. In literature review section, several studies that
explains similiar problems are explained in recent years. Implementation and result sections
shows the solutions reached and scenarios considered while problems are developed. At last,
conclusion section gives some information about limitations and hypothesis of this work.

24
 2. Literature review
General hub location problem deals with locating facilities (hub) and allocating point of
demands (node) to facilities on a network, in order to provide a transshipment service, routed
between origin-destination. Service delivery to customers in most location problems depends
on the distance between customers and their assigned facilities. While trying to minimize the
average distances in some location problems, this purpose is not suitable for some situations.
For example, in the choice of location of emergency service vehicles such as fire brigade or
ambulance, it is vital to meet the demand in a short time, so a maximum reasonable travel
distance or time should be determined. The keyword here is the word of coverage (Owen and
Daskin, 1998). Coverage problems are divided into set covering problems, which require full
demand coverage, and maximum covering problems, which give the best coverage of
demand. The aim of the maximum coverage problem, first formulated by Church and ReVelle
(1974), is to maximize the amount of demand covered within the reasonable service distance
by placing a fixed number of facilities (ReVelle et al., 1977). Objective is to minimize the
total transportation cost of moving flows from origin node to destination node via hub. Initial
literature is shown on Table 2.

Table 2: Studies on hub location problems

Author Problem Type

O’Kelly (1987) Quadratic integer programing for the single allocation p-hub median
problem
Campbell (1992) First linear integer programming for the single allocation p-hub
median problem
Skorin-Kapov, et al. (1996) Mixed integer programing formulation
Ernst and Krishnamoorthy Improved linear integer programming formulation
(1996)
O’Kelly (1992) Single allocation hub location problem with fixed costs
Campbell (1994) Capacitated multiple allocation p-hub median problem
Kara and Tansel (2003) Single allocation p-hub covering problem
Wagner ( 2008) İmproved p-hub covering location problem with bounded path
lengths
Hwang and Lee (2012) Uncapacitated single allocation p-hub maximal covering problem
for CAB data
Peker and Kara (2015) p-hub maximal covering problem with gradual decay function as
partial coverage
Musavi and Bozargi-Amiri Scheduled and sequenced of vehicles hub in perishable supply chain
(2017)
Chanta and Sangsawang Optimization model for determining locations of railway stations in
(2018) Thailand
Ma et al. (2020) Hierarchical Multimodal hub location with time restriction for
China railway
Chanta and Sangsawang, Optimal station locations for HSR based on partial coverage and
(2021) passenger cost savings
Isler et al. (2021) Designing strategic railway freight network in Brazil via GIS.

25
For more details on hub location problems in maximal covering models, see Alumur and Kara
(2008) and Farahani, et al. (2013). On the other hand, there a alot of application models exist
in literature. There are also examples applied for different areas, not only for hub location or
facility location selection. One of these studies was conducted by Li et al. (2020). In this
framework, the model of the study is designed as a problem of maximum coverage and is as
stated below ( ReVelle et al. 1997).

𝑀𝑎𝑥 ∑ 𝑎𝑖 𝑦𝑖 (1)
𝑖∈𝐼

𝑠. 𝑡
∑ 𝑥𝑗 ≥ 𝑦𝑖 ∀𝑖 ∈ 𝐼 (2 )
𝑗∈𝑁

∑ 𝑥𝑗 = 𝑝 (3 )
𝑗∈𝐽

𝑥𝑗 , 𝑦𝑖 ∈ {0,1} ∀𝑖 ∈ 𝐼, ∀𝑗 ∈ 𝐽 (4)

Objective (1) maximizes the number of people covered within the critical distance. Constraint
set (2) allows coverage of demand point i if one or more facilities are established within
critical distance. Constraint (3) limits the number of established facilities to p. Constraint set
(4) ensures all variables to be binary. In a classical MCLP, one seeks the location of a number
of facilities on a network in such a way that the covered population is maximized. A facility
covers a demand node, if it is established in a distance less than the threshold to the demand
node. This pre-defined threshold is often called the coverage radius, which directly affects the
solution of the problem.

2.1. Maximal Coverage Model

Since the primary and secondary supply stations are considered in the present study, the
partial coverage problem proposed by Karasakal and Karasakal (2004) was used. Therefore,
there are 6 different critical distances. These are T1, T2, T3 and S1, S2, S3. Following values
are used in model. Since the data of cities were not shared while collecting demand data, data
extraction was used. As shown in Figure.1, the amount of rail passengers varying over the
years varies between 1% and 3% of the total population. Since the highest rate was
determined as 3%, the demands in the study were used as 3% of the total population of the
specified cities. In addition, the railway distance chart used is obtained from open source web
pages. The MCLP-P is modeled by Karasakal and Karasakal (2004) as follows;

𝑀𝑎𝑥 ∑ ∑ 𝑐𝑖𝑗 𝑥𝑖𝑗 (5)

𝑖∈𝐼 𝑗∈𝑀𝑖
𝑠. 𝑡
∑ 𝑦𝑗 = 𝑃 (6)
𝑗∈𝐽
𝑥𝑖𝑗 ≤ 𝑦𝑗 ∀𝑖 ∈ 𝐼, 𝑗 ∈ 𝑀𝑖 (7)
∑ 𝑥𝑖𝑗 ≤ 1 ∀𝑖 ∈ 𝐼 (8)
𝑗∈𝑀𝑖
𝑦𝑗 ∈ {0,1} ∀𝑗 ∈ 𝐽 (9)

26
 𝑥𝑖𝑗 ∈ {0,1} ∀𝑖 ∈ 𝐼, 𝑗 ∈ 𝑀𝑖 (10)
Where
𝐼 ∶ 𝑖𝑛𝑑𝑒𝑥 𝑠𝑒𝑡 𝑜𝑓 𝑑𝑒𝑚𝑎𝑛𝑑 𝑝𝑜𝑖𝑛𝑡𝑠
𝐽 ∶ 𝑖𝑛𝑑𝑒𝑥 𝑠𝑒𝑡 𝑜𝑓 𝑝𝑜𝑡𝑒𝑛𝑡𝑖𝑎𝑙 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑠𝑖𝑡𝑒𝑠
𝑃 ∶ 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑖𝑡𝑒𝑠 𝑡𝑜 𝑏𝑒 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑎𝑡𝑒𝑑
𝑀𝑖 : 𝑠𝑒𝑡 𝑜𝑓 𝑠𝑖𝑡𝑒𝑠 𝑡ℎ𝑎𝑡𝑒𝑖𝑡ℎ𝑒𝑟 𝑓𝑢𝑙𝑙𝑦 𝑜𝑟 𝑝𝑎𝑟𝑡𝑖𝑎𝑙𝑙𝑦 𝑐𝑜𝑣𝑒𝑟 𝑡ℎ𝑒 𝑑𝑒𝑚𝑎𝑛𝑑 𝑓𝑟𝑜𝑚 𝑖
𝑆: 𝑚𝑎𝑥𝑖𝑚𝑢𝑚 𝑐𝑜𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒
𝑇: 𝑚𝑎𝑥𝑖𝑚𝑢𝑚 𝑝𝑎𝑟𝑡𝑖𝑎𝑙 𝑐𝑜𝑣𝑒𝑟 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒 (𝑇 > 𝑆)
𝐷𝑖𝑗 : 𝑡𝑟𝑎𝑣𝑒𝑙 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑗𝑎𝑛𝑑 𝑑𝑒𝑚𝑎𝑛𝑑 𝑠𝑖𝑡𝑒 𝑖
𝐶𝑖𝑗 : 𝑙𝑒𝑣𝑒𝑙 𝑜𝑓 𝑐𝑜𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑟𝑜𝑣𝑖𝑑𝑒𝑑 𝑏𝑦 𝑡ℎ𝑒 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑗 𝑡𝑜 𝑡ℎ𝑒 𝑑𝑒𝑚𝑎𝑛𝑑 𝑝𝑜𝑖𝑛𝑡 𝑖

1, 𝑖𝑓 𝐷𝑖𝑗 ≤ 𝑆 , 𝑓(𝐷𝑖𝑗), 𝑖𝑓 𝑆 < 𝐷𝑖𝑗 < 𝑇, 0 < 𝑓(𝐷𝑖𝑗) < 1

𝐶𝑖𝑗 : {
0, 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

1,
𝑖𝑓 𝑎 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑖𝑠 𝑠𝑖𝑡𝑒𝑑 𝑎𝑡 𝑗
𝑦𝑗 : {
0, 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
1, 𝑖𝑓 𝑖𝑓 𝑡ℎ𝑒 𝑑𝑒𝑚𝑎𝑛𝑑 𝑎𝑡 𝑝𝑜𝑖𝑛𝑡 𝑖 𝑖𝑠 𝑒𝑖𝑡ℎ𝑒𝑟 𝑝𝑎𝑟𝑡𝑖𝑎𝑙𝑙𝑦 𝑜𝑟 𝑓𝑢𝑙𝑙𝑦 𝑐𝑜𝑣𝑒𝑟𝑒𝑑 𝑏𝑦 𝑎 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑎𝑡 𝑗
𝑥𝑖𝑗 : {
0, 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

Objective function maximizes the coverage level within the maximum critical distance T in
Equation 5. Constraints are sequentially limits the number of facilities to be sited to P on
Equation 6, ensures that if a facility is not sited at j, then demand at i can not be covered by j
in Equation 7, all demand points can be covered by at most one facility in Equation 8. If there
are more than one facilities covering a demand point, the facility that provides the maximum
coverage will be selected which is forced by the objective function. Last two constraint sets
impose binary restriction on the decision variables.

Table 3 : Initial parameters used in this models and its values

Parameters Value
(km)
S1 : minimum critical distance for demand-by intermediate station coverage 50
S2 : minimum critical distance for demand-by-main station coverage 452
S3 : minimum critical distance for intermediate-by-main station coverage 851
T1 : maximum critical distance for demand-by-instermediate station coverage 451
T2 : maximum critical distance for demand-by-main station coverage 850
T3: maximum critical distance for intermediate station-by-main station coverage 1250

To identify primary and secondary train stations in railway transportation networks. Railway
distances between provinces and each other were used. Problem is modeled and solved by
GAMS. By keeping the critical distances used in the problem constant, selected base station
numbers and their effects on the objective function were also investigated.

3. Implementation and Results

Using the parameters given in Table 3, Problem is solved by GAMS CPLEX software on
version 23.5 by using a intel i7 2.00 GHz processor and 8 GB RAM capacity. Number of
possible main station is used as input arbitrarily to model. None of the critical distances are
changed initially to understand model behaviour for assignment of main stations. However,

27
for the optimal solution and the number of stations, different station quantity parameters have
been tried with fixed covering distances. Number of possible main station is limited to 5 and
intermediate stations for each main station is limited to 9 for an initial solution. Change on
objective function value according to the number of main stations are shown in Figure 2. It is
possible to say that there is two minimum points for assignment of main stations. According
to the result, the maximum number of main stations to be selected as four and the number of
intermediate stations is determined under the condiditons of more than one main station
should be selected. Y axis shows objective function value and X axis shows the selected
number of hubs.

Objective function value

1860000

1850000

1840000

1830000

1820000

1810000

1800000

1790000
1 2 3 4 5 6 7 8 9
Number of hubs

Figure 2 : Change of objective function value by number of main stations

According to the Figure 2, number of main stations should be selected as for. These stations
are Afyon,Kırıkkale,Malatya and Mersin. Assignments of intermediate stations to main
stations are shown in following table 4.

Table 4 : Main station demand by Intermediate station Assignments

Main Stations Afyon Kırıkkale Malatya Mersin

Assigned Edirne Amasya Ankara Istanbul
Intermediate Karaman Kırklareli Erzurum Mardin
Stations Tekirdağ Kars Samsun

This assignment is performed for direct demand by main station coverage. This coverage is
performs if intermediate station at i is opened and assigned to main station at j originally. It is
possible to say that those assigned demand points will be serve directly to main stations
detected shortly. Another assignment is expressed for intermediate station demand from
remote points through intermediate stations to main stations as third coverage. Although
results were obtained, scenarios were established to measure the effect of the primary

28
coverage distances used in model on the cost. Changed coverage distances in established
scenario and the number of base stations to be selected are shown in the table 5 below as Cov.

Table 5: First coverage distance parameters used in each scenario

First Coverage
S1 T1
Cov1 50 451
Cov2 50 400
Cov3 50 350
Cov4 50 300
Cov5 50 250
Cov6 50 200

The purpose of each coverage scenario is actually to identify the most efficient network that
can be transported at the least cost. These parameters were placed in the existing model and
outputs were obtained for each scenario. Obtained results are shown below on Table 6.

Table 6: Objective function value change due to each scenario

Number of main stations

1 2 3 4 5
Cov1 1851180 1830610 1823610 1819610 1817610
Cov2 1851180 1830610 1823610 1819610 1817610
Cov3 1851180 1830482 1823072 1819199 1817610
Cov4 1851180 1825814 1815293 1813602 1814983
Cov5 1851180 1818712 1804003 1801576 1806725
Cov6 1851180 1802613 1781785 1768611 1787203

As Table 6 shows, costs decrease as coverage gets smaller. Although the reduced cost seems
to be a better result, it can only be an initial solution as the infrastructure costs for the selected
main stations are not taken into account. The lowest costs were obtained from the Cov6
scenario. In this case, the main station locations selected for Cov6 scenario need to be known
and examined. According to Cov6 scenario Kayseri is selected while number of hub is 1.
Afyon and Elazig is selected while number of hub is 2. Afyon, Malatya and Tekirdağ is
selected when n = 3. If we push model to select more than three hubs, locations are
reorganized and selected as Izmit, Malatya, Manisa and Nigde. When number of hubs
selected is increased to 5, only Tekirdag location is added up to n=4 situation locations.

4. Conclusion and Discussion

Although the use of trains is much less than other vehicles today, where logistics activities are
inevitable, they have a great place in industrial transportation. It can be said that train
transportation is very popular, especially between cities close to each other. However,
increasing the use of trains among other means of transport can be a good alternative to both

29
reducing the traffic level and reducing carbon emissions. In this study, an existing model was
used to design NHSR networks operating in Turkey and make them capable of carrying
maximum passengers, and the results were discussed by usinh hierarchical maximal covering
model. Scenarios were created and examined in order to examine the model's behavior and
geographical choices. It is a known fact that airplanes with high landing/take-off activities in
terms of emission measurements increase carbon emissions considerably. Frequent and short-
haul flights create a lot of pollution for the environment at this point. It is thought that a more
efficient transportation will be provided as a result of the replacement of these flights with
some flights with the help of HSR networks. Based on shared data, many scenarios can be
produced to save time and reduce carbon emissions. Although there is an airport, high-speed
trains can be substituted for flights that takes lower than 45 minutes from hub locations. For
the Istanbul-Ankara journey, the distance between which is 453 km, 4 hours and 15 minutes
travel by high-speed train can be envisaged, instead of losing time while arriving at the
airports 45 minutes earlier and airport shuttle time. In total this flight takes approximately 4.5
hours in total.

However, the study was carried out with a limited data set. Due to the limited data access,
some problem assumptions are considered absent. One of these constraints is capacity
constraint. Because the qualifications and capacities of the stations in each province are
variable. However, since these data are not accessible, the capacity of all of them is not
considered as a constraint. This indicates that the problem is more static and is one of the
typical problems encountered in modeling a real life problem. In addition, each train type and
capacity and fuel consumption are not equal. These factors vary from route to route and train
type. In other words, it was assumed that they all have enough capability to meet the demand.
Therefore, manipulative results are likely to be found in the model results. However, it is
thought that this study will be an idea for projects and breakthroughs to be made for public
transportation, on which the foundations can be laid in the future, also can serve greener and
wider communities more efficiently. At the same time, the development and operation of
public transportation and transportation skills within the country will make a greater
contribution to the country's economy. It may be possible to reduce travel costs and carbon
emissions, which endanger the environment, by increasing the use of public transportation
and time savings by making comprehensive plans in line with the scenarios. In this way, it is
possible to make very efficient journeys by saving our time with a green transportation.

30
REFERENCES

Alumur, A., Kara, B.Y.: Network hub location problems: the state of the art. Eur. J. Oper.
Res. 190, 1–21 (2008)

Baumeister, S., & Leung, A. (2021). The emissions reduction potential of substituting short-
haul flights with non-high-speed rail (NHSR): The case of Finland. Case Studies on Transport
Policy, 9(1), 40-50.

Campbell, J.F.: Integer programming formulations of discrete hub location problems. Eur. J.
Oper. Res. 72, 387–405 (1994)

Campbell, J.F.: Location and allocation for distribution for distribution systems with
transshipments and transportation economies of scale. Ann. Oper. Res. 40, 77–99 (1992)

Chanta, S., & Sangsawang, O. (2018, October). A single allocation p-hub maximal covering
model for optimizing railway station location. In International Conference on Intelligent
Computing & Optimization (pp. 522-530). Springer, Cham.

Chanta, S., & Sangsawang, O. (2021). Optimal railway station locations for high-speed trains
based on partial coverage and passenger cost savings. International Journal of Rail
Transportation, 9(1), 39-60.

Ernst, A.T., Krishnamoorthy, M.: Efficient algorithms for the uncapacitated single allocation
p-hub median problem. Locat. Sci. 4, 139–154 (1996)

Farahani, R.Z., Hekmatfar, M., Arabani, A.B., Nikbakhsh, E.: Hub location problem: A
review of models, classification, solution techniques, and applications. Comput. Ind. Eng. 64,
1096–1109 (2013)

Hwang, Y.H., Lee, Y.H.: Uncapacitated single allocation p-hub maximal covering problem.
Comput. Ind. Eng. 63, 382–389 (2012)

Isler, C. A., Asaff, Y., & Marinov, M. (2021). Designing a Geo-Strategic Railway Freight
Network in Brazil Using GIS. Sustainability, 13(1), 85.

Kara, B.Y., Tansel, B.C.: The single-assignment hub covering problem: models and
linearizations. J. Oper. Res. Soc. 54(1), 59–64 (2003)

Karasakal, O. & Karasakal, E. (2004). A Maximal Covering Location Model in The Presence
of Partial Coverage. Computers & Operations Research, 31, 1515-1526.
Li, H., Mukhopadhyay, S. K., Wu, J. J., Zhou, L., & Du, Z. (2020). Balanced maximal
covering location problem and its application in bike-sharing. International Journal of
Production Economics, 223, 107513.

Loh, C. (2021). France Moves Closer To Barring Short-Haul Domestic Flights. Simpleflying.
/ Retrieved from https://simpleflying.com/france-short-haul-domestic-ban-closer/ 24.04.2021/

31
Ma, Y., Shi, X., & Qiu, Y. (2020). Hierarchical multimodal hub location with time restriction
for China Railway (CR) Express network. IEEE Access, 8, 61395-61404.

Musavi, M., & Bozorgi-Amiri, A. (2017). A multi-objective sustainable hub location-

scheduling problem for perishable food supply chain. Computers & Industrial Engineering,
113, 766-778.

O’Kelly, M.E.: A quadratic integer program for the location of interacting hub facilities. Eur.
J. Oper. Res. 32, 393–404 (1987)

O’Kelly, M.E.: Hub facility location with fixed costs. Pap. Reg. Sci. 71, 292–306 (1992)

Owen, S.H. and Daskin, M.S. (1998).Strategic facility location: A review.111(3), 423-447,
European Journal of Operational Research.

Peker, M., Kara, B.: The p-hub maximal covering problem and extensions for gradual decay
functions. Omega 54, 158–172 (2015)

ReVelle, C., Bigman, D., Schilling, D., Cohon, J. and Church, R. (1977).Facility location: a
review of context-free and EMS models. 12 (2), 129146, Health Services Research.

Skorin-Kapov, D., Skorin-Kapov, J., O’Kelly, M.E.: Tight linear programming relations of
uncapacitated p-hub median problems. Eur. J. Oper. Res. 74, 582–593 (1996)

Tollefson, J. (2021). COVID curbed carbon emissions in 2020-but not by much. Nature./
Retrieved from https://www.nature.com/articles/d41586-021-00090-3/ 25.03.2021 /

TURKSTAT, Passenger Transportation by Railways Statistics, 2020. Released on

29.09.2020/https://data.tuik.gov.tr/Bulten/DownloadIstatistikselTablo?p=Qu1pZ0OkJ3RgwW
uwOBfpqZycq14fQJRJw0Doa9Z4BLtENdLXlPS2INUor/5hJJ6w / Accessed on 24.08.2021

Wagner, B.: Model formulations for hub covering problems. J. Oper. Res. Soc. 59, 932–938
(2008)

32
 A COMPARATIVE STUDY FOR THE MAXIMUM FLOW PROBLEM
ARISING AT ROAD NETWORKS IN KOTA KINABALU
Şeyme Safadi1, Alptekin Durmuşoğlu2, Eren Özceylan3

Abstract ⎯ Maximum flow problems are defined as the maximum amount of flow in a network from
source to sink while considering the capacity of each edge to find the feasible flow in a directed
network. There are plenty of areas that apply maximum flow methodology in finding feasible/optimal
solution, and so one of the most famous real world applications is transportation/traffic problems. In
this paper, a real-life traffic problem arising at Kota Kinabalu, Malaysia from literature is taken into
consideration. Three well-known algorithms namely Dinic's algorithm, Edmonds–Karp algorithm,
Ford–Fulkerson algorithm are applied to the problem including 24 nodes and 60 edges. A
comparative analysis is conducted in terms of CPU time and objective function values. The results of
the algorithms are also compared with the known optimum solution. Finally, the three methodologies
show the same amount of flow in the traffic network while there is a significant difference in the
central process unit “CPU” time whereas the CPU time to solve the problem using Edmond’s Karp is
significantly less comparing with the others.

Keywords ⎯ Dinic's algorithm, Edmonds–Karp algorithm, Ford–Fulkerson algorithm, Maximum flow

problem, Road networks.

INTRODUCTION
Maximum flow problem was introduced for the first time in 1950sby T. E. Harris and F. S. Ross as
indicated in the review paper by Schrijver (2002), during the second world war where, the US wanted
to know the amount of supply the Soviet Union could get through its rail network and a year after
Lester R. Ford, Jr.(2010) and Delbert R. Fulkerson (1962) created the first known algorithm Ford
Fulkerson (1956) which explained the concept of maximum flow min cut and showed a prove about
its effectiveness in the real life applications, and after that the algorithm is started to be used in many
applications in life such as electrical power systems, communication networks, computer networks
and logistic networks. Thus, the study of maximum flow problems has had a huge impact in the last
fifty years.
The maximum flow problems in general are considered to find the optimal or feasible flow. In the
past few years many efficient algorithms arise to solve network problems by C. Stein (2009), notably
Ford- Fulkerson (1956), Edmond’s Karp by Brilliant.org et al. (2021) and Dinic’s algorithm Dinic,
E.A. (1970); uses the concept of blocking flow; uses the shortest augmenting path approach to find
the maximum flow in directed and capacitated graph Gentilini, R., Piazza, C. & Policriti (2008), on
the other hand, the pre-flow approach Parpalea, M. (2010)is used by the push-relabel
algorithm of Goldberg and Tarjan (1993); and the binary blocking flow algorithm of Goldberg and
Rao. The augmenting path is constrained to find the flow in a directed capacitated network from
source (s) to sink (t) and by considering that the flow must not exceed the capacity of the edge in
residual network.
With the increase number of transportation networks, the number of researches that studies the
road network from different aspects has increased significantly, for instance the study implemented in
the western of Moldavia by Daniel, Tudora & Eva, Mihail. (2015). tries to assess the nodes to
formulate a road network and another node network research apply shortest path approach to a road
network in Malaysia. But, there are limited amount of transportation that consider the capacities of the
routes as the main objective of the study, the reason is due to the existence of many conditions which

1
Şeyme Safadi, Gaziantep Branch of Association of Independent Doctors, Gaziantep, Turkey, [email protected]
2
Alptekin Durmuşoğlu, Gaziantep University, Department of Industrial Engineering, Gaziantep, Turkey, [email protected]
3
Eren ÖZCEYLAN, Gaziantep University, Department of Industrial Engineering, Gaziantep, Turkey, [email protected]
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

33
affect the transportation such as rain and car accidents, and considering the capacity allows us a faster
and smoother transportation with no late. Therefore, analyzing and calculating maximum traffic
capacity in road network is a significant method to make a decision if there is a need for maximum
flow solution and which method will provide the optimal solution and central processing unit time
(CPU) time needed to get this optimal solution.
Because each of the algorithms uses different approach in finding the solution, the road network, at
Kota Kinabalu, Malaysia, A. Noraini & H. Ting. (2017) is solved using four different algorithms;
mathematical model, ford Fulkerson, Depth First Search approach, Edmond’s Karp uses Breadth First
Search approach while Dinic’s algorithm Y. Dinitz (2006). uses level graph and blocking approach to
find the maximum flow in a graph then, the results of the algorithms are compared in term of
maximum flow or number of vehicles that can path through the network per hour to find the
maximum flow and the desired routes in the network and also the CPU time is compared between the
four algorithms to select the fastest and most optimal algorithm.
The study paper proposes an effective comparison among four main maximum flow network’s
problem-solving algorithm, it also enhance on the idea that even though the flow of the network does
not change if the approach used to find the optimal solution has changed, but the time needed to find
this solution depends significantly on the approach used.
In section 1 Explains what is maximum flow problems and its algorithms. In section 2, the
methodologies show the implementation of the four different methodologies on the network in term of
writing the Pseudo cod and forming the mathematical model. in section 3, the data are inserted and
road network is formed then the each of the algorithms is run alike. In section 4, the results of the
algorithms are compared to each other and the in term of execution time and the optimal solution.

METHODOLOGY
This paper studies, the capacitated direct network of single source (s) and single sink (t) is formed by
first define each traffic light as a node, each street or route as an edge and the capacity of each edge in
the network is defined as the maximum amount of vehicles can pass throughout the edge and it is a
non-negative value.
To find the optimal solution for this maximum flow problem, the four algorithms to be implemented
to the capacitated graph must follow the following constraints mentioned by Ahuja R.K (2001)
1. The flow of an edge must not exceed the capacity of the same edge:
G (V, E); F (u, v) ≤ C (u, v)
2. The capacity must be a non-negative value:
C (u, v) ≥ 0
3. Conservative of flow throughout the graph:
ϵV: s ≠ u and t ≠ v ∑ =∑
Where the amount that flow out of the source in a network equals the amount that flow to the sink
node.
MATHEMATICAL MODEL:
To solve maximum flow problems using mathematical model using L.R. Ford, D.R. Fulkerson (1962)
we first define decision variables which will formulate a Linear Programming model to find
maximum flow within the road network.
Decision variable is:
= the amount of flow from node i to node j
The objective function in Eq. (1) states that as maximum amount flow out of source node (s) or the
amount come into the sink node (t):
MAX ∑ (1)
The constraints for the maximum flow are assigned, therefore Eq. (3) shows that the flow of an edge
must not exceed the capacity of the same edge:
(2)
Eq. (3) proposes that the amount of flow of an edge must be a non-negative value:
(3)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

34
The total flow out of a node must equals the amount of flow to the flow excluding the source and sink
nodes is shown in Eq. (4):
∑ { } ∑ { } (4)

FORD FULKERSON:
Ford Fulkerson adapts Depth First Search Approach (1962). First, set an augmenting path from source
to sink which does not include any cycles and that pass only through positive weighted edges and it is
non-full in forward and non-empty in backward after that draw residual graph which is just a network
with the same vertices and same edges, indicates how much more flow is allowed in each edge in the
network graph. Usually, we use the residual capacities as remaining capacities to calculate the flow
easily and save time by subtracting flow from the capacity.
The previously mentioned information is transferred into pseudo code, Table 1, which helps in
formatting Java programming language code using Brilliant.org. (2021) to get the optimal flow within
the network.
The following steps are followed to find the optimal solution of this algorithm using Java
programming language such as:
Table1. Ford Fulkerson’s Pseudo Code
Pseudo code of Ford Fulkerson

The term forward edge is defined as an edge which existed in the original network, while reversed
edges are the edges which are created by residual graph.
According to time complexity which its time complexity is a formula that equals O(E*F) to solve
network problems where E is the edges and F is the flow in a network while its time complexity
depends on capacity value of edges in the graph because it only pushes one unit in edge in every
iteration
EDMOND’S KARP: The algorithm finds optimal solution to flow problems by finding the
bipartite matching while solving the problem according to time.
What makes Edmond's Karp algorithm J. Edmonds (1972) different from Ford Fulkerson algorithm is
that the technique Edmond's use is not only finding the augmenting paths but also find the shortest
augmenting paths and the approach adopted here is Breadth-First Search (BFS) which sends one by
one flow each time. So, if there are multiple augmenting paths to choose from, Edmonds-Karp will be
sure to choose the shortest augmenting path from the source to the sink every time. Table 2 displays
the mentioned constraints to generate a pseudo code of this algorithm and shows the main steps to
code the algorithm using a Java programming language from Brilliant.org (2021):
Table2. Edmond’s Karp Pseudo Code
Pseudo code of Edmond’s Karp

[ ]
[ ]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

35
 [ ]

//

[ ]
[ ] [ ]
[ ]
[ ]

The algorithm includes additional techniques that reduce the running time by using O(E ) formula
that each augmenting path can be found in O(E*F) time, that every time at least one of the E edges
becomes saturated (an edge which has the maximum possible flow), that the distance from the
saturated edge to the source along the augmenting path must be longer than last time it was saturated,
and that the length is at most V. Another property of this algorithm is that the length of the shortest
augmenting path increases monotonically, which its time complexity is no longer depend on the
capacity value of any edges.
DINIC’S: in Dinic’s algorithm Y. Dinitz (2006). BFS approach is adopted and it is similar to
Edmond’s Karp it uses shortest distance (in terms of number of edges) of the node from source to find
optimal solution, while the difference comparing to Edmond’s Karp is that DA uses:
 A blocking flow of some network is such a flow that every path from s to t contains
at least one edge which is saturated by this flow. Note that a blocking flow is not
necessarily maximal.
 A layered network of a network GG is a network built in the following way. Firstly,
for each vertex v we calculate level[v] - the shortest path (unweighted) from s to this
vertex using only edges with positive capacity. Then we keep only those edges (v,
u) for which level[v]+1= level[v]+1=level[u]. Obviously, this network is acyclic.

We use BFS to check if more flow is possible and to construct level graph Table 3 from
GeeksForGeeks.org (2021). In level graph, we assign levels to all nodes, level of a node is shortest
distance (in terms of number of edges) of the node from source. Once level graph is constructed, we
send multiple flows using this level graph. This is the reason it works better than Edmond Karp. In
Edmond Karp, we send only flow that is send across the path found by BFS:

Table 3. Pseudo Code of Dinic’s Algorithm

The pseudo code of Dinic’s

[ ]
[ ][ ]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

36
 [ ]
[ ][ ] [ ][ ]

[ ]
[ ][ [ ]]
[ ]
{

}
}
}
{

[ ]

{
[ ] [ ]
[ ][ [ ]] [ ] [ ]
{
[ ][ [ ]]
[ ][ [ ]]
[ ] [ ]
}
}
}

Time complexity for Dinic’s algorithm is Doing a BFS to construct level graph takes O(E) time.
Sending multiple more flows until a blocking flow is reached takes O(VE) time. The outer loop runs
at-most O(V) time. In each iteration, we construct new level graph and find blocking flow. It can be
proved that the number of levels increase at least by one in every iteration). So, the outer loop runs at
most O(V) times. Therefore, overall time complexity is O(EV2).

APPLICATION AND COMPARISON

The data used in this paper are from Dewan Bandaraya Kota Kinabalu (DBKK), a city council in Kota
Kinabalu and Google maps.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

37
 Figure1. Macroscope view of the network

In this paper, traffic signal timing manual was used to get the capacity; where the 25 nodes represent
are the locations of the intersections between streets or traffic lights, the 61 edges are the routes
connect between locations or intersection, and the capacity of the routes are assigned using direct
empirical method was used in paper Y.K. (2010). for capacity estimation to all the edges to form a
directed capacitated network which is used in this study.
Table (4) shows the collected data:
Table 4. Network’s Nodes, Edges and Capacity
Number Location Name From To capacity
1 Jalan Sepanggar S V1 914
2 Jalan UMS S V3 3103
3 Jalan Tuaran 1 V1 V2 1394
4 Jalan Tuaran 2 V2 V4 1394
5 Jalan Bypass (North) V2 V5 1905
6 Jalan Bypass (South) V5 V2 1151
7 Jalan Tuaran 3 V4 V5 2161
8 Jalan Lintas 1 V5 V8 2877
9 Jalan Tuaran 4 (North) V5 V6 2559
10 Jalan Tuaran 4 (South) V6 V5 2832
11 Jalan Paris (North) V3 V6 1533
12 Jalan Paris (South) V6 V3 2500
13 Jalan TunFuad Stephen 1 V3 V7 2000
14 Jalan TunFuad Stephen 2 V7 V10 1948
15 Jalan Istiadat (North) V7 V9 846
16 Jalan Istiadat (South) V9 V7 2120
17 Jalan Tuaran 5 (North) V9 V5 1902
18 Jalan Tuaran 5 (South) V6 V9 1648
19 Jalan Damai (North) V9 V12 1687

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

38
 20 Jalan Damai (South) V12 V9 1289
21 Jalan Tuaran 6 (North) V9 V15 1807
22 Jalan Tuaran 6 (South) V15 V9 813
23 Jalan KK Bypass (North) V10 V13 1897
24 Jalan KK Bypass (South) V13 V10 1394
25 Jalan Tunku (North) V13 V16 1482
26 Jalan Tunku (South) V16 V13 2575
27 Jalan Liaman Diki V13 V14 809
28 Jalan Tun Razk V10 V14 1995
29 Jalan Coastal V14 V17 2223
30 Jalan Kolam 1(North) V12 V15 2161
31 Jalan Kolam 1(South) V15 V12 2161
32 Jalan Kolam 2(North) V8 V12 1902
33 Jalan Kolam 2(South) V12 V8 2877
34 Jalan Lintas 2 V8 V11 2223
35 Jalan Lintas 3 V11 V20 1561
36 Jalan Penampang (North) V11 V21 1388
37 Jalan Penampang (South) V21 V11 1514
38 Jalan Tuaran 7 (North) V15 V18 3600
39 Jalan Tuaran 7 (South) V18 V4 2726
40 Jalan Tunku 2 (North) V16 V18 3757
41 Jalan Tunku 2 (South) V18 V16 1482
42 Jalan Nenas (North) V15 V16 1482
43 Jalan Nenas (South) V16 V15 1877
44 Jalan Kemajuan V18 V17 2262
45 Jalan Tunku 3 (North) V18 V21 1078
46 Jalan Tunku 3 (South) V21 V18 3757
47 Jalan Pintas 1 (North) V20 V22 1251
48 Jalan Pintas 1 (South) V22 V20 2483
49 Jalan Pintas 2 (North) V22 V23 2200
50 Jalan Pintas 2 (South) V23 V22 1251
51 Jalan Tunku 4 (North) V21 V23 2200
52 Jalan Tunku 4 (South) V23 V21 1948
53 Jalan Coastal 2 V17 V19 2559
54 Jalan Sembulan (North) V19 V23 3300
55 Jalan Sembulan (South) V23 V19 1419
56 Jalan Coastal 3 V19 V24 4428
57 Jalan Mat Salleh (North) V23 V24 895
58 Jalan Mat Salleh (South) V24 V23 2200
59 Jalan Kepayan V24 T 2426
60 Jalan Lintas 3 V20 T 1600

Then the nodes and edges are connected and the network is formed while considering the capacity of
each edge; as shown in the figure below:

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

39
 Figure2. Road Network Graph

The Mathematical Model:

The maximum flow of the road network is solved by formulating a mathematical model and solved by
using an optimization program, M.S. Bazaraa (1977), LINGO version 19.0.
Starting by assigning objective function:
Max Z=XS1 + XS3
S.T;
XS1 <= 914; XS1 >= 0; X1316 <= 1482; X1316 >= 0;
XS3 <= 3103; XS3 >= 0; X1417 <= 2223; X1417 >= 0;
X12 <= 1394; X12 >= 0; X159 <= 813; X159 >= 0;
X24 <= 1905; X24 >= 0; X1512 <= 2161; X1512 >= 0;
X25 <= 1151; X25 >= 0; X1516 <= 1482; X1516 >= 0;
X36 <= 1533; X36 >= 0; X1518 <= 3600; X1518 >= 0;
X45 <= 2161; X45 >= 0; X1613 <= 2575; X1613 >= 0;
X52 <= 1151; X52 >= 0; X1615 <= 1877; X1615 >= 0;
X56 <= 2559; X56 >= 0; X1618 <= 3757; X1618 >= 0;
X58 <= 2877; X58 >= 0; X1719 <= 2559; X1719 >= 0;
X63 <= 1533; X63 >= 0; X1815 <= 2726; X1815 >= 0;
X65 <= 2832; X65 >= 0; X1816 <= 1482; X1816 >= 0;
X69 <= 1902; X69 >= 0; X1817 <= 2262; X1817 >= 0;
X79 <= 846; X79 >= 0; X1821 <= 1078; X1821 >= 0;
X710 <= 1948; X710 >= 0; X1923 <= 3300; X1923 >= 0;
X811 <= 2223; X811 >= 0; X1924 <= 4428; X1924 >= 0;
X812 <= 1902; X812 >= 0; X2022 <= 1251; X2022 >= 0;
X96 <= 1648; X96 >= 0; X20T <= 1600; X20T >= 0;
X97 <= 2120; X97 >= 0; X2111 <= 1514; X2111 >= 0;
X912 <= 1687; X912 >= 0; X2118 <= 3757; X2118 >= 0;
X915 <= 1807; X915 >= 0; X2123 <= 2200; X2123 >= 0;
X1013 <= 1897; X1013 >= 0; X2220 <= 2483; X2220 >= 0;
X1014 <= 1995; X1014 >= 0; X2223 <= 2200; X2223 >= 0;
X1120 <= 1561; X1120 >= 0; X2322 <= 1251; X2322 >= 0;
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

40
X1121 <= 1388; X1121 >= 0; X2321 <= 1948; X2321 >= 0;
X128 <= 2877; X128 >= 0; X129 <= 1289; X129 >= 0;
X1215 <= 2161; X1215 >= 0; X1310 <= 1364; X1310 >= 0;
X1314 <= 809; X1314 >= 0; X2319 <= 1419; X2319 >= 0;
X2324 <= 894; X2324 >= 0; X2423 <= 2200; X2423 >= 0;
X24T <= 2426; X24T >= 0;

XS1 - X12 = 0;
X12 + X52 - X24 - X25 = 0;
XS3 + X63 - X36 - X37 = 0;
X24 - X45 = 0;
X25 + X45 - X56 - X58 = 0;
X36 + X56 + X96 - X63 - X65 - X69 = 0;
X37 + X97 - X710 = 0;
X58 + X128 - X811 - X812 = 0;
X79 + X69 + X129 + X159 - X97 - X96 - X912 - X915 = 0;
X710 + X1310 - X1013 - X1014 = 0;
X811 + X2111 - X1121 - X1120 = 0;
X812 + X912 + X1512 - X128 - X129 - X1215 = 0;
X1013 + X1613 - X1310 - X1316 - X1314 = 0;
X1014 + X1314 - X1417 = 0;
X915 + X1215 + X1615 + X1815 - X159 - X1512 - X1518 - X1516 = 0;
X1316 + X1516 + X1816 - X1613 - X1615 - X1618 = 0;
X1417 + X1817 - X1719 = 0;
X1618 + X1518 + X2118 - X1815 - X1816 - X1821 - X1817 = 0;
X1719 + X2319 - X1923 - X1924 = 0;
X1120 + X2220 - X2022 - X20T = 0;
X1121 + X1821 + X2321 - X2111 - X2118 - X2123 =0;
X2022 + X2322 - X2220 - X2223 = 0;
X1923 + X2123 + X2223+ X2423 - X2319 - X2321 - X2322 - X2324 = 0;
X194 + X2324 - X2423 - X24T = 0;

And the maximum flow of the traffic network by using a mathematical model to solve it is 4017:
From the source, we have flow of 4017 while the flow for the sink could be 4026 so we choose the
minimum capacity which is the flow out of the source. And the result as shown in figure 4, the CPU
time is 110000000 nanoseconds, are obtained using a computer with specification processor Intel core
i5 5200u cpu @ 2.20ghz and an operating system of 64-bit and 12.0 Gb rams.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

41
 Figure 3. Mathematical Model’s results

Figure 4. Augmenting Paths’ of Mathematical Model

Ford Fulkerson:
By implementing L.R. Ford, D.R. Fulkerson (1962) to the same road network problem I by using Java
programming language, where the coding process used Depth first search approach during the
detecting the augmenting paths and according to the pseudo cod mentioned previously.
The results of this method showed the same amount of flow in the network which is 4017 while the
time of the process is significantly different, where the CPU time 1482642 milliseconds
Augmentation path
S --> 1 --> 2 --> 5 --> 8 --> 11 --> 20 --> 25
bottleneck (min flow on path added to max flow) = 914

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

42
Augmentation path
S --> 3 --> 6 --> 5 --> 8 --> 11 --> 20 --> T
bottleneck (min flow on path added to max flow) = 647

Augmentation path
S --> 3 --> 7 --> 10 --> 14 --> 17 --> 19 --> 24 --> 25
bottleneck (min flow on path added to max flow) = 1948

Augmentation path
S --> 3 --> 6 --> 5 --> 8 --> 11 --> 21 --> 23 --> 24 --> 25
bottleneck (min flow on path added to max flow) = 478

Augmentation path
S --> 3 --> 6 --> 5 --> 8 --> 11 --> 21 --> 23 --> 22 --> 20 --> 25
bottleneck (min flow on path added to max flow) = 30

Figure 5. Augmenting Paths’ of Ford Fulkerson

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

43
Edmond’s Karp:
By using java compiler, a code is written and implemented using time complexity approach and
breadth-first search shown in the attached file and the maximum flow Edmond’s Karp algorithm is
obtained and the maximum number of vehicles per hour is calculated.
The maximum number of vehicles is 4017 vehicles per hour and the augmenting paths shown and the
program took CPU of 91808458 nanoseconds to calculate the maximum flow of the graph with 25
nodes and 60 edges

Augmentation path
S --> V1 --> V2 --> V5--> V8--> V11--> V20 --> T

Augmentation path

S --> V3 --> V6--> V5--> V8 --> V11-->V21--> V23-->V24--> T

Augmentation path

S--> V3-->V7-->V10-->V14--> V17--> V19--> V24-->T

Figure 6. Augmenting Paths of Edmond’s Karp Algorithm

Dinic’s Method:
In Edmond’s Karp algorithm, we use BFS to find an augmenting path and send flow across this path.
In Dinic’s algorithm by Asano & Yasuhito. (2000), BFS is used to check if more flow is possible and
to construct level graph.
 In level graph, we assign levels to all nodes, level of a node is shortest distance (in terms
of number of edges) of the node from source. Once level graph is constructed, we send

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

44
 multiple flows using this level graph. This is the reason it works better than Edmond
Karp. In Edmond Karp, we send only flow that is send across the path found by BFS.
The results of the code which is written using CPP programming language and the maximum number
of vehicles that can flow within the network is of the same, which means the algorithm shows a
maximum flow of 4017 vehicles/ hour but the code does not show the augmenting paths, while the
CPU time is 12 nanoseconds to find the optimal solution.

Finally, the maximum flow of the road network is obtained using four different algorithms,
Mathematical model using LINGO optimization package, Ford Fulkerson and Edmond’s Karp using
Java programming language and Dinic’s using CPP programming language. All the four methods are
run using a computer with specification processor intel core i5 5200u cpu @ 2.20ghz and an operating
system of 64-bit and 12.0 Gb rams, have the same maximum flow which is 4017 vehicles per hour
while the is a significant difference in the CPU time, Mathematical model needed 11 nanoseconds to
find the optimal solution for 25 nodes and 60 edges for road network, while Ford Fulkerson took CPU
time 1482642 nanoseconds which is the most amount of CPU time because it uses DFS approach with
time complexity of O(E * F), Edmond’s Karp find the solution taking 91808458 nanoseconds by
using BFS approach which its time complexity is O (V * E^2) and Dinic’s algorithm spent 12
nanoseconds by using the same approach such as Edmond’s Karp but its time complexity is O (V^2 *
E) to find the optimal solution.

CONCLUSION
The study paper proposes four different maximum flow problem based approaches, the objective of
this approaches is to compare between these approaches in term of maximum number of vehicles per
hour and the execution time CPU to obtain the result, the four algorithms are namely mathematical
model, Ford Fulkerson, Edmond’s Karp and Dinic’s algorithms, starting by applying the mathematical
model on the road network which consists of 25 nodes and 60 edges, the model is formulated and the
results are compared with the other method, then the same data are introduced to Ford Fulkerson
algorithm which is uses depth first search approach is written using Java programing language to
obtain the results and by using the same programming language Edmond’s Karp which adopt Breadth
first search is implemented and Dinic’s algorithm has shown its results using level graph and blocking
approach during the process of finding the maximum flow within the network.
The maximum flow problems have shown a wide range of application in various of life fields, that led
to arise of new algorithms with different approaches, according to this study paper, Ford Fulkerson
and Mathematical Mode were the simplest methodologies to follow in solving network road
transportation and the easiest to code using optimization package or programming language and both
gave the same optimum solution, but on the other hand with the increase number of nodes and arcs of
the network, the CPU time to find the optimal solution increased significantly because is that it took a
long time for the correct nearest distance to surface from the calculations. Another main difference we
can notice is that, however Dinic’s algorithm and Edmond’s Karp both use the same approach, the
approach is used in different way, Dinic’s algorithm sends several flow at the same time using level
graph and use BFS to check if more flow is possible and to construct level graph, while Edmond’s
Karp algorithm sends only one flow each time and this is why the complexity time of Dinic’s took 12
nanoseconds, while Edmonds Karp has a CPU time which is 91808458 nanoseconds comparing
with Ford Fulkerson and the mathematical model. The study also shows that the change of the
algorithms does not affect the flow in the road network, where all the algorithms showed the same
optimum solution.
The proposed maximum flow problem approaches will give the same results according to capacity of
the network used in the research. The proposed algorithms will give different CPU time even if the
number of nodes and edges are the same. By means of these algorithms proposed in this study, the
approach which every algorithm uses to get the final result impacts the execution time to run the
process CPU time of the problem.in this concern, the following work can be regarded as future
research direction, when the running time is an important and inevitable factor which has a significant
impact on the process of maximum flow problem.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

45
 REFERENCES

A. Noraini & H. Ting. (2017). Traffic congestion problem of road networks in Kota Kinabalu via
network graph Sabah, Malaysia Retrieved from Faculty of Science & Natural Resources, Universiti
Malaysia Sabah.
Ahuja, R. K., Magnanti, T. L., & Orlin, J. B. (1993). Networks Flows. New-York: Prentice-Hall
Massachusetts, US. Sloan school of management of institute of technology.
Ahuja R.K., Magnanti T.L., Orlin J.B. (2001) Maximum Flow Problem. In: Floudas C.A., Pardalos
P.M. (eds) Encyclopedia of Optimization. Springer, Boston, MA. https://doi.org/10.1007/0-306-
48332-7_275.
Asano, Takao & Asano, Yasuhito. (2000). RECENT DEVELOPMENTS IN MAXIMUM FLOW
ALGORITHMS. Journal of the Operations Research Society of Japan. 43. 2-31. 10.15807/jorsj.43.2.
C. Stein, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest (2009) ‘Maximum Flow’
Introduction to Algorithms Third Edition Cambridge, Massachusetts London, England708-714.
Daniel, Tudora & Eva, Mihail. (2015). A geographical methodology for assessing nodality of a road
network. Case study on the western Moldavia. Acta geographical Slovenica. 54. 10.3986/AGS54107.
CASE STUDY ON THE WESTERN MOLDAVIA A geographical.
Dinic, E.A. (1970) Algorithm for Solution of a Problem of Maximum Flow in a Network with Power
Estimation. Soviet Math Doklady ,11,
12771280.http://www.citeulike.org/user/Fujisaki/author/Dinic:EA.
Dinic’s Algorithm. GeeksForGeeks.org retrieved 13.21, April.12,2021 from
https://www.geeksforgeeks.org/dinics-algorithm-maximum-flow/.
Edmonds-Karp Algorithm. Brilliant.org. Retrieved 13:19, April 12,
2021, from https://brilliant.org/wiki/edmonds-karp-algorithm/.
Ford, L., & Fulkerson, D. (1956). Maximal Flow Through a Network. Canadian Journal of
Mathematics, 8, 399-404. doi:10.4153/CJM-1956-045-5
Ford-Fulkerson Algorithm. Brilliant.org. Retrieved 13:18, April 12,
2021, from https://brilliant.org/wiki/ford-fulkerson-algorithm/
Gentilini, R., Piazza, C. & Policriti (2008) A. Symbolic Graphs: Linear Solutions to Connectivity
Related Problems. Algorithmica 50, 120–158 https://doi.org/10.1007/s00453-007-9079-5
J. Edmonds and R.M. Karp. 1972. Theoretical Improvements in Algorithmic Efficiency for Network
Flow Problems. <i>J. ACM</i> 19, 2 (April 1972), 248–264.
DOI:https://doi.org/10.1145/321694.321699
L.R. Ford, D.R. Fulkerson, Flows in Networks, Princeton Univ. Press, Princeton, NJ, 1962.Method.
doi:10.4153/CJM-1956-045-5
L.R. Ford, D.R. Fulkerson (1962), Flows in Networks, Princeton Univ. Press, Princeton, NJ,
1962.Method.US Retrieved from https://www.rand.org
M.S. Bazaraa, J.J. Jarvis (1977) Linear programming. 2. Network analysis (planning), school of
industrial and systems engineering georgia institute of technology retrieved from http://www.b-
farhadinia.ir
Parpalea, M. (2010); Min-Max algorithm for the parametric flow problem, Bulletin of the
Transylvania University of Brasov, Series III: Mathematics, Informatics, Physics, 3(52):191–198

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

46
Ramlan, Rohaizan & Shamsuddin, A. & Shamsuddin, Fahd. (2012). Shortest path performance
within real road network: Case study in Malaysia. 6. 278-281.
Schrijver, A. (2002). On the history of the transportation and maximum flow problems. Mathematical
Programming, 91(3), 437–445. doi:10.1007/s101070100259.
Teo, K.T.K. & Kow, W.Y. & Chin, Y.K.. (2010). Optimization of Traffic Flow within an Urban
Traffic Light Intersection with Genetic Algorithm. 172 - 177. 10.1109/CIMSiM.2010.95.
Y. Dinitz (2006). "Dinitz' Algorithm: The Original Version and Even's Version"
Negev. Israel Retrieved University of the Ben-Gurion Negev. Israel

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

47
 IMPROVING SERVICES WITH AUGMENTED REALITY IN AIRPORTS:
FACILITATING AIRPORT NAVIGATION AND IMPROVING CUSTOMER
SERVICE
Sercan Demir, Ersin Aktas, Turan Paksoy

Abstract ⎯ The aviation sector has been rapidly evolving to better meet passenger expectations over the last
decade. Improving the passenger service level during flight and airport operations is crucial for commercial
airlines' profitability. Airlines have begun to focus on technology-based solutions to enhance their passengers’
experience as the competition is getting stiffer in recent years. Augmented reality (AR) is one of these emerging
technologies deployed at the airport to increase passengers’ satisfaction. AR provides an efficient and cost-
effective way to improve the operational efficiency of airport service with its ability to integrate computer-
generated data and the real-world environment. Merging digital information with a passenger’s environment
via a smart device brings great convenience before and during flight. Some AR applications in airports include,
but are not limited to, navigating passengers to the check-in counters and gates, informing passengers about
the nearby shops and discounts, cabin luggage verification, guiding passengers between various terminals, and
entertaining passengers during long layovers. This paper will explore the AR applications in airports and the
potential benefits for commercial airliners and provide case studies from real-life examples.

Keywords ⎯ Airport Navigation, Augmented Reality, Aviation Safety, Digital Technology

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

48
 1 AUGMENTED REALITY TECHNOLOGY

Augmented reality (AR) is one of the fundamental technologies of Industry 4.0, and it is an important
research area in the field of computer science. AR combines real-world and computer-generated data
on a screen of a device. An AR application uses elements such as a display, a camera, and application
software on a computer to combine and align real and virtual objects. Devices such as camera phones,
handheld PCs, laptops, and head-mounted displays can run AR applications. AR applications improve
a person’s grip and awareness regarding her surroundings by overlapping useful information on the
screen of a device (Demir et al., 2020). The goal of AR is to simplify people’s lives by merging virtual
information with a view of the real-world environment on a screen. AR improves the users’ perception
and interaction with the real environment (Carmigniani & Furht, 2011).
Sword of Damocles was the first ever created head-mounted display (HMD) system designed and
created by Ivan Sutherland and Bob Sproull at Harvard University in 1968. Sword of Damocles was
the first prototype of the AR system, and it could create three-dimensional components, then
superimpose them on the real environment (Billinghurst et al., 2015). Figure 1 shows the history and
development of phases AR in a timeline.

Figure 1: History and Development of Phases Augmented Reality (Demir et al., 2020)

The reality—virtuality continuum, which presents the mixture of objects classes, was defined by
Milgram and Kishino (1994). The left side of the continuum represents real environments, while the
right side of the continuum represents virtual environments. The left side of the continuum, real
environments, defines environments consisting of only real objects which can be viewed directly
without any electronic device. The right side of the continuum, virtual environments, includes only
virtual objects. Real and virtual environment objects coexist within a single display in mixed reality
(MR) environment. Any object located between the extreme points of Milgram’s reality—virtuality
continuum is a member of MR (Milgram and Kishino 1994). Figure 2 depicts the virtuality continuum.

Figure 2: Virtuality Continuum (Milgram and Kishino 1994)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

49
Azuma (1997) explores six potential AR applications in his paper: medical visualization, maintenance
and repair, annotation, robot path planning, entertainment, and military aircraft navigation and
targeting. Billinghurst et al. (2015) investigate the AR application in marketing, medicine, education,
entertainment, and architecture. Demir et al. (2020) discuss the companies, such as DHL, Boeing and
Airbus Case, IKEA, and Coca Cola, that highly utilize AR applications.
As Industry 4.0 is reshaping our lives from manufacturing to education, many emerging
technologies are taking the place of the existing ones. AR is one of the key technologies that
Industry 4.0 built upon. AR offers a rapid, efficient, and productive solution in many industries
such as manufacturing, health, transportation, education, entertainment, and hospitality. As
companies adopt AR applications to their processes and services, the customer satisfaction rate
increases while nonproductive time in processes is eliminated. AR applications are redesigning
many operations at airports. These operational enhancements include navigating passengers to
the check-in counters and gates, informing passengers about nearby shops and discounts,
confirming cabin luggage, guiding passengers between various terminals, and entertaining
passengers during long layovers. In the next section, we will discuss the prominent AR
applications at airports.

2 AUGMENTED REALITY APPLICATIONS IN AIRPORTS

2.1 AR Navigations at Airports

Google Maps introduces AR indoor navigation system in airports that help passengers find their way
in an airport. “Indoor Live View” application uses the phone’s camera to provide walking directions
superimpose onto the image of surroundings in real-time (Waldek, 2021). Google’s Live View
augmented reality (AR) is an extra feature added to Google maps. It will help users to find a specific
store or terminal at airports. The use of the feature is not limited only to airports; it works in malls and
transit stations too. The application helps users find check-in counters, baggage claim centers,
restrooms, elevators, ATMs, and more (Hendrickson, 2021). Another feature added to the application
is the new weather and air quality layers that provide weather information and air quality in the region.
Also, the estimated trip times for different transportation modes are part of the application
(Carrasqueira, 2021).

Figure 3: Google Maps Live View (Carrasqueira, 2021)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

50
Airport AR applications help the passenger optimally plan their time by informing them about
departure times and possible flight delays. Additionally, these apps provide information about baggage
registration, locations, flight check-in counters, parking reservations, restaurants, and nearby stores.
AR applications can measure the size of the cabin bag and let the users if the size of their bag is within
the permitted dimension. Hence, passengers avoid the last-minute hassle before landing (Sharma,
2020).

Figure 4: AR App to determine whether cabin luggage is within permissible limits (Sharma, 2020)

2.2 AR Applications in Handling Services

SATS, a gateway service provider at Singapore Changi Airport, cuts loading time for baggage and
cargo from an hour to about 45 minutes using smart glasses. The company has trained 600 ramp
handling employees to use AR glasses. Their ramp handling employees use AR glasses to scan QR
codes placed on cargo containers to see the baggage loading instructions (flight number, placement
location, etc.) in real-time. The AR glasses replace the paper-based loading; hence it will dramatically
reduce loading time for each flight. Ramp handling staff can easily look at QR codes on baggage and
cargo containers and instantly navigate information that pops up on small screens. For instance, the
weight and loading sequence of luggage and containers (Abdullah, 2017). The control center staff can
see what the ramp handler sees in real-time since cameras are embedded in the AR glasses. This AR
application will help ease employees' jobs, and passengers at Singapore's Changi airport will have
shorter waiting times before their flights (Tan, 2017).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

51
 Figure 5: AR glasses for airport ramp handling operations (Tan, 2017)

The AR system used in Singapore's Changi airport results in a reduction in aircraft turnover times
which leads to benefits for the airlines by increasing the aircraft’s hours of use, for the airport by
reducing capacity problems on the platform, and for the passengers by reducing the probability of
delays (Perez, 2021).

3. AR Application for Passenger Entertainment

John f. Kennedy airport has started a VR and AR-based service called PeriscapeVR that has 12 stations
that are designed for entertainment purposes. The entertainment center is located in a retail lounge in
one of the terminal buildings. Games last from five to ten minutes and cost between $1 and $2 per
minute. Passengers wear HP Windows MR headsets and experience anything from reef migration and
whale encounters to exploring imaginary worlds in deep space. Hence, the passengers have an option
to stay away from boredom while waiting for their flight at JFK airport (Robinson, 2018).

Figure 6: The PeriscapeVR center in JFK Airport (Robinson, 2018)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

52
 3 CONCLUSIONS

AR technology has impacted and become a part of our daily life since the beginning of the last
decade. AR technology superimposes computer-generated data or images on a user’s view of the real
world. It combines our surrounding with useful information that ease tasks, hence save time and
energy. AR has many application fields such as interactive gaming (Pokemon Go), more efficient
shopping experience (AR mirrors), maintenance, education, architecture, and sports. The vast
application fields of AR make this technology an important investment opportunity. AR increases
engagement and interaction with the surroundings while enriching the user experience.
Airports are one of the facilities AR has been implemented and yields productive results. AR
applications help passengers to navigate the check-in counters and gates, provide information about
the nearby shops and discounts, verify the size of cabin luggage, guide passengers between various
terminals, and offer entertainment options during their layover. The merits of AR will continue to
revamp airport operations and offer a satisfactory passenger experience. As AR technology advances,
it will offer more applications that will create convenient travel solutions.

REFERENCES

Abdullah, Z. (2017). Sats to boost operations with smart glasses. Retrieved 25 August 2021, from
https://www.straitstimes.com/business/sats-to-boost-operations-with-smart-glasses

Azuma, R. T. (1997). A survey of augmented reality. Presence: teleoperators & virtual environments, 6(4),
355-385.

Billinghurst, M., Clark, A., & Lee, G. (2015). A survey of augmented reality.

Carrasqueira, J. (2021). Google Maps gets improved indoor navigation with Live View and more. Retrieved 25
August 2021, from https://www.neowin.net/news/google-maps-gets-improved-indoor-navigation-with-live-
view-and-more/

Demir, S., Yilmaz, I., & Paksoy, T. (2020). Augmented Reality in Supply Chain Management. In Logistics 4.0
(pp. 136-145). CRC Press.

Carmigniani, J., & Furht, B. (2011). Augmented reality: an overview. Handbook of augmented reality, 3-46.

Hendrickson, J. (2021). Google Maps Adds Indoor AR Navigation for Malls and Airports. Retrieved 25 August
2021, from https://www.reviewgeek.com/75736/google-maps-adds-indoor-ar-navigation-for-malls-and-
airports/

Milgram, P., & Kishino, F. (1994). A taxonomy of mixed reality visual displays. IEICE TRANSACTIONS on
Information and Systems, 77(12), 1321-1329.

Perez, A. (2021). Augmented Reality in Airports. Retrieved 26 August 2021, from

https://aertecsolutions.com/en/2019/01/14/augmented-reality-in-airports/

Robinson, D. (2018). JFK Airport is using VR and AR to entertain passengers - will more follow?. Retrieved
27 August 2021, from https://www.ns-businesshub.com/science/jfk-airport-vr-ar-entertain-passengers/

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

53
Sharma, A. (2020). Enhancing passenger experience in aviation with augmented reality. Retrieved 27 August
2021, from https://www.nagarro.com/en/blog/enhancing-airline-passenger-experience-augmented-reality

Tan, Y. (2017). The world's best airport has handed out smart glasses to ground staff. Retrieved 25 August 2021,
from https://mashable.com/article/changi-airport-ar-sats

Waldek, S. (2021). Google Maps Introduces AR Navigation in Airports, But Is It Helpful?. Retrieved 25 August
2021, from https://www.tripsavvy.com/google-ar-navigation-review-5120744

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

54
 A NEW ERA IN MARKETING: FACIAL RECOGNITION PAYMENT IN
RETAIL INDUSTRY

Sercan Demir, Ersin Aktas, Turan Paksoy

Abstract ⎯ The shopping experience is evolving rapidly with the advancement in digital technology. As
retailers transform shopping landscapes using the merits of digital technology, consumer behaviors are also
influenced by this technology. Online shopping replaces the traditional shopping environment, and retailers
are looking for opportunities to improve the customer shopping experience, such as convenient and secure
payment. Different payment methods, such as e-wallets, offer customers a convenient, fast, and easy in-store
or online payment experience. However, security is still an important issue in shopping transactions. Facial
recognition technology offers secure payment options, and these systems are gradually taking place in our
daily lives. Facial recognition technology is capable of matching human faces with stored images in a
database. For instance, China has already initiated the facial recognition payment method at retail stores. A
payment device scans the faces, recognizes the consumers, and payments are processed through an associated
bank account of the customers. Facial recognition payment offers improved customer security by minimizing
financial crime concerns, rapid identity verification, convenient paying, and substituting tickets for
transportation payments. This paper investigates facial recognition payment, its advantages in the retail
industry, and its future applications.

Keywords ⎯ Face Recognition Technology, Facial Recognition Payment, Digital Payment, Digital
Technology, Retail Industry

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

55
 1 INTRODUCTION

Applications of facial recognition technology are pretty new to us. However, the applications of this
technology are growing rapidly and find practice areas from high-tech companies to airlines. Many
enterprises are willing to engage facial recognition systems in their products and services due to their
potential to add convenience to the customer experience. The market value of the facial recognition
systems was the US $3.2 billion in 2019, and it is expected to reach US$7 billion by 2024 (Ar,
2020). Facial recognition technology (FRT) has arisen as a practical solution to address identification
and identity verification problems. This technology uses biometric systems and tries to reveal the
identity by analyzing distinctive features of the body (Introna & Nissenbaum 2010).
Biometrics are unique physical characteristics, such as fingerprints, that can be used for automated
recognition (Biometrics, 2021). There are two types of biometrics: behavioral and physiological.
Physiological biometrics includes individually varying body characteristics such as fingerprints, the
iris, the face, and hand geometry. On the other hand, behavioral biometrics study the habits and
behaviors of a person, such as odor, signature, computer keystroke dynamics, and voice. Every
biometric measure has positive and negative features. For instance, iris recognition offers high
accuracy; however, it is expensive to implement this system. Fingerprints can be easily collected, but
the method is inappropriate when people are not eager to collaborate. The face is considered a
reliable parameter for biometric measures since recognition accuracy is high for the human face.
Everyone has a face, and it is usually easy to capture a face on a display. Hence, properties such as
collectability, universality, and acceptability make face recognition a viable option comparing to
other biometrics (Kaur et al. 2020).
Facial recognition has many application areas: biometrics, information security, access control, law
enforcement, and surveillance system. A new trend of shopping is to make payment with the merits
of facial recognition technology. This paper explores the benefits and advantages of merging facial
recognition technology into our shopping experience. This paper is organized as follows. In section
2, a detailed review of facial recognition technology will be provided. Applications of facial
recognition technology will be discussed in section 3. In section 4, new trends in retail shopping and
facial recognition payment technology will be explored by discussing the real-life applications.
Section 5 will conclude the paper with insights, actions, and future research directions.

2 FACIAL RECOGNITION TECHNOLOGY

The rapid development of facial recognition technology (FRT) is attributed to the advancement in
computer vision processing technology in conjunction with the improvements in digital video camera
technology. Facial recognition technology works based on extracting facial features captured on a
digital video image and matching this data with previously analyzed faces stored on a database
(Andrejevic & Selwyn, 2020).
Hamann & Smith (2019) state that FRT creates “a template” of the target’s facial image and
compares the template to photographs of preexisting images of known faces. The known face images
can be found in driver’s license databases, government identification records, mugshots, or social
media accounts, such as Facebook.
Automated facial recognition technology started in the 1960s after the invention of the first
semiautomated facial recognition system. The system was used to calculate distances and ratios in a
human face feature which an observer pinpointed. The comparison of the images was based on a
reference point, which was common for a facial feature, and calculated distances and ratios. Woody
Bledsoe, Helen Chan Wolf, and Charles Bisson initiated the automated facial recognition technology
in 1964-1965. Nowadays, research on facial recognition systems has become an important area for

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

56
researchers and industry, and many FRT applications are being implemented daily (Kaur et al.,
2020). The workflow of a basic face-recognition system is illustrated in figure 1.

1. Image
Capturing

5. Person
2. Face
Verification and
Detection
Identification

4. Database 3. Feature
Matching Extraction

Figure 1: Workflow of a basic face-recognition system (adapted from Kaur et al., 2020)

The basic steps of an FRT shown in Figure 1 are clarified below (Kaur et al., 2020):

1. Image Capturing: An image might be captured by a still or CCTV camera with or without the
person’s permission or knowledge. This captured image is also called probe image.
2. Face Detection: The face of the subject is detected among the whole captured images.
3. Feature Extraction: The unique facial features are extracted from the detected face. Face
template is generated in this step.
4. Database Matching: The features of the detected face in the previous step are matched with
the corresponding images in the database. If the facial features on the face template match
properly, verification of the person is successful.
5. Person Verification and Identification: After the success of verification (face detection), the
last step is identifying the subject. The identification of a person relies on the face-detection
abilities of the used recognition algorithms.

In the early development stages of the FRT during the 1970s, it was treated as a 2D pattern
recognition problem. Measuring the distance between eyes and other points was the main focus for
researchers. Also, the development of an automatic face recognition system was necessary. FRT
became a common research area among researchers from different disciplines such as psychology,
computer science, and law enforcement. The methods below are used in face recognition systems
(Parmar & Mehta, 2014). Various face recognition methods can be classified into three main groups:
holistic matching, feature-based (structural), and hybrid methods.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

57
2.1 Holistic Matching Method

This method uses the whole face region as input data for face catching system. Eigenfaces, the most
widely used method for face recognition, principal component analysis, linear discriminant analysis,
and independent component analysis, are the most known examples of the holistic matching method
(Parmar & Mehta, 2014).

2.2 Feature-based (structural) Methods

According to Parmar & Mehta (2014), local features such as eyes, nose, and mouth are first
extracted, and their locations and local statistics are fed into a structural classifier in these methods.
This method analyzes local features such as eyes, nose, and mouth and their geometric relationships.

2.3 Hybrid Methods

Hybrid face recognition systems employ both holistic and feature extraction methods. 3D Images are
used in hybrid methods. These 3D systems work on five steps: detection, position, measurement,
representation, and matching. The image of a person's face is caught in 3D allow the system to
capture curves of the eye sockets, the shapes of the chin or forehead (Parmar & Mehta, 2014).
Human facial features are the main subject in face recognition studies. Researchers state that eyes,
mouth, and nose are amongst the most crucial feature for recognition. Feature points such as eyes,
nose, and mouth are extracted and used as input data in applications by some image processing
techniques (Bakshi & Singhal, 2014). A geometric representation of a human face is given in figure
2.

Figure 2: A Representational Image of Aspects Highlighting Uniqueness of a Face (Dangi, 2011)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

58
 3 APPLICATIONS OF FACIAL RECOGNITION TECHNOLOGY

Face recognition is an important research area and has been maintaining its popularity since its early
development years. Researchers and practitioners have been focusing on the prominent applications
of FRT, some of which are shown in figure 3.

Law
Enforcement

Security and
Forensic
Surveillance

Facial
Recognition
Mobile Technology
Applications
Health

Reconstrution Identification

Figure 3: Prominent applications of Facial Recognition Technology

• Law enforcement applications: FRT is used to create mug-shot albums for static matching and
operate video surveillance systems for real-time matching by video image sequences (Kaur et al.,
2020).

• Security and Surveillance applications: Public security and surveillance applications are one of
the most prominent fields for the applications of FRT. For instance, facial biometric scanning
applications are increasingly being adopted at border controls in train stations and airports worldwide
(Zuo et al., 2019). FRT is used to access restricted areas where only specific people have the
authorization to enter. Technology companies develop applications that use deep learning algorithms
for fraud detection and differentiating a human face from a photograph (Kaur et al., 2020).
Parmar & Mehta (2014) report 300 cameras linked to the closed-circuit TV (CCTV) controller room
in Newham Borough of London to strengthen city center surveillance. As the city council states, the
crime rate has dropped 34 % since implementing the technology.

• Health applications: Medical applications of FRT have been expanded in recent years. As
reported by Oxford University, facial analysis software can detect dysmorphic craniofacial features
suggestive of syndromic genetic disorders. The algorithm, which works based on unsupervised
machine learning techniques, has been trained with thousands of face images to learn diagnostically
relevant phenotypic features. The software can detect the faces of patients suffering from
developmental disorders such as Angelman, Apert, Cornelia de Lange, Down, Fragile X, Progeria,
Treacher-collins, and Williams-Beuren (Zuo et al., 2019).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

59
• Identification applications: Face recognition systems can identify people by their face images.
Facial recognition technology detects the presence of an authorized person instead of just checking
whether a valid identification (ID), key, or password is being used (Parmar & Mehta, 2014).

• Reconstruction applications: If a person’s identity is lost, the facial recognition system can
compare photographs with the existing data and recreate the person’s identity (Kaur et al., 2020).

• Mobile and laptop applications: Today, pin numbers and passwords in mobile phones and laptops
are being replaced by FRT applications. The main motive behind this is to secure the users’
confidential data (Kaur et al., 2020). For instance, Zuo et al. (2019) state that Apple iPhone X uses
Face ID biometric authentication technology based on facial depth image acquisition to unlock the
smartphone. Apple’s TrueDepth camera system projects 30,000 invisible infrared spectrum points
onto the user’s face to form a depth map and a two-dimensional infrared image. Then, the captured
data are transformed and compared to the user’s enrolled facial data. This facial data is captured
from different perspectives at the time of smartphone setup. If the compared data match, the phone
unlocks.

• Forensic applications: Forensics addresses applying scientific principles to analyze data collected
by law enforcement agencies. Forensic face recognition systems introduce a group of tools that help
investigators narrow the identity of a subject. Forensic face recognition systems experience a great
improvement due to recent advances in facial aging, facial marks, sketch to photo matching, video-
based face recognition, and NIR image to photo match (Jain et al., 2011).

4 FACIAL RECOGNITION PAYMENT

As digital technology is advancing, it becomes easier for hackers to get personal details. For this
reason, many people are hesitant to make online payments. Security is a critical issue in digital
transactions. Biometric authenticity offers unique identification for every individual; hence it has
been gaining attention in digital payment research. For instance, fingerprint, hand geometry, iris,
face, and palm are biometric authentication applications. Face recognition is the most popular,
convenient, and widely acceptable technique (Aria et al., 2020). Facial recognition has been gaining
increasing popularity because of its benefits over traditional surveillance techniques, like biometrics.
Governments make huge investments for the development of FRT. China and the USA are the
pioneer countries in FRT (Wei & Rodrigo, 2021).
Gao et al. (2021) discuss two payment technologies available for customers to use at self-checkout:
QR and FR payment technologies. QR payment technology is a contactless payment technology
where a customer makes the payment by scanning a QR code from a mobile app. QR code on a
customer’s phone is directly linked to the customer’s mobile payment account. In contrast, customers
do not need to use smartphones for facial recognition payments. A customer can make the payment
transaction by standing in front of a built-in webcam on the self-checkout. The customer locates her
face within the frame on the screen of the self-checkout machine. Once the customer aligns her face,
an algorithm-based identification process runs in the backend and matches the customer in front of
the screen with the photo on the ID card stored in the database. If the match is successful, the
payment transaction is authorized (Gao et al., 2021). The demonstration of QR and FR payment is
given in figure 4.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

60
 Figure 4: Demonstration of QR and FR payment technologies (Gao et al., 2021)

Compared to QR payment, FR significantly reduces the effort and time of costumers during the
payment process because it requires fewer actions steps to be taken. The comparison between the
actions steps of QR and FR payments is shown in figure 5.

Figure 5: Action steps for QR and FR payment technologies (Gao et al., 2021)

According to Vazquez-Fernandez and Gonzalez-Jimenez (2016), as a payment method, FR payment

must possess the following characteristics to achieve an effective system:

1. Usability: Ease of use and achieving low false rejection rates are the key factors.
2. Security: the systems should prevent fraud (i.e., low false acceptance rate).
3. Availability: the verification method should be usable anywhere and at any time.

The development and widespread use of FR payment generate great value in improving customers’
shopping experience, increasing business efficiency, and boosting the smart development of the
economy and society. The face recognition payment shortens payment time by significantly reducing
checkout processes. The rapid advancement in technologies such as 5G, AI, and IT, have promoted
the FR payment. Some industries adopting FR payment in China are restaurants, shopping centers,
supermarkets, online shopping, hotels, transportation facilities (Wei & Rodrigo, 2021).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

61
Facial recognition payment is built upon FRT. According to Zhang & Kang (2019), facial-
recognition payment has the following characteristics:

1. FRT is cheap comparing iris recognition and fingerprint recognition since they require
additional devices.
2. Face recognition is one of the few biometric methods that provide both high accuracy and
low intrusiveness.
3. FR payment offers a more secure and convenient authentication method than traditional
passwords payments.
4. Simple accessibility and flexibility are major advantages of FR payment. It eliminates the
need for cash or credit cards and still offers secure payment since each person’s facial
features are different.

A comparison of three novel payment methods, apple pay, QR-code scanning, and FR payment, are
given in table 1.

Payment
Device Requirements Software Requirements Time Cost
Methods
Apple Pay or A phone that supports Credit/debit card preset in Apple
Within 15s
Android Pay facial recognition Pay
QR-code
scanning A smartphone with a An Ali Pay/WeChat account and
30s – 1min
camera an associated bank account

10 – 15s
Facial An Ali Pay/WeChat account and
None (well, a face) (For frequent users, it
recognition an associated bank account
could be less than 10
seconds)

Table 1: Comparison of novel payment methods (Liu, 2020)

Contactless payment with consumers tapping their cards onto POS machines or using their
smartphones linked to Google Pay, Apple Pay, or Samsung Pay accounts has gained huge attention
for the last decade. Besides, contactless payment helps consumers be safe during the COVID-19
pandemic in physical stores (Whitehead, 2021). Facial recognition with its potential benefits will be
more appealing in the near future. The market segment of FR payment is increasing, and it is
expected to reach the US $7 billion by 2024.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

62
 5 CONCLUSIONS

Consumers do not need wallets, cards, and cash since the development of convenient payment
allows them to pay through mobile passwords, biometrics features, or just showing their mobile
devices. FR payment offers non-contact advantages over other methods (Zhang et al., 2021). FR
payment is an inevitable result of the technological advancements starting with Industry 4.0, and it
offers convenient, free from the restrictions of mobile phone-based payment methods. Consumers are
no longer have to worry about how to pay, even if their mobile phones are not with them or powered
off. Hence, FR payment serves to satisfy modern shopping needs (Zhong et al., 2021).
FR payment technology offers potential opportunities for addressing many issues regarding
security, time, and the overall customer experience. Many countries like China and the US have
already made huge investments in developing this technology and its unique merits that will lead to
the development of the economy and society. FR payment will continue to attract attention from
researchers and practitioners because of its convenient, easy, reliable, and flexible payment
characteristics.

REFERENCES

Andrejevic, M., & Selwyn, N. (2020). Facial recognition technology in schools: Critical questions
and concerns. Learning, Media and Technology, 45(2), 115-128.

Ar, A. (2020). Can consumers benefit from facial recognition technology?. Retrieved 18 August
2021, from https://techwireasia.com/2020/03/can-consumers-benefit-from-facial-recognition-
technology/

Aria, M., Agnihotri, M. V., Rohra, M. A., & Sekhar, M. R. (2020). Secure online payment with
facial recognition using mtcnn. International Journal of Applied Engineering Research, 15(3), 249-
252.

Bakshi, U., & Singhal, R. (2014). A survey on face detection methods and feature extraction
techniques of face recognition. International Journal of Emerging Trends & Technology in Computer
Science (IJETTCS), 3(3), 233-237.

Biometrics. (2021). Retrieved 18 August 2021, from https://www.dhs.gov/biometrics

Dangi, R. (2011). Face Recognition. Retrieved 21 August 2021, from

https://www.engineersgarage.com/face-recognition/

Gao, J., Rong, Y., Tian, X., & Yao, Y. O. (2021). Save Time or Save Face? The Social Presence
Effect and Herding Effect in the Use of Facial Recognition Payment Technology in Retail. The
Social Presence Effect and Herding Effect in the Use of Facial Recognition Payment Technology in
Retail.

Hamann, K., & Smith, R. (2019). Facial Recognition Technology. CRIM. JUST, 9.

Introna, L., & Nissenbaum, H. (2010). Facial recognition technology a survey of policy and
implementation issues.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

63
Jain, A. K., Klare, B., & Park, U. (2011). Face recognition: Some challenges in forensics. In 2011
IEEE International Conference on Automatic Face & Gesture Recognition (FG) (pp. 726-733).
IEEE.

Kaur, P., Krishan, K., Sharma, S. K., & Kanchan, T. (2020). Facial-recognition algorithms: A
literature review. Medicine, Science and the Law, 60(2), 131-139.

Liu, F. (2020). Case Study of Facial-Recognition Payment in China. Retrieved 23 August 2021, from
https://www.nngroup.com/articles/face-recognition-pay/

Parmar, D. N., & Mehta, B. B. (2014). Face recognition methods & applications. arXiv preprint
arXiv:1403.0485.

Vazquez-Fernandez, E., & Gonzalez-Jimenez, D. (2016). Face recognition for authentication on

mobile devices. Image and Vision Computing, 55, 31-33.

Whitehead, J. (2021). Face Pay: Is the Future of Payment with Facial Recognition? - ArganoUV
(United Virtualities) – Tech Consultants & Salesforce Commerce Cloud Experts. Retrieved 23
August 2021, from https://weareuv.com/face-pay-is-the-future-of-payment-with-facial-recognition/

Wei, M., & Rodrigo, V. P. L. (2021). The Impact of Face Recognition Payment in the Economic. In
2021 International Conference on Diversified Education and Social Development (DESD 2021) (pp.
305-310). Atlantis Press.

Zhang, W. K., & Kang, M. J. (2019). Factors affecting the use of facial-recognition payment: An
example of Chinese consumers. Ieee Access, 7, 154360-154374.

Zhang, L. L., Xu, J., Jung, D., Ekouka, T., & Kim, H. K. (2021). The Effects of Facial Recognition
Payment Systems on Intention to Use in China. Journal of Advanced Researches and Reports, 1(1),
33-40.

Zhong, Y., Oh, S., & Moon, H. C. (2021). Service transformation under industry 4.0: Investigating
acceptance of facial recognition payment through an extended technology acceptance model.
Technology in Society, 64, 101515.

Zuo, K. J., Saun, T. J., & Forrest, C. R. (2019). Facial recognition technology: a primer for plastic
surgeons. Plastic and reconstructive surgery, 143(6), 1298e-1306e.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

64
 SUSTAINING A RESILIENCE OF COMPETITIVE ADVANTAGE
DURING COVID-19 IN MARITIME SUPPLY CHAINS:
FROM PERSPECTIVE OF RESOURCE-BASED THEORY

Kazım YENİ1, Sedat BAŞTUĞ2

Abstract ⎯ The aim of the study is to investigate what challenges supply chain companies have been
faced and how supply chain companies (ocean carriers, port terminals and other supply chain actors
etc.) have sustained the resilience of competitive advantage during Covid-19 in maritime supply chains
by exploring required resources and capabilities of firms. Additionally, the study also aims to find out
the firm competitive advantage strategy from the lessons earned during pandemic. The data collection
was carried out on the 31 supply chain stakeholders. Two types of method used in analysing a
literature and interview. For the methodology, a literature review was conducted and afterward
interviews were deeply analyzed with content analysis. The findings of study reveals that the many
firms have developed their capabilities (such as digital transformation etc.) to ensure the effective
competitive advantage during Covid-19. For the practitioners, the study will aid them to build their
resilience strategy by using their own resources and capabilities.

Keywords: Maritime Supply Chain, Resilience, Resource-Based View Theory, Competitive

Advantage, Content Analysis

INTRODUCTION
Global manufacturing companies continue their activities within the supply chains covering large
geographical areas that serve the trading system built in today's global economic system. However,
Global supply chains have always been sensitive to shocks that contains trade wars, pandemics such as
Ebola, Covid-19 and domestic instability etc. This is specifically due to the factors that could block flow
of goods and services from the exporting countries to major import trading stakeholders. Covid-19 has
disrupted the global transportation activities across all economies and industries. The disruptions are
widely because of the lockdown measures adopted and enforced by countries as a health strategy to
mitigate the impact of Covid-19 spread on the humanity. Manucfacturing halted, movement of people
and goods restricted, border closures implemented and therefore supply chain partners faced logistical
constraints, as well as the slowdown of trade and transportation activities are drops of the Covid-19
lockdown measures. With the introduction of the lockdown, the transportation industry has remained

1
Asst.Prof. Kazım YENİ, İskenderun Teknik University, Barbaros Hayrettin Naval Architecture and Maritime Faculty, Management in Marine
Transportation, [email protected], ORCID: 0000-0002-6876-9903
2
Asst.Prof. Sedat BAŞTUĞ, İskenderun Teknik University, Barbaros Hayrettin Naval Architecture and Maritime Faculty, Maritime Business
Administration , [email protected], ORCID: 0000-0002-7121-2882

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

65
partially closed even if the global supply chain activities are dependent. World Trade Organization
(WTO) predicted that global trade has slowdown from 32% to 13% at end of the year 2020. Therefore,
the disruptions has impacted the demand on logistics services, slowly. Although the demand for logistics
services is less affected by the outbreak compared to other manufacturing sectors, there are vital lessons
for the logistics industry to be drawn from the Covid-19 pandemic. The maritime industry plays a key
role in facilitating international trade within the global economic system, ship operators and ports take
their place in the global trade scene as indispensable main actors of maritime trade (Lam, 2015). The
aim of the study is to investigate what challenges maritime supply chain companies have been faced and
how they have sustained the resilience of competitive advantage during Covid-19 in maritime supply
chains by exploring required resources and capabilities of firms. Additionally, the study also aims to
find out the firm resilience strategies from the lessons earned during pandemic.

The study has structured as follows: First of all, a proposed theoretical framework on resource-based
view theory and supply chain resilience has been presented. The methodology has been presented for
introducing research design. Upon satisfactorily completion of the data analysis, the findings of the
study have been submitted and discussed. In conclusion, the results of the study have been explained.

THEORETICAL FRAMEWORK
The theoretical framework is the main body that can hold or support a theory of the research study
(Abend, 2008). In this section, all theory and fact where stands for the research aim, have been described.
Upon explaination of theories and facts, the theoretical framework has been established.

RESOURCE-BASED VIEW THEORY AND RELATEDNESS WITH COMPETITIVE

ADVANTAGE
The resource-based view theory (RBV) is a theory that examines and understand internal resources and
capabilities of the organizations and maintains resources and capabilities in developing strategy to
ensure competitive advantages (Barney, 1986; Hamel and Prahalad, 1996). In this theory, resources can
not simply moved or purchased, which need an extended learning curve or big change in the organization
climate and culture, are expected to be sole to the organization, and that’s why, it is not easy to imitate
by competitors. Regarding Conner (1991), performance variance between companies depends on its
possession of particular inputs and capabilities.

Regarding RBV theory, an organization can be seen as the group of pyhsical resources, human resources
and organizational resources (Barney, 1991; Amit and Shoemaker, 1993). The resources of an
organization are beneficial, unique, imperfectly imitable and substitutable which are the particular roots
of sustainable competitive advantage for sustainable top performance (Barney, 1991). Resources should

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

66
cover assets, organizational process, company attributes, knowledge, or information ruled by the
company which can be utilized to visualize and use their strategies (Daft, 1983; Barney, 1991; Mata et
al., 1995). However, the term “resources” are identified different by scholars, for instance, some of them
(Wernerfelt, 1984; Olavarrieta and Ellinger, 1997; Spanos and Lioukas, 2001) are defined as brand
names, technological abilities and effective procedures etc. Other researchers (Itami and Roehl, 1987;
Hall, 1992; Hall, 1993) have classified them into tangible and intangible resources. Barney (1991)
categorized them into three types of resources: (1) Physical capital resources (technological, physical,
plant and equipment), (2) Human capital resources (experience, training, insights), and (3)
Organizational capital resources (legal structure). Brumagim (1994) offers a hierarchy of resources with
four levels of corporate resources: (1) Production/maintenance (the most basic or lowest level), (2)
Administrative, (3) Organizational learning, and (4) Strategic vision (the most advanced or the highest
level).

Apart from Barney (1991), Grant (1991) has substantially made a contribution to the resource-based
theory from the perspective of competitive advantage. He mentioned that there is a unclear implications
of this theory for the strategic management because of the several reasons, including the lack of a
practical framework and a few efforts to integrate this theory with competitive advantage. Therefore, he
finally proposed a new model to make a strategy formulation which coordinates a number of key themes
in the literature. A proposed framework includes five-level procedure for strategy formulation:
analyzing the firms’s resource base; appraising the firm’s capabilities; examining the profit-earning
potential of firm’s resources and capabilities; choosing a strategy; and expanding and upgrading the
firm’s pool of resources and capabilities. Figure 1 illustrates his original framework on the theory.

Figure 1. Resource-Based Approach to Strategic Analysis for Competitive Advantage:

A Practical Framework
Source: Grant (1991)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

67
Grant (1991) identified the resources and capabilities as the foundation for competitive advantage
strategy into two categories, including source of direction and corporate profitability. At this point under
pandemic conditions, the understanding of the source of direction is more important than corporate
profitability due to the urgent customer’s needs, business sustainability and resilience. As the starting
point, source of direction answers simple questions for developing a strategy: “What is our trade?”,
“Who are the customers?”, and “Which of their requirements do we serve?”. However, the customers’
needs are volatile, and any global climate change (such as pandemic) can suddenly occurs. Externally
focused orientation may not ensure a good foundation for formulating long-term strategy. Therefore,
Levitt (1993) offered a practical solution that companies can easily identify their served markets widely
rather than narrowly. For example, major shipping lines usually perceives themselves to be in the
maritime market but not the whole transportation business. Such assessing of the target market is of
little value if the firm may not simply improve the capabilities needed for serving customer’s
requirements. Grant (1991) evaluates the companies’ resource-based strategies by looking the
determinants of sustainable competitive advantage, including (a) durability, (b) transparency, (c)
transferability and (d) replicability.

a) Durability: It decides how long the competitive advantage can be sustained and is considered in terms
of the ability of rivals to imitate through ensuring access to the resources on which the competitive
advantage is built. It considerably differs regarding different cases, such as, the speed of technological
change or the great ripple effect of epidemic on the economies (Grant, 1991).

b) Transparency: It defines the firm’s ability to sustain its competitive advantage over time depends
upon the pace with which other companies can imitiate its strategy. Imperfect transparency is the basis
of Lippman and Rumelt’s (1982) theory of “uncertain imitability”, that defines the greater uncertainity
within the market over how firms “do it”. Imitation needs to solve two problems for a rival. First
dilemma is what the competitive advantage of the successfull rival is and how can be ensured. The
second dilemma is how the rival can collect the resources and capabilities needed to imitiate the
favorable strategy of the competitor.

c) Transferability: When the company has setup the sources of the superior performance, imitation needs
amassing the resources and capabilities important for a competitive challenge. Many resources and
capabilities can not be easily transferable between companies, therefore, would be competitors are
unable to ensure the resources required to replicate the competitive advantage of the company. The
transferability is directly deal with imperfect information, geographical immobility, the immobility of
capabilities, and company-specific resources.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

68
d) Replicability: The limited transferability of resources and capabilities affects the imitation success of
the company. Some of them easily imitiated by replication. In service market, new process innovation
(such as video communication tools, digital services, shortening of the customs clearance processes) are
infamous for their fast imitation by competitors. Grant (1991) argues that hardly replicable are
capabilities based upon complicated organizational routines.

SUPPLY CHAIN RESILIENCE DURING COVID-19

Resilience is a reactive situation when disruption or shock happens (Holling, 1973). It is the sole
competence of supply chain to survive disruptions and mitigate operational capacity after disruptions
(Zavitsas et al., 2018). It is defined as a term to underline that logistics disruptions can have both
“negative” and “positive” effects. Risk is generally known as a potential future bad event (Manuj and
Mentzer, 2008).

Logistics disruptions may have long-run negative effects with various risks (see Fig. 2). They appear in
different forms. At the first stage, the functional risks may occur regularly but with minor results.
Particularly, the risks may create disturbances in the humanitarian supply chain that are not admitted to
be important if they do not happen simultaneously. At the second stage, the disruptive risks may have a
ripple or snowball effect with major results.

Figure 2. Supply chain risks

Source: Ivanov (2018)

Current trends turned supply chain into less controllable and more complex. Regarding complexity
theory, these utilizations are more delicate to disruptions. Disruptions in the humanitarian supply chain,
only in its supply base, may immediately affect the whole supply chain process. At the era of Covid-19,
the risk may bring about a disruption with a ripple effect because a risk may not be limited. Alternatively,
it may lead to damage of supply chain process. This problem is referred to a ripple effect that is a
disruption diffusion in the humanitarian supply chain that hinder supply chain performance (such as

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

69
order cycle time, perfect condition, and in stock percentage etc.). It may also have more important
consequences than just lowering short-run performance. The ripple effect is a very important stressor to
supply chain and their ongoing collapses amid pandemic coming from the disruption. Regarding Vilko
and Halikas (2012), in a supply chain, the time delay is another disruption risk, especially in the context.
It broadly damages on the nature of cargo. For example, humanitarian aid supplies are a time sensitive
goods in that a logistics delay may lead to greater loss of human life. On the contrary, a week’s stoppage
of machine spare parts would only have urgent economic and social results.

A PROPOSED THEORETICAL FRAMEWORK

The theoretical framework is based on the resource-based view theory. In this theory, resources
classified into tangible and intangible resources (see Figure 3).

-Financial Strength -Organizational Culture -Process innovations

-Land / Space -Organizational Learning -Collaboration
-Vehicles/Equipment -Employee recruitment, compensation,
-Organizational flexibility
-Buildings and other reward, and training policies
-Personel Management
physical structures -Company reputation
-Decision-making
-Cost Efficiency
-Customer Orientation
Tangible Intangible -Communication

Resources Capabilities

Sustaining A Resilience Of
Competitive Advantage During
Covid-19

Building a Firm Strategy

⁃ Durability ⁃ Transparency ⁃ Transferability ⁃ Replicability

Figure 3: A Proposed Theoretical Framework

Source: Model adapted from Barney (1991) & Grant (1991)

From the perspective of supply chain, “tangible resources” dimension consists of criteria (1) financial
strenght, (2) land/space, (3) vehicles/equipment and (4) buildings and other physical structures.
“Intangible resources” dimension contains criteria, including (1) organizational culture, (2)
organizational learning, (3) Employee recruitment, compensation, reward, and training policies and (4)
company reputation. Making a resilience of competitive advantage during Covid-19 depends firm
capabilities as well as company resources. In figure 3, the model identifies constructs including

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

70
tangible/intangible resources and capabilities of companies in supply chain network. “Firm capabilities”
dimension includes criteria: (1) process innovations, (2) collaboration, (3) organizational flexibility, (4)
personel management, (5) decision-making, (6) cost-efficiency, (7) customer orientation and (8)
communication. In table 1, all criteria are identified in the model with their sources. Every company in
supply chain has own resources and has ability to efficiently make use of its resources by capabilities.
Companies should have to align their resources and capabilities to provide a sustainable logistics
operations and it will provide competitive advantage which is to a great extent shaped by the uniqueness
of the organization’s resources and capabilities during Covid-19. Companies that has well-build
resources and capabilities, can develop their own company strategies within the framework of
competitive advantage which is utilized by determinants (durability, transparency, transferability, and
replicability).

Table 1. Resource-Based View Theory Constructs and Criteria of Resilience of competitive advantage
during Covid-19 in Maritime Industry
Constructs Criteria Sources Description
It is the ability to create profits and cash flow to fund bills and
Financial Strenght Čater and Čater, 2009
refund debt or investors.
Kamasak, 2017; Čater Land is a geographical location, infrastructure assets of a
Land / Space
Tangible and Čater, 2009 company.
Resources The term “vehicles / equipment” is defined as the pyhsical
Vehicles / Equipment Kamasak, 2017
equipments in the company assets.
Buildings and Other It refers to factories, offices, warehouses, stores, showrooms and
Kamasak, 2017
Physical Structures IT structure etc.
Kamasak, 2017;
Drnevich &
It refers to the underlying assumptions, beliefs, values and ways
Kriauciunas, 2011;
Organizational Culture of communicating that providing a contribution to the unique
Pandza & Thorpe, 2009;
social and psychological environment of an organization.
Lawson & Samson,
2001
It is the process by which an organization develops itself over
Intangible Organizational
Bueno et al. 2010 time through gaining experience and using that experience to
Resources Learning
create knowledge.
Employee Recruitment,
It refers to the process of identifying, attracting, interviewing,
Compensation, Reward, Kamasak, 2017
selecting, hiring, rewarding, training and onboarding employees.
and Training Policies
It is a total of whole views and beliefs held about the firm based
Kamasak, 2017;
Company Reputation on its history and its future prospects, in comparison to close
Rindova et al. 2010
rivals.
It refers to several innovative technologies (e.g. automation and
Kamasak, 2017,
Process Innovations digitalization etc.) for optimizing the service production of
Carraresi et al. 2004
organization.
It refers to network responsiveness, co-ordination and adaptation
Collaboration Kamasak, 2017
of networks with customers
Organizational It is defined as the ability to adapt quickly to new or changing
Fredericks, 2005
Flexibility environments for organization.
It defines the management of overall skills, innovativeness,
Personel Management Kamasak, 2017
Capabilities creativity, and know-how of employees
It refers to the skills, expertise and decision-making abilities of
Decision-making Kamasak, 2017
managers.
It is a type of business strategy which is simply act of saving
Cost Efficiency Wade & Hulland, 2004
money by performing an activity in a better way.
It is a business philosophy that puts the requirements of the
Customer Orientation Ketchen JR et al., 2007
customer over the needs of the trade.
Kamboj & Rahman, It utilizes marketing communication to gain the confidence of
Communication
2015 customers regarding positive perception about products.
Source: Authors

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

71
 METHODOLOGY
This study has two research approach, including literature review and content analysis. In the maritime
supply chain literature, the relevant keywords’ terms “maritime supply chain”, “resilience”, and
“resource-based view theory” were used in the electronic databases “Abi Inform”, “Ebsco Host”, and
“Web of Science”. The study has limited search in the following set of leading transportation and
maritime journals: “Maritime Policy and Management”, “Transportation Research”, “Transport
Policy”, “Transport Review”, “Supply Chain Management: An International Journal”, “International
Journal of Physical Distribution and Logistics Management” and “Journal of Supply Chain
Management”. They are in the same way and very popular double blind peer-reviewed. Overall, a group
of 112 articles found in the journals. However, some articles has just cited other articles that conducted
content analysis. Therefore, they were excluded from the research and afterward, 52 articles have been
set in the final sample. The identified articles were then consolidated into the section of literature review
in this study.

Following the literature section, the steps of content analysis has been illustrated in Figure 4. Neuendorf
(2002) presented a methodology that can be used to text or other data, including voice or video data.
This methodology offers a guidelines classified into four steps: (1) variable conceptualization and
operationalization, (2) coding schema, (3) training and coding, and (4) realibility checks.

Variable Conceptualization and

Operationalization

Code Book Creation Detailed

Data Collection
Coding Schemes Defines

Testing and Pilot Realibility 31

Interviews Coded

Coding Process Coding By Two

Independent Coders

Final Results Calculate Inter-Coder Realibility

Figure 4. Overview of Content Analysis Process

Source: Adapted from Craighead et al. (2007) & Neuendorf (2002)

The data source of depth-interview has been offered before conducting the content analysis. The relevant
data is collected from three types of interviewers. The total number of attendants from shipping line
operators (10 persons), non vessel operating common carrier (NVOCC, 13 persons), and port operators
(8 persons) is 31 persons in the sample. In depth-interviews were conducted on Zoom using a semi-

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

72
structured interview form which had been tested with academic and professional experts to clarify
statements.

OPERATIONALIZATION AND CONCEPTUALIZATION ON VARIABLES

A first phase in content analysis is the identification of the variables that will be operationalized.
Neuendorf (2002) argues that the operationalization of variables related to research question has the
flexibility to conduct content analysis. He also suggests that scholars should concentrate on content
validity at this phase. If variable operationalization does not fit with definitions of variables, the coding
will have errors and lead to wrong assumptions. Referring to the Figure 1, the construct of “sustain a
resilience of competitive advantage during Covid-19” is identified on the basis of resource-based view
theory. The connected constructs were already illustrated in the section of theoretical background.
Regarding this theory, this study differentiates constructs between resources and capabilities.
Additionally, resources involves tangibles and intangibles. To measure the constructs, the study has
selected the criteria (i.e. financial strenght, land / space, vehicles / equipment, buildings and other
physical structures) under the constructs.

CODING SCHEMA
Following this section, codebook has been developed with set of keywords that reflects each of the
constructs and criteria. Especially, a dictionary of keywords has been developed to capture references
on the depth-interviews. The study also defined synonyms for a dictionary of keywords to assess the
sensivity of the results with respect to how widely or narrowly the list of keywords terms are identified.
The list of comprehensive and restrictive dictionaries are presented in Table 2.

Table 2. Coding Schema

Criteria Comprehensive Dictionary Restrictive Dictionary
commercial power, financial power, financial
financial, commercial, power,
Financial Strenght stability, financial vigor, commercial health, financial
stability, health
health
Land / Space district, shore, territory, field shore, territory
reach stacker, ship to shore crane, container, crane, forklift, container, ship,
Vehicles / Equipment container ship, rubber tyred crane, mobile harbour MHC, STS, laptops, mobile
crane, IT equipment, communication equipment phones
container freight station, gate, pier, lighthouse, jetty,
Buildings and other physical
fire station house, storage area, warehouse, IT CFS, gate, pier, jetty,
structures
infrastructure, communication infrastructure
knownledge, society, beliefs, assumptions, values, beliefs, assumptions, values,
Organizational Culture
morale, morale
information, organizational
Organizational Learning information, organizational training, science
training, science
Employee recruitment,
compensation, reward, and employee training, employee rewarding training, rewarding
training policies
Company reputation prestige, history, future works prestige, history, future works
novelty in process, process renewal, process change, novelty, renewal, change, remote-
Process innovations
remote-working working

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

73
 corporate cooperation, corporate teamwork, cooperation, participation,
Collaboration
leadership teamwork, leadership
organizational elasticity, organizational adaptability,
Organizational flexibility elasticity, adaptability, resilience
organizational resilience,
talent management, human resource certification,
Personel Management talent, certification, salary,
personal salaries,
problem solving, decision-making models, decision-
Decision-making problem, model, process,
making process,
savings, capital, energy,
Cost Efficiency cost savings, capital costs, cold ironing, outsourcing
outsourcing
market orientation, responsiveness, claim handling,
Customer Orientation responsiveness, claims, flexibility
marketing flexibility
high-end technology, industry 4.0, cloud systems,
technology, industry 4.0, cloud,
Communication video conferencing, electronic data interchange, port
video conferencing, EDI, POS
operating system

CODING AND TRAINING

Coding is the main part of content analysis which convert textual into the statistics. However, the coder
should be trained for consistency of the content analysis (Neuendorf, 2002). The first step of the coder’s
training consists of pilot testing the training instructions to provide that they share a general
understanding of the codebook variables. Upon completion of the training, the coders should
independently start coding the documents. When the finalization of exercise, inter-rater realibility should
be evaluated. The reliability scores should meet satisfactory level between coders.

There are many computer-assisted qualitative data analysis softwares (CAQDAS) to be used in content
analysis. Thanks to MAXQDA is the most popular software for qualitative research aims, it is possible
to detect user-specified keywords in the datasources. The associated frequency counts and the location
of the keywords were exported to the Excel data for the tabulation purposes. In the Excel, the Visual
Basic scripting was preferred to measure the constructs related to the resilience of competitive
advantage. Spot checks were randomly conducted to validate the accuracy of automated coding exercise
and results found logical with human coding.

RELIABILITY CHECKS
Final step is the measurement of inter-rater reliability statistics. A very common metric is the percentage
agreement. Neuendorf (2002) argues that scores of 0,90 percent and above are acceptable. 0,80 or greater
is acceptable in most situations, and 0,70 may be appropriate in some exploratory studies for some
indices. More metrics, including Scott’s 𝜋 Cohen’s Kappa, and Krippendorff’s 𝑎 offer a chance
corrected assesments. For these metrics, a 0,70 is acceptable agreement in the reliability checks.

The making use of CAQDAS software diminishes human bias and therefore, does not need to check
reliability checks. But, the sensivity in coding should be checked regarding coding rules, for example,
the use of comprehensive and restrictive dictionaries. The degree of consistency test conducted via

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

74
ReCal2 document and highest consistency scores (Scott’s 𝜋 Cohen’s Kappa, and Krippendorff’s
𝑎=0,91) were reported.

FINDINGS
The findings of this study illustrates research themes, profiles of the respondents, tools and techniques,
and finally, content analysis about the constracts of resilience of competitive advantage during Covid-
19 in maritime industry.

RESEARCH THEMES
This section offers a summary in Figure 5, explains that content analysis has been conducted widely in
the “resilience of competitive advantage during Covid-19 in maritime shipping” area. Since content
analysis has been specifically well-known in this field, this study reveal important highlights that
published in this area.

The most important topic of all respondents was determined as process innovation for capabilities, while
the buildings and other pyhsical structures is the most common topic for company resources. Disruption
conditions need process innovation that firms prefer to survive (Siagian et al., 2021). The firms should
establish boundries between employees to keep adequate distance during Covid-19, the use of
technology for integrating offices, and supply chain partners. In process innovation, the companies
adopts innovative technology such as, the using video conference softwares for remote access.
Similarly, Liu et al. (2020) and Tarigazan et al. (2019) found that process innovation was positively
supported by information system integration. Successful information system integration generates a
good supply chain integration and finally results to the a good coordination concerning on-time delivery,
efficient tracking shipments, and of course, efficient planning in the supply chain management.

On the other hand, organizational culture and financial strength were the less important topics that all
respondents were agreed. The characteristics of organizational culture directly affect the productivity of
the organization and especially the morale of the employess in the organization. Just as the pandemic
has changed the organization's supply chain processes, the resulting uncertainty has greatly changed the
way of the business in the maritime industry. However, respondents argued that they developed limited
organizational precautions including verbal instructions, training and warning signs, and near miss
procedures against infection of Covid-19 to develop safe behaviors in the workplace. They have mostly
encouraged employees work from their homes.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

75
Figure 5. Research Themes (Resources and Capabilities) in Content Analysis

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

76
 Table 3. Profiles of the Respondents
Shipping Line Operators Port Operators NVOCCs
Sector Company Position Sector Company Position Sector Company Position
Resp. Resp. Resp.
Position Experience Experience Experience Position Experience Experience Experience Position Experience Experience Experience
No No No
(Year) (Year) (Year) (Year) (Year) (Year) (Year) (Year) (Year)
Country
1 20 11 5 1 Trade Manager 16 2,5 2,5 1 Director 19 15 5
Manager
Procurement
2 10 6 1,5 2 General Manager 21 6 6 2 General Manager 21 4 4
Manager
Deputy
Seaway product
3 Managing 19 19 5 3 Sales Manager 10 4 1 3 13 4 4
manager
Director
Sales and
Head of Value Added
4 Marketing 10 4 4 4 CEO 8 5 5 4 20 16 7
Services
Manager
Sales and
Operation
5 Marketing 24 17 17 5 22 14 5 5 Regional Director 16 9 9
Director
Specialist
Regional
Deputy General Transatlantic seatrade
6 Sales 16 2 2 6 14 4 4 6 8 6 1
Manager manager
Manager
Outside
Supply Chain
7 Sales 18 6 6 7 General Manager 17 17 2 7 12 1,5 1,5
Manager
Officer
Operation Procurement and
8 CEO 20 20 7 8 20 11 7 8 15 3 1
Manager supply manager
Marketing
9 16 16 5 9 General Manager 17 7 7
Manager
Short Sea
Line Sales
10 24 14 14 10 Sales Director 19 12 3
Country
Manager
Ocean Export
Average > 17 > 11 >6 Average > 15 >7 >4 11 18 2,5 2,5
Specialist
12 Operation Manager 7 5,5 1
Sales and Marketing
13 8 8 5
Manager
Average > 15 >8 >5

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

77
 TOOLS AND TECHNIQUES
CAGDAS softwares offer an easy analysis of the great volumes of text data (Lissack, 1984; Morris,
1994). Whereas the technological improvements has raised the potential and viability of content
analysis, only 14 papers (28%) of the maritime supply chain articles preferred no more than one function
in CAGDAS software.

RESULTS: CONTENT ANALYSIS OF THE STUDY

With regard to the methodology explained above and subsequent coding results, the study had created
the coding schema. In figure 6, the responses of respondents are displayed. The study notes that most
commonly words are simplistic and not consist of such words (a, can, could, did, do, have, had, made,
no, not, and some etc.). On the other hand, the study reveals what the respondents frequently discuss
about the resilience against competitive advantage from the perspective of resource-based theory.
Measures, port and problems (challenges) have been emphasized more than all the other words.

Figure 6. Word Cloud for Content Analysis

To analyse the logic of their research, looking for connections and interrelations between the analytical
categories and also sub-criteria is one of the most popular methods of any research project. MAXQDA
qualitative content analysis software enables the researchers to examine the relationships between pairs
of codes, beyond searching for intersections and the proximity of codes (Kuckartz and Rädiker, 2019,
p.160-162). Figure 7 demonstrates the most significant and meaningful relationships among the
pandemic related challenges. The code relations matrix shows a meaningful relationship among “Blank
sailing”, “Shortage in equipment”, and “Decrease in cargo volume”. However, it is very simple
explanation and require deeper insight for further analysis. Regarding Kuckartz and Rädiker (2019), The
more two codes overlap, i.e. the more similar they are in terms of their use in the data, the closer they

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

78
are marked together on the map, that is, this implies that the more intersections the two codes have, they
will appear closer on the map. Both, the nearness of and the line width between two codes build on the
same piece information: the similarity matrix that contains the amount of co-occurences between the
two codes. The nearness is calculated by the method multidimensional scaling. The line width reflect
the similarity directly. For example, the code “Lockdown” is placed near “Blank sailing”, because it has
co-occurences with “Shortage in equipment”, which also have co-occurences with “Decrease in cargo
volume”. But the lines tells other truth, that there is no direct connection between “blank sailing” and
“Communication disruption” in pandemic related challenges. On the other hand, the code “Buildings
and other physical structures” is mostly connected node with other nodes and placed near the codes
“Cost efficiency”, “Personel management”, “Employee recruitment, compensation, reward, and
training policy”, “Organizational flexibility” and “Vehicles / equipment”. Meanwhile, every node is
connected with each other in the graph of the resilience measures. The findings indicate that the
measures taken by companies are interconnected, although all challenges are not related to each other.

Pandemic related Challenges Resources and Capabilities for Resilience

Measures Resilience Measures

Figure 7. Codes Relationships

In this study, the visualization makes it easier for scholars to determine patterns and uncommon or
different cases. In this study, the visual representation shows that the respondents have easily managed
the challenges with numerous resilience measures. Thanks to these measures, it is seen that they
maintain their competitive advantage and suppress the negative effects of Covid-19. Figure 7 examines
which codes have been appointed to which documents and how often. In the MAXQDA, the
visualization tool helps researchers to illustrate the discovery of codes interrelations. This is called
“Code Matrix Browser” which can be used not only to produce “documents x codes” matrices for
specific documents, but it can also match group of documents.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

79
Resources
Capabilities

Figure 8. Codes Matrixes for Pandemic related Challenges, Resources and Capabilities

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

80
In the code matrix browser, the 31 documents are organized in the columns of the matrix and the codes
“pandemic related challenges”, resources and capabilities for resilience measures coded as “resilience
measures” are arranged in the rows. The documents covers all interviews of the respondents from
shipping line operators (coded as Line), non vessel operating common carriers (coded as NVC), and sea
ports (coded as Port). The individual nodes show how many times the relevant code has been assigned
to the respective document (Kuckartz and Rädiker, 2019). Regarding Kuckartz and Rädiker (2019), the
larger node represents how larger code has been assigned. For example, the document (Line 8) only
mentions the code “Buildings and other physical structures”. The bottom row is a summary row; it
shows how many of the displayed codes are present in each document. This total is only 2 for this
document. In opposite of this document (Line 8), the document (Line 9) mentions lots of problems,
following Line 3, Line 4 and Line 7. Therefore, the analyse indicates that the shipping line operators
have faced lots of challenges and taken resilience measures more than other groups (NVOCC and Ports).
On the other hand, the code “Buildings and other physical structures” is the most mentioned code in the
code system. This code covers not only physical assets (i.e. piers, and their office buildings etc.) of the
respondents’ companies but it also covers their information technologies (IT) infrastructures. For that
reason, it extremely differs from the code “Vehicles / equipment”. This code is related to the operator’s
vessels, port handling equipments or containers. Following this code at the section of resilience
measures, the codes (“Cost efficiency”, “Communication”, “Employee recruitment, compensation,
reward, and training policy”, “Personal management”, “Process innovation” and “Vehicles /
equipment”) are arranged respectively. Additionally, the most commonly mentioned challenges are
respectively “Shortage in equipments”, “Blank sailings” and “Decrease in cargo volumes”. These
challenges appeared to be the same for almost every respondent. Regarding Kuckartz and Rädiker
(2019), this illustration of the codes assigned per document in the Code Matrix Browser should show
that visualizations are far more than just eye-catching gimmicks.

DISCUSSION
The Covid-19 outbreak declared as a pandemic by the WTO in March 2020, will most likely be referred
to by the economic chaos as well as the deaths it caused by future generations. The economic
consequences of the epidemic were felt especially in the 2nd half of 2020. (UNCTAD, 2020). For this
reason, the interviews which constitute the data of this study were conducted between in the middle of
April, 2020 and in the middle of June, 2020.

This paper discusses what challenges supply chain companies have been faced and how supply chain
companies (ocean carriers, port terminals and other supply chain actors etc.) have sustained the
resilience of competitive advantage during Covid-19 in maritime supply chains by exploring required
resources and capabilities of firms.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

81
The respondents expressed various impacts of Covid-19 that are multidimensional including IT
solutions, remote office works, blank sailings, excessive container waiting times at port, shortage of
truck drivers, decreasing and changing demand, communication, personnel management and others.

All of the interviewees emphasized that they started working-from-home. Working-from-home

challenged companies in providing necessary IT structure and equipment, managing people, i.e. their
motivation, who work from home, and adapting the organization of daily operations. However, port
workers had to keep working at the ports by taking serious measures at workplace similiarly most
seafarers were stranded on board of vessels all around the world for many months. The interviewees
pointed out various impacts that Covid-19 disrupted the physical movements of containers, which
caused extreme delays and longer lead times, even cancellations. One of the most frequently mentioned
impact in the interviews was that loaded containers were piled up at ports for extremely longer times.
The stranded containers at ports complicated operations of ports in our sample, especially the ones with
limited container storage area. Almost all interviews reported that cargo owners had to pay high amount
of demurrage and storage fees when Covid-19 became a pandemic. The excessive amount of waiting
time occurred either because consignees cancelled / delayed their orders or they did not have sufficient
space in their warehouse because the demand on their products plummeted or stopped. Another reason
was that inland transportation could not be organized due to driver shortages. The inability to organize
inland transportation also contributed to another major disruption: Empty-container shortage. Ocean
carriers and NVOCCs particularly emphasized the problem of empty-container shortage as many
shippers had to delay or cancel their shipments. This problem was pointed out by interviewees from
different regions. Empty container re-positioning emerged as a major problem for container ship
operators. Another reason of empty-container shortage was the blank sailings. The shipping lines
cancelled many of their scheduled voyages to some specific ports either due to quarantine restrictions
or insufficient demand in those regions. Some lines in interviews reported they employed smaller feeder
vessels to serve some of those ports, which resulted in lengthened lead times. In fact, blank-sailings
were considered as a solution by lines to reduce their operating costs, to increase capacity utilization
and to prevent decrease of freight rates. Nonetheless, blank-sailings caused challenges in empty
container repositioning, delays for shippers and anomalies in the operation of ports.

The interviewees stated that the demand to shipping services fluctuated heterogeneously among
industries. The demand in some industries such as food and medical boomed while it stopped for others
such as the automotive industry. Anomalies in demand resulted in fiercer competition in the market.
Majority of the interviewees, especially NVOCCs and ocean carriers pointed out that they pursued a
more aggressive marketing for industries in which the demand increased.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

82
The companies should establish resource-based strategies by looking the determinants of sustainable
competitive advantage on the basis of durability, transparency, transferability and replicability. Besides,
the readers should consider three types of groups to make a clear understanding of the research insight:
shipping line operators, NVOCC and terminal operators. Shipping line operators are the naturally rival
of the NVOCC companies in a some way but port terminals are placed as their service suppliers for both
sides. Therefore, it is necessary to consider that there may be rivalries between groups as well as intra-
group rivalries.

Durability: Firm capabilities have the potential to be more durable than the resources upon which they
are based because of the company’s ability to maintain capabilities through replacing individual
resources (including people) as they wear out or move on (Grant, 1993). For instance, shipping lines,
NVOCCs and ports’ capabilities in the tranportation industry have been maintained by the protective
measures (increased mobilization, remote office works and digitalization) taken by companies against
Covid-19. Some of the companies (Line 3 and Port 2) have especially ensured the resilience with the
additional protective measures and have achieved sustainable competitive advantage compared to their
competitors in their groups. Another example of the “durability” is that the shipping line (Line 9) made
negotiations to get the tender works that they are in contact with, trying to carry the loads they did not
carry before. They tried to take the tender loads by trying to offer attractive offers at local costs and
freights. For example, the shipping line company offered £ 350 / 18 days transit time versus £ 500 / 9
days transit time for the shipment from X-line Izmir, Turkey to UK. The shipping line would not
normally accept the price at this discount rate, but the coronovirus pandemic has also forced shipping
lines to do so in response to falling freight rates.

Transparency: Transparency clearly needs to overcome both problems stated in the literature and the
information problem is a result of imperfect information between competitors. If a company would like
to imitiate the strategy of competitors, it should firstly build the capabilities which underline the
competitor’s competitive advantage and should be determined what resources are required to replicate
these capabilities. During the pandemic, there is a need to increase the internal and external
communications of the companies as their capabilities. Regarding findings, major companies (especially
Line 3, Line 9, Line 10 and Port 3) has to decide the investment on resources (such as IT systems) to
sustain their business against Covid-19. Others have only have updated IT systems but their changes
were not so deep and occasionally required to be conducted the business manually.

Grant (1993) defines this as the “transparency” of competitive advantage. They have scaled the business
in their services, self-sufficiency through backward integration, and superior customer service through
excellence in reservation, cargo handling, loading and discharging of the cargoes, transportation, and

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

83
other technical services. Regarding the other problem of “transparency”, the capability which needs a
complex arrangement of configuration between huge numbers of various resources is more difficult to
understand than a capability which rests upon the exploitation of a single dominant resource. For
example, container shipping lines or NVOCCs had to organize the empty containers for the next
shipments and needs a close cooperation between various stakeholders, especially ports. Thanks to the
Covid-19, the empty containers have been remained at the ports and caused the demurrage as long as
detention expenses. Therefore, shipping lines or NVOCC have turned to respond to the needs of their
customers with various resources since the capability of the companies to access the equipment has
decreased. For example, Line 3 operator has used its own resources (“cost efficiency” or “buildings and
other physical structures”) by providing freight discounts to their customers to cope with the huge
demurrage costs occured by Covid-19. Meanwhile, shipping line has also offered import storage area to
avoid the demurrage in case of any failure to pick up the empty container at ports.

Transferability: As stated in the literature, the imperfection transferability arise from different sources:
(1) geographical immobility, (2) imperfect information, (3) firm-specific resources and (4) the
immobility of capabilities. With regard the geographical immobility, the people can not move from one
district to another district due to the lockdowns occured by Covid-19. Therefore, the demand of
consumer buying substantially decreased and had gradually affected the industrial markets. The cargo
transportation demand was also decreased %25 in especially European region. On the other hand, Covid-
19 has increased imperfect information and caused greater uncertainity. Evaluating the value of a
resource is made difficult by the heterogeneity of resources and of course, imperfect knowledge of
potential productivity of individual resources. During Covid-19, the imperfect information has resulted
in resources being underpriced or overpriced. Even if most of the respondents feel that the bunker prices
in the pandemic may be the main reason of increasing freight rates, the actual reason is the imperfect
information as well as uncertainity occured by the pandemic. Apart from immobility and imperfect
information, the value of resource may be damaged on transfer due to the decline in productivity (Grant,
1993). When the pandemic started to hit global trade in March, the shipping line alliances were able to
move quickly and decisively to reduce overcapacity and prevent a sharp drop in freight rates. Shipping
line alliances helped container industry boost freight rates in coronavirus pandemic. What happened in
the pandemic did not lead to the results mentioned in the theory of “uncertain imitability”. Finally, the
immobility of capabilities needs the interactive teams of resources, are quite more immobile than
individual resources. This is because they need the transfer of the whole team. While such situations are
frequently seen in wide corporate mergers, they can also occur in some special conditions such as
pandemics.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

84
Replicability: Some resources and services may be easily imitiated through replication. In transportation
services, especially new market innovations (i.e. mega containerships, greener shipping, ant
digitalization etc.) can be copied fairly by competitors. Some of the shipping companies have utilized
their capabilities to motivate the people, but some of them have only focused on developing the
availability of their equipment. In complex organization routines, some capabilities seem to be a simple
but prove exceptionally difficult to replicate. For example, Line 9 changed the method of cargo
transportation from road to rail during the lockdown. They also offered to arrange extra capacity to carry
the cargo and ramped up the intermodal supply to the customer’s plant from 1 train rake per week to 5
trains per day. They supplied over 200 rakes of rail cargo transportation capacity to keep their supply
chain on track.

CONCLUSION, LIMITATION AND FURTHER RESEARCH

This study show that the capabilities given by different players has a different motives but mostly
focused on the sustainability of their competitive advantage during coronovirus pandemic. Capabilities
requires the design of company’s resources regarding situations such as, technology, global economic
climate and unusual cases like a pandemic.

With outbreak of coronovirus pandemic, shipping industry has been essential in terms of guaranteeing
supply chain lines throughout the world, and moving vital stocks of goods, fuel, and medical supplies
from one destionation to other. Just as, the number of port of calls are substantially decreased on
maritime supply chains. Moreover, the shipping routes from Europe to China and from Europe to the
US have been affected. Shipper complaints of excessive and unwarranted charges by providing that
charges may well be found unreasonable unless they further the prompt movement of cargo (“freight
fluidity”), thus shifting the burden of delays that are beyond the control either party towards the carrier
and away from the shipper. In Turkey, ports, NVOCCs and shipping lines have adopted their capabilities
for preventing this catastrophy of their own shippers. Therefore, they have strengthened their resources
(financial power, more poweful IT structure, and cost effectiveness) to add more capabilities for
overcoming the coronovirus pandemic. While the longer-term impact of the Covid-19 outbreak is yet to
be fully understood, all indicators are pointing to significant immediate challenges for the sector. These
differ depending on the maritime transport segment (e.g. container, bulk, reefer, tanker) and whether the
transport operation is domestic or international. They also vary by region, level of development, and the
state of prior preparedness to shocks and disruptions.

This study has several limitations. First, the critical comments on only choosing the respondents are
from the Turkish representatives of line operators, NVOCCs and terminal operators, due to the
difficulties in data collection and the lack of fund and time availability. Secondly, the datas of this study

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

85
were obtained between April – June, 2020 just after Covid-19 became a pandemic. For further study
suggestion, the studies, about first impact of the Covid-19 in maritime supply chain, can be analyzed to
compare our study. Or this study can be renewed to see the changes in this period when the second year
of the Covid-19 pandemic.

REFERENCES
Abend, Gabriel. "The Meaning of Theory." Sociological Theory 26 (June 2008): 173–199; Swanson,
Richard A. Theory Building in Applied Disciplines. San Francisco, CA: Berrett-Koehler
Publishers 2013.
Amit, R. and Schoemaker, P. J. (1993), “Strategic assets and Organizational rent”, Strategic
Management Journal, 13, pp. 33-46.
Barney, J. B. (1986), “Strategic factor markets: Expectations, luck, and business strategy”, Management
Science, 32(10), pp. 1231-1242.
Brumagim, A. L. (1994), “A hierarchy of corporate resources”, Advances in Strategic Management,
10A, pp. 81-112.
Bueno, E., Aragón, J. A., Paz Salmador, M., & García, V. J. (2010). Tangible slack versus intangible
resources: the influence of technology slack and tacit knowledge on the capability of
organisational learning to generate innovation and performance. International Journal of
Technology Management, 49(4), 314-337.
Čater, T., & Čater, B. (2009). (In) tangible resources as antecedents of a company's competitive
advantage and performance. Journal for East European Management Studies, 186-209.
Carraresi, L., Mamaqi, X., Albisu Aguado, L. M., & Banterle, A. (2012). Strategic capabilities and
performance: an application of resource-based view in Italian food SMEs. Universitat Bonn-
ILB Press.
Conner, K. (1991), “Historical comparison of resource-based theory and five schools of thought within
industrial organization economics: Do we have a new theory of the firm?” Journal of
Management, 17(1), pp.121-154.
Daft, R. (1983), Organizational Theory and Design. New York: West.
Drnevich, P. L., & Kriauciunas, A. P. (2011). Clarifying the conditions and limits of the contributions
of ordinary and dynamic capabilities to relative firm performance. Strategic management
journal, 32(3), 254-279.
Fredericks, E. (2005). Infusing flexibility into business-to-business firms: A contingency theory and
resource-based view perspective and practical implications. Industrial Marketing
Management, 34(6), 555-565.
Grant, R. M. (1991). The resource-based theory of competitive advantage: implications for strategy
formulation. California management review, 33(3), 114-135.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

86
Hall, R. (1992). “The strategic analysis of intangible resources”. Strategic Management Journal, 13, pp.
135-144.
Hall, R. (1993), “A framework linking intangible resources and capabilities to sustainable competitive
advantage”, Strategic Management Journal, 14, pp. 607-618.
Hamel, G. and Prahalad, C. (1996), Competing for the Future, Harvard Business School Press,
Paperback edition, Boston (Massachusetts).
Itami, H. and Roehl, T. (1987), Mobilizing Invisible Assets. Cambridge, MA: Harvard University Press.
Kamasak, R. (2017). The contribution of tangible and intangible resources, and capabilities to a firm’s
profitability and market performance. European Journal of Management and Business
Economics.
Kamboj, S., Goyal, P., & Rahman, Z. (2015). A resource-based view on marketing capability, operations
capability and financial performance: An empirical examination of mediating role. Procedia-
Social and Behavioral Sciences, 189, 406-415.
Ketchen Jr, D. J., Hult, G. T. M., & Slater, S. F. (2007). Toward greater understanding of market
orientation and the resource‐based view. Strategic management journal, 28(9), 961-964.
Kuckartz, U., & Rädiker, S. (2019). Analyzing qualitative data with MAXQDA. Switzerland: Springer
International Publishing.
Lam, J. S. L. (2015) ‘Designing a sustainable maritime supply chain: A hybrid QFD-ANP approach’,
Transportation Research Part E: Logistics and Transportation Review. doi:
10.1016/j.tre.2014.10.003.
Lawson, B., & Samson, D. (2001). Developing innovation capability in organisations: a dynamic
capabilities approach. International journal of innovation management, 5(03), 377-400.
Levitt, T. (1993). The globalization of markets. Readings in international business: a decision approach,
249, 249-252.
Lippman, S. A., & Rumelt, R. P. (1982). Uncertain imitability: An analysis of interfirm differences in
efficiency under competition. The bell journal of Economics, 418-438.
Liu, X., Lin, K., Wang, L., & Ding, L. (2020). Pricing decisions for a sustainable supply chain in the
presence of potential strategic customers. Sustainability, 12(4), 1655.
Mata, F. J., Fuerst, W. L. and Barney, J. B. (1995), “Information technology and sustained competitive
advantage: A resource-based analysis”, MIS Quarterly, December, pp. 487-505.
Olavarrieta, S. and Ellinger, A. E. (1997), “Resource-based theory and strategic logistics research”,
International Journal of Physical Distribution & Logistics Management, 27(9/10), pp. 559-
587.
Pandza, K., & Thorpe, R. (2009). Creative search and strategic sense‐making: missing dimensions in
the concept of dynamic capabilities. British Journal of Management, 20, S118-S131.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

87
Rindova, V. P., Williamson, I. O., & Petkova, A. P. (2010). Reputation as an intangible asset:
Reflections on theory and methods in two empirical studies of business school reputations.
Journal of Management, 36(3), 610-619.
Siagian, H., Tarigan, Z. J. H., & Jie, F. (2021). Supply chain integration enables resilience, flexibility,
and innovation to improve business performance in COVID-19 era. Sustainability, 13(9),
4669.
Spanos, Y. E. and Lioukas, S. (2001), “An examination into the causal logic of rent generation:
Contrasting Porter’s competitive strategy framework and the resource based perspective”,
Strategic Management Journal, 22, pp. 907-934.
Tarigan, Z. J. H., Siagian, H., & Bua, R. R. (2019, February). The impact of information system
implementation to the integrated system for increasing the supply chain performance of
manufacturing companies. In IOP Conference Series: Materials Science and Engineering (Vol.
473, No. 1, p. 012050). IOP Publishing.
UNCTAD (United Nations Conference on Trade and Development). (2020). Covid-19 and maritime
transport: Impact and responses. pp. 12 – 15. Geneva.
Wade, M., & Hulland, J. (2004). The resource-based view and information systems research: Review,
extension, and suggestions for future research. MIS quarterly, 107-142.
Wernerfelt, B. (1984), “A resource-based view of the firm”, Strategic Management Journal, 5, pp. 171-
180.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

88
 IMPACT OF COVID-19 PANDEMIC ON AIR TRANSPORTATION
INDUSTRY

Ersin Aktas1, Sercan Demir2, Turan Paksoy3

Abstract ⎯ COVID-19 is the most serious event that has occurred recently, and it continues to
pose a risk to our lives. Since Covid-19 is effective worldwide, implementing stringent measures
on a global scale becomes imperative to combat it. In response to this requirement, governments
all around the world have closed borders imposed travel restrictions and lockdown rules.
Besides, wearing a mask, social distancing, and hygiene became the daily routine we feel obliged
to comply with. This change brought by COVID-19 has caused serious economic and financial
difficulties in many business areas (or industries). Especially in industries that are vulnerable to
crises or external shocks, the effect of these challenges has become more evident. The air
transportation industry is among the industries in which such difficulties are experienced
seriously. Following Covid-19 measures, air travel demand declined, flights were canceled, and a
serious survival problem emerged in the aviation industry. Although many governments
announced bailout packages to support the industry, the negative effects of COVID-19 on
aviation are lasting. The purpose of this study is to investigate the effect of COVID-19 on the
aviation industry. The decline in air traffic, flight, and airport and airline income was examined
and presented in line with this purpose. Also, changing trends in air cargo and new income
sources are discussed.

Keywords ⎯ Aviation, Air Transportation, Airline, Airport, Covid-19

1
Ersin Aktas, Necmettin Erbakan University, Faculty of Aviation and Space Science, Department of Aviation Management, Konya,
Turkey, [email protected]
2
Sercan Demir, Harran University, Faculty of Engineering, Department of Industrial Engineering, Sanliurfa, Turkey,
[email protected]
3
Turan Paksoy, Necmettin Erbakan University, Faculty of Aviation and Space Science, Department of Aviation Management, Konya,
Turkey, [email protected]

19th International Logistics and Supply Chain Congress October 21-22, 2021, 1
Gaziantep, Turkey

89
Introduction
Aviation is highly vulnerable to epidemics and crises that pose a threat to public health
(Tay et al. 2021). Severe Acute Respiratory Syndrome (SARS) in 2003, Avian influenza
H5N1 in 2006 and Swine flu H1N1 pandemic in 2009 were among the crises that
affected the aviation industry in the recent past (Chung, 2015, p. 1). Since mid-2020, the
industry has been struggling to cope with the COVID-19 pandemic, which was first
occurred in December 2019 in the Wuhan province of China and has affected the all
world (WHO, 2020).
Isolation and lockdown have been accepted an effective method in the prevention of such
infectious diseases (Chung, 2015). This method was also applied during the COVID-19
pandemic. World Health Organization (WHO) took some stringent measures to interrupt
transmission of the virus. The WHO developed guidelines on COVID-19. In cooperation
with the International Air Transport Association (IATA), the organization has prepared
guide documents to protect cabin crew and airport workers from the transmission of the
virus (WHO, 2020). Following the announcement of COVID-19 as a pandemic by the
WHO, many countries imposed lockdown rules and travel restrictions. These decisions
had caused many serious consequences for the aviation industry. In this study, we aim to
explain effect COVID-19 pandemic on aviation industry.

Aviation Value System

Transportation is a set of activities that enable people and goods to be moved from one
place to another. Road, rail, air, and water are the main modes of transportation. In
addition, pipelines, cables, and space transportation are considered as a mode of
transportation for a certain purpose (Schmitt and Gollnick, 2016). Air transportation
refers to transportation activities performed by airplanes and aircraft (Wensveen, 2011,
p.35). It can be conceptualized as a value chain. Airline companies are at the center of
this value chain. The upstream part of the aviation value chain includes aircraft and
component manufacturers, infrastructure (airport, air navigation service provider), and
other service providers (fuel suppliers, insurance providers and ground services
providers). The downstream sector of the aviation value chain includes cargo and
passenger distribution (Tretheway and Markhvida, 2014).

19th International Logistics and Supply Chain Congress October 21-22, 2021, 2
Gaziantep, Turkey

90
Impact of Covid-19 on Aviation
COVID-19 continues to affect the global economy. Although almost every sector has felt
its impact, especially the airline sector has felt it seriously. Airline demand declined more
sharply than the combination of the September 11, 2001 event and the 2008 financial
crisis. Although airlines had stronger balance sheets prior to COVID-19 compared to
other crises, COVID-19 has hardly affected even key players in the industry. Some
airlines went to bankruptcy. States around the world announced bailout packages to
support airlines (Curley et al., 2020). More than 16,000 aircraft making up the global
commercial airline fleet grounded through. According to CIRIUM estimates 6000
airplanes will be in storage by the end of 2021 (CIRIUM, 2021). Around 30 airlines went
bankrupt or collapsed bankrupt in 2020 (CAPA, 2021).
In 2020, developments related to COVID-19 such as the number of people affected by
the virus, testing regime, restriction levels, openness of borders have been the main
determinants of airline passenger demand. From end of 2020, the pace of vaccination
significantly affected passenger demand for the airline (IATA WATS, 2021). In the
second half of 2020, less stringent travel restrictions were imposed in Africa and the
Middle East, as well as in North and Latin America than elsewhere. This had a relatively
positive effect on the air traffic in the regions. On the other hand, the strict
implementation of travel restrictions throughout the year in the Asia Pacific region has
seriously affected air traffic (IATA WATS, 2021).
According to the IATA statistics, in 2020, revenue passenger-kilometers (RPKs)
decreased by 65.9%. During the year, around 1.8 billion passengers traveled by air.
Industry-wide available seat-kilometers (ASK) fell 56.7% for the same year. Overall
available cargo ton-kilometers (ACTKs) dropped 21.2% in 2020 (IATA WATS, 2021a).
Globally, airlines closed 2020 with a net profit loss of $126.4bn. Airlines in North
America were most suffered from the net profit losses. These were followed by airlines in
the Asia Pacific and Europe with $35bn and $34bn revenue losses, respectively. African,
Middle Eastern, and Latin American airlines suffered net revenue losses of $2, $7.9 and
$11bn in 2020, respectively (IATA WATS, 2021b).

19th International Logistics and Supply Chain Congress October 21-22, 2021, 3
Gaziantep, Turkey

91
Covid-19 had a serious negative impact on air cargo between the time period covering
March and May 2020, due to the border closures and the suspension of flights. The
easing of restrictions after this period resulted in a rapid increase in air cargo demand.
This allowed the industry to recover close to the pre-crisis period. Still, cargo ton
kilometers fell 9.1% compared to 2019. This decline was the largest on an annual basis
after the 2009 financial crisis (IATA WATS, 2021a)
Airline operators are not expected to achieve positive financial results until 2022 and to
reach pre-crisis traffic levels until 2024. As a response to such situation governments
around world provide financial aid to airlines (IATA, 2020c). State aids to the aviation
industry to prevent bankruptcies arise from two main reasons. First, passenger and freight
transport is a basic public service, and disruptions in this area due to the bankruptcy of
airlines can have serious economic and social consequences. Second, air transport has a
significant impact on the competitiveness of domestic industries in terms of making
international markets accessible (OECD, 2021). In other words, states attach great
importance to air transport connectivity in order to maintain economic activities and
protect employment. In line with this approach, states generally provided aids to major
players in their aviation industry (Abeta et al., 2020).
According to IATA, as of the end of March 2021, the amount of financial aid given to
airlines reached to $226.9 billion. As shown in figure 1, a large amount of government
financial support to airlines was provided in form of direct aid including loans, cash
injections, and equity financing, followed by wage subsidies, loan guarantees, corporate
taxes, ticket taxes and fuel taxes (IATA, 2021c).
An increasing amount of government support is provided to airlines on a commitment to
fulfill certain conditions (Borko et al., 2020). The French government announced an
approximately $7.6 billion support package to Air France airlines. The state has pledged
to provide this support on the conditions that the airline will be the most environmentally
friendly airline in the world, reduce its carbon emissions by half, and obtain 2% of its fuel
from alternative and sustainable sources (Reuters, 2020). In addition, the French
government has decided to provide financial support to major parts suppliers and
manufacturers in exchange for more investment and research in low-emission or fuel-
efficient aircraft (Alderman, 2020). The Austrian government offered a bailout package

19th International Logistics and Supply Chain Congress October 21-22, 2021, 4
Gaziantep, Turkey

92
to Austrian Airlines on the condition of fulfillment of strict ecological requirements. The
ecological requirements, with the emphasis on sustainability, to be met by the airline
were specified as shifting passenger traffic to rail on short-haul flights, reducing carbon
dioxide emissions by half and increasing jet fuel efficiency by 1.5% percent annually
(Austrianairlines, 2020). State aids offered to airlines on a conditional basis will have the
potential to change the aviation industry in the future, as it encourages research,
development, and innovation (Borko et al., 2020).

Reimbursable/ deferral only

46%
Non-reimbursable / waiver
Direct aid 101.11 54% discount
Wage subsidies 76.5

Loan guarantees 25.5

Corporate taxes 12.0

Ticket taxes
11.0
Fuel taxes
0.8

226.9

Figure 2. Government aid made available to airlines due to COVID-19 (IATA, 2021c)

In addition to airlines, states have provided aid to airports and operators in various forms.
The largest part of the supports consisted of cash supports, deferral of payments, and
taxes and wage support. These were followed by grants, financial transfers, state loans,
and loan guarantees. The least amount of the aid was provided in form of equity injection
(OECD, 2021).
The COVID-19 pandemic seriously affected airports. The COVID-19 pandemic has
severely affected airports. With the suspension of flights, aeronautical revenues that
airports generate from airlines in the form of landing fees, passenger fees and security
fees plummeted. Similarly, non-aeronautical revenues of airports including parking areas,
terminal facilities, duty-free and restaurants fell (Twinn et al., 2020).
Before the COVID-19 pandemic, the airport industry was estimated to generate $180
billion in revenue for 2020. However, the pandemic had an unexpected impact on airport
revenues, resulting in $125 billion losses from projected revenue. Similar to passenger

19th International Logistics and Supply Chain Congress October 21-22, 2021, 5
Gaziantep, Turkey

93
demand, the pandemic has mostly hit the revenues of airports in Europe and the Middle
East. Europe suffered an estimated 44.4 billion revenue losses for 2020. Europe and the
Middle East declined 70.5% in their revenues for 2020 from the projected level (ACI,
2021).
Table 1: Projected and estimated airport revenue for 2020 by regions (Source: Airports Council
International (ACI), 2021)

Region Projected revenue Projected under Estimated Estimated

(Pre COVID-19) COVID-19 revenue loss (%)
Africa 3,950 1,280 -2,670 -67.6%
Asia-Pacific 59,060 22,855 -36,206 -61.3%
Europe 62,953 18,584 -44,368 -70.5%
Latin America-Caribbean 11,596 4,506 -7,090 -61.1%
Middle East 15,260 4,498 -10,762 -70.5%
North America 35,398 11,674 -23,723 -67.0%
World 188,217 63,397 -124,820 -66.3%
Source: (ACI, 2021)

Airports generate some of its revenues from airline cargo operations. However, since a
significant part of the cargo is carried on passenger planes, the cargo capacity offered by
the airlines decreased with the suspension of passenger flights (Twinn et al., 2020).
Industry-wide revenue passenger kilometers (RPC) decreased 94.3% year-on-year in
April 2020, the largest decline in recent times. Industry-wide available seat kilometers
(ASK) fell 87% in April 2020 compared to the same period of 2019 (IATA, 2020a). The
decrease in cargo tonne-kilometers (CTK) reached 27.7% in April 2020 (IATA, 2020b).
The aviation industry contributes to approximately $3.5 trillion of economic activity and
87.7 million employments under normal conditions on a global scale. Approximately
11.3 million people are directly employed by the aviation industry. Aviation-supported
employment is expected to temporarily decrease by 46 million to 41.7 million due to the
impact of the COVID-19 pandemic. Direct employment supported by the industry is
projected to decrease by 43% compared to the pre-COVID-19 period, to 4.3 million. In
addition, it is estimated that the economic value created by the sector, including direct
and indirect value, has the potential to decrease by 52% to $1.7 trillion dollars. Aviation
companies employ a high percentage of licensed professionals. These employee groups
have to receive some training in order to be able to do their jobs again after a certain
break. Rehiring, licensing and training these personnel requires significant resources and
19th International Logistics and Supply Chain Congress October 21-22, 2021, 6
Gaziantep, Turkey

94
time. Aviation companies prefer to retain their employees through wage reduction
methods. However, due to the long duration of the impact of COVID-19, a serious
decrease in the employment of licensed personnel is expected (ATAG, 2020).
The performance of airlines is expected to be worse than forecast due to difficulties in
controlling Covid-19 variants and the slower pace of vaccination in some regions.
Total airline passenger demand is expected to be around 2.4 billion for 2021. Airlines' net
loss after tax is estimated to be around $47.7 bn for 2021. North American airlines are
expected to show an improvement in performance thanks to the recovery in the domestic
market and close 2021 with a loss of $5 billion. European airlines are forecast to close
2021 with a net loss of $22 billion due to the slow pace of vaccination and little
relaxation of travel restrictions. Airlines in other regions will continue to show negative
financial performance due to low demand for international travel (IATA, 2021c).
Air cargo will continue to be a profitable source of revenue for airlines in 2021, as the
recovery in the economy provides an increase in cargo demand and the capacity offered
by airlines to meet this demand is not at the desired pace. Cargo revenues are estimated to
reach $152 billion in 2021. However, the improvement in air cargo will not be enough to
compensate for the loss of revenue due to weak demand in the passenger market. Overall,
operating revenues of airlines are forecasted to increase by 23% for 2021. However, this
increase corresponds to 55% of 2019 (IATA, 2021c).
The grounding of planes due to the pandemic has caused airlines to face some new
challenges. Airlines need more equipment and engine covers to protect grounded
airplanes. In particular, the storage of different types of aircraft on a certain base has
revealed the need for aircraft type-specific technical personnel and ground handling
equipment. In addition, highly skilled personnel of aviation industry experienced a long
period of layoffs or remained redundant. This resulted in the need for refreshing training
to keep their skills current (Adrienne et al., 2020).

Conclusion
Infectious diseases that pose a risk to public health have always been a threat to the
aviation industry. COVID-19 has impacted the aviation industry in unprecedented ways
that threaten its future sustainability (Tay et al. 2020). In this study, we examined the
impact of COVID-19 on air transport, which can be considered as a value chain. Some of
19th International Logistics and Supply Chain Congress October 21-22, 2021, 7
Gaziantep, Turkey

95
the key results have been revealed that needs to be emphasized. First, the air transport
sector suffered serious economic losses mainly due to the suspension of flights of airlines
located at the center of the aviation value chain. Although some of these losses are
compensated by government aid and increased air cargo demand, it can be said that it will
take time for airlines to make a profit due to the continuing impact of Covid-19. Second,
although there is some recovery in airline demand, it will take time to reach pre-
pandemic levels. Therefore, some aircraft are still in ground position. This indicates that
airlines will continue to incur extra maintenance costs to keep planes operational.
Thirdly, state aids provided on a conditional basis can encourage elements of the air
transport value system to prioritize environmentally friendly technologies and operating
methods.

19th International Logistics and Supply Chain Congress October 21-22, 2021, 8
Gaziantep, Turkey

96
 REFERENCES

ACI (2021) The impact of COVID-19 on the airport business and the path to recovery
https://aci.aero/news/2021/03/25/the-impact-of-covid-19-on-the-airport-business-and-the-
path-to-recovery/

Adrienne, N., Budd, L., & Ison, S. (2020). Grounded aircraft: An airfield operations
perspective of the challenges of resuming flights post COVID. Journal of Air Transport
Management, 89, 101921.

Alderman, L. (2020) France’s Aerospace Industry to Get $17 Billion in Government

Support. Retrieved from https://www.nytimes.com/2020/06/09/business/France
aerospace-industry.html

ATAG (2020). Aviation: Benefits Beyond Borders. Air Transport Action Group.
https://aviationbenefits.org/media/167186/abbb2020_full.pdf

Austrain Airlines (2020) Austrian Airlines Coronavirus Update: Austrian Airlines

Receives Financial Aid from the Federal Government and Lufthansa. Retrieved from
https://www.austrianairlines.ag/Press/PressReleases/Press/2020/06/038.aspx?sc_lang=en
&mode=%7B30999B4B-42D0-45A6-B671-FE5E3CB68ED8%7D

Borko, S., Geerts, W., & Wang, H. (2020). The Travel Industry Turned Upside Down:
Insights, Analysis and Actions for Travel Executives. Skift Research

CAPA (2021). Only a few airlines collapsed in 2020. (CAPA Center for aviation).
https://centreforaviation.com/analysis/reports/only-a-few-airlines-collapsed-in-2020
550663

Chung, L. H. (2015). Impact of pandemic control over airport economics: Reconciling

public health with airport business through a streamlined approach in pandemic control.
Journal of Air Transport Management, 44, 42-53

Cirium (2020). Cirium Weekly Covid-19 Aviation Industry Update – April 5,

2021. 12 Months Of In-Storage Passenger Aircraft.
https://www.cirium.com/thoughtcloud/cirium-weekly-covid-19-updates/

Curley, A., Dichter, A., Krishnan, V., Riedel, R., & Saxon, S. (2020). Coronavirus:
Airlines brace for severe turbulence. McKinsey & Company.

IATA (2020a) Air Passenger Market Analysis April 2020 Air passenger demand comes
to a standstill amidst lockdowns. https://www.iata.org/en/iata
repository/publications/economic-reports/air-passenger-monthly-analysis---apr-20202/

IATA (2020b) Air Cargo Market Analysis April 2020 Air cargo volumes fall further but
load factors up sharply. https://www.iata.org/en/iata-repository/publications/economic-
reports/Air-Freight-Monthly-Analysis-Apr-2020/
19th International Logistics and Supply Chain Congress October 21-22, 2021, 9
Gaziantep, Turkey

97
IATA (2021b) Industry Statistics Fact Sheet April 2021. https://www.iata.org/en/iata-
repository/pressroom/fact-sheets/industry-statistics/

IATA (2021c) IATA Outlook for the global airline industry April 2021 update.
https://www.iata.org/en/iata-repository/publications/economic-reports/airline-industry-
economic-performance---april-2021--
report/#:~:text=But%20the%20average%20for%202021,air%20travel%20growth%20ove
r%202020.&text=pent-up%20demand%20and%20the,%25%20of%20pre-
%20crisis%20levels.

OECD (2021) State Support to the Air Transport Sector: Monitoring developments
related to the Covid-19 crisis. https://www.oecd.org/corporate/State-Support-to-the-Air
Transport-Sector-Monitoring-Developments-Related-to-the-COVID-19-Crisis.pdf

Reuters (2020) Air France must cut emissions, domestic flights for aid: minister.
https://www.reuters.com/article/us-health-coronavirus-france-economy/air-france-must-
cut-emissions-domestic-flights-for-aid-minister-idUKKBN22B2EL?edition-redirect=uk

Schmitt, D., & Gollnick, V. (2016). The Air Transport System Springer,Werlag Wien.

Tay, D., Du, K., Ho, J., Liu, F., Chan, C., & Cao, C. (2020). The Aviation Industry:
Tackling the turbulence caused by COVID-19. IETI Trans. Econ. Manag, 1, 44-56.

Tretheway, M. W., & Markhvida, K. (2014). The aviation value chain: Economic returns
and policy issues. Journal of Air Transport Management, 41, 3-16

Twinn, I and Qureshi, N., Rojas, D. S. P and Conde, M. L. (2020) The Impact of
COVID-19 on Airports. International Finance Corporation (IFC)
https://www.ifc.org/wps/wcm/connect/26d83b55-4f7d-47b1-bcf3-01eb996df35a/IFC
Covid19-Airport FINAL_web3.pdf?MOD=AJPERES&CVID=n8lgpkG

Wensveen, J. G. (2011). Air transportation: A management perspective. Seventh edition.

Ashate.

WHO (2020) Coronavirus disease 2019 (COVID-19) Situation Report – 94, retrieved
from https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200423-
sitrep-94-covid-19.pdf

19th International Logistics and Supply Chain Congress October 21-22, 2021, 10
Gaziantep, Turkey

98
 THE IMPACT OF LOGISTICS INNOVATIVENESS ON PERCEIVED
ORGANIZATIONAL PERFORMANCE: APPLICATION IN THE
LOGISTICS INDUSTRY
Karahan KARA1

Abstract - Companies exhibit innovation approaches in many areas such as product, process, technology,
and management. Logistics innovativeness approach is the innovation approach of companies in their
logistics activities. Innovative behaviors in logistics activities provide companies with competitive power.
At the same time, companies can increase their current performance levels with logistics innovativeness.
The aim of this study is to reveal the relationship between the logistics innovativeness levels of logistics
companies operating in the eastern black sea region and their perceived organizational performance
levels. Another aim of the study is to determine whether logistic innovativeness and perceived
organizational performance levels differ according to the demographic status of the participants. Within
the scope of the research, the reliability and validity of logistics innovativeness and perceived
organizational performance scales were used in the literature. Scale expressions were distributed to
logistics companies with the created questionnaire application. The sampling area was determined by
simple random sampling. The data set was created according to the results of a total of 351 surveys from
58 logistics companies in the eastern black sea region, Turkey. Hypotheses have been created within the
scope of the research. SPSS and AMOS package programs were used to test the hypotheses. As a result of
the research, it was concluded that logistics innovativeness has a significant positive effect on perceived
organizational performance. In addition, perceived organizational performance has been showed
differentiation according to gender, age and tenure, but it has been determined that there is no
differentiation in all demographic variables within the scope of logistic innovativeness variable.
Keywords: Logistics Innovativeness, Perceived Organizational Performance, Structural Equation
Modeling

INTRODUCTION
Businesses are structures that can create competitive power and establish an advantage over their
competitors as a result of joint action with other stakeholders of the value chain they are in.
Logistics service providers, which play an active role in the value chain, provide added value to
companies in achieving their goals. The nature of added value is the quality of the service
provided by logistics companies and the distinctive features of the service from other logistics
companies (Kemppainen & Vespalainen, 2007). The innovation mentality created within
logistics companies contributes to the provision of qualified and new logistics services. It also
helps the organization created by the employees to be successful. Innovation-oriented studies in
logistics activities, which are conceptualized as logistics innovativeness in the literature, emerge
as the driving force that enables logistics companies to generate new ideas and transform these
ideas into new logistics services (Raltson et al., 2013).

1
Ass.Prof., Artvin Coruh University, Hopa Vocational School, Logistics Program, [email protected].

99
Various performance measurement methods are applied to determine the customer value and
market value of the efforts made by the companies. These measurement methods are created by
considering the financial, market, organization, operation, etc. reference points. Organizational
performance measurement is a form of evaluation created by taking into account all of the
company's activities (Bekmezci et al., 2020). Perceived organizational performance is an
organizational performance evaluation approach that is not based on the quantitative output of
firms, but on the perceptions of employees (Dess & Robinson, 1984). With this method,
managers obtain the internal performance perspective by reaching the evaluations of employees'
organizational performance. In the light of these data obtained, activities for performance
improvement are revealed by taking managerial decisions.
The purpose of this study is to determine the relationship between logistics service providers'
logistics innovativeness and perceived organizational performance. The fact that innovation
performance is among the basic indicators of organizational performance (Drucker, 2000)
supports our stated goal. Therefore, in the conceptual framework of the study, the concepts of
logistics innovativeness and perceived organizational performance are explained. In the method
section, the research model, sampling, and scales used in the research are explained. In the
findings section of the study, the results of the analysis made using SPSS and AMOS package
programs are included. In the last part of the study, the results obtained based on the findings are
explained and presented with literature support.

CONCEPTUAL FRAMEWORK
Logistics Innovativeness
Competition strategies, which are among the management strategies, explain the necessity of
companies to be superior to their competitors in the current market conditions and bring
companies to the foreground to create competitive power. In the literature, innovation-oriented
competitive strategies are emphasized as one of the main ways of gaining competitive advantage
at both national and global levels (Salunke et al., 2011). Developing innovative designs within
the value chain to increase customer satisfaction and creating innovative behavior patterns form
the basis of innovation-oriented competitive strategies (Weerawardena & Mavondo, 2011).
Organizations develop new management models in management practices in order for employees
to gain innovative behaviors and to adopt organizational goals and to work in line with the
determined goals (Gebauer, 2011). The basic expectation of innovative management models is to
display innovative attitudes and behaviors from employees. At this point, the concept of
innovation emerges. Midgley and Dowling (1978) explain the concept of innovativeness as the
personality trait that all individuals in society have more or less. Subramanian (1996)
reconstructed the conceptual framework of the concept of innovation. In the study, the approach
of handling the concept of innovation in the literature was criticized, and the characteristic
features of the organization are the determining variables of innovativeness, as well as the
necessity of innovativeness among the determining variables of organizational performance.

100
The organizational performance level of the organizations varies depending on the performance
success of the elements within the organization as well as the success of the value chain in which
the organizations are located. Logistics activities, like all activities in the value chain, have a
significant impact on organizational performance. Raltson et al. (2013) explains that the
innovative behaviors performed within the scope of logistics activities are logistics
innovativeness and evaluates the new ideas, methods and techniques developed by personnel
working in logistics departments to improve logistics activities within the scope of logistics
innovation. In addition, the creation of innovations that create value in all logistics activities,
from the production stage of the product to the delivery to the customer and even after-sales
services, is considered at the same level as the innovative value created by the development of a
new product or service (Lee & Song, 2010; Rahim et al., 2020).
The theoretical basis of the concept of logistics innovativeness is the resource-based view and
diffusion theory. The resource-based view suggests that the development of the resources owned
by the firms can provide a competitive advantage (Barney, 1991), while the diffusion theory
suggests that the speed of information dissemination can be increased by increasing the intra-
organizational communication efficiency (Bass, 1969). Considering logistics innovation among
the capabilities of the company also helps to use the innovative talents of the personnel involved
in logistics processes as a resource. Thus, improving the logistics capabilities of the companies
with innovative approaches gives companies a competitive advantage in logistics processes
(Fawcett et al., 1993; Raltson et al., 2013). In order to develop solution ideas for the problems
encountered in logistics activities and for the company to get maximum benefit from these
solution proposals, the innovative idea should reach decision makers through effective
communication channels as soon as possible. For logistics innovation to be created within the
company and employees to feel this approach, it is necessary to create an innovative
management approach that supports employees to develop new ideas and designs.
Perceived Organizational Performance
Performance is about the determination of the degree of achievement of predetermined targets,
taking into account company or situation-specific constraints (Lebas, 1995). Decision makers
who determine the performance measurement criteria set the measurement criteria by acting on a
case-by-case basis. This situation makes performance evaluation approaches subjective.
Managers use perceived organizational performance measurement methods to learn the
perceptions of employees about firm performance and to examine the effects of these perceptions
on employee performance (Zeffane & Melhem, 2017). This approach guides the creation of
environmental conditions necessary for managers to get high performance from their employees
(Berberoglu, 2018). It is seen in the literature that organizational performance levels can be
calculated with different perspectives such as financial, technical, managerial, etc. However,
perceived organizational performance measurement, with a holistic approach, explains the
overall performance of the organization based on the perceptions of all stakeholders of the
organization (Delaney & Huselid, 1996).
Objective and subjective criteria are two different approaches used in organizational
performance measurement. Performance evaluation calculations based on objective criteria

101
provide companies with absolute information. However, there are difficulties in transforming the
obtained absolute information into inferences about the general performance status of
organizations. On the contrary, while subjective criteria are insufficient to provide absolute
information about the performance of organizations, their generalizability levels are high (Allen
et al., 2007). At the same time, significant relationships are observed between subjective
performance measurement results based on perception and objective performance measurement
results based on absolute calculations (Carmeli & Schaubroeck, 2008). Thus, perceived
organizational performance calculations can be expressed as a form of evaluation that reflects the
general performance of companies and points to absolute performance situations.
METHODOLOGY
Research Model
Our research model, which was created to reveal the relationship between the logistics
innovation levels of individuals working in the logistics sector and their perceived organizational
performance levels, is presented in Figure 1. Within the scope of the research, it was also aimed
to conduct difference tests of logistic performance and perceived organizational performance
responses according to demographic variables.

Perceived
Logistics
Organizational
Innovativeness
Performance

Figure 1: Research Model

The aim of the research is to determine the effect of logistics innovation approach in increasing
the organizational performance of logistics companies. The importance of the research is the
acceptance of the concept of logistics innovativeness in innovative approaches and the transfer of
the relationship between logistics innovativeness and organizational performance to the
literature.
Hypotheses
Considering the impact of logistics activities within the scope of supply chain applications on
firm performance, it is expected that there will be a meaningful relationship between logistics
activities and performance. It is thought that logistics innovativeness, which increases the quality
of logistics activities and creates a competitive advantage, can have a significant impact on firm
performance in this direction. In the literature, the concept of innovation is mostly
conceptualized as product innovation, process innovation, managerial innovation, and market
innovation. Logistics innovation, on the other hand, ensures that innovation activities are handled
in logistics activities. Therefore, in the literature, it is seen that the studies on the concept of
logistics innovativeness have not been done at a sufficient level.
As a result of reviewing the literature on the relationship between the concept of innovation and
performance, Lee et al. (2011) collected data from 243 hospitals within the scope of their study

102
on the relationship between hospitals' supply chain innovation and organizational performance.
Analyzes based on the data obtained conclude that supply chain innovation has a statistically
significant positive effect on organizational performance. In their study on supply chain and
logistics managers, Kim et al. (2006) have explained that applied technological supply chain
innovation influences channel capabilities and channel capabilities (information exchange and
responsiveness) have a significant effect on market performance. According to the analysis
findings based on the survey data collected from logistics and supply chain managers by Ralston
et al. (2013), who consider logistics innovation among logistics capabilities, it is seen that
logistics innovativeness has a significant positive effect on logistics performance. Panayides
(2006) has found that innovativeness strengthens the relationship between the logistics service
provider company and the customer, increases the quality of logistics service and affects the
enhancement of firm performance. Rahim et al. (2020) have found that logistics innovativeness
has a significant positive effect on logistics performance, because of the regression analysis of
data collected from 150 supply chain and logistics department employees. Ferraresi et al. (2012)
obtained the conclusion that knowledge management and strategic focus affect innovation and
that innovation has a significant effect on firm performance by analyzing the data collected from
241 Brazilian companies. Eris and Özmen (2012) determined by structural equation modeling
that innovation in the logistics sector in Turkey has a significant positive effect on firm
performance. Seo et al. (2014) determined that innovation has a significant effect on supply
chain integration and performance, with the data collected from South Korean manufacturing
companies and the application of structural equation modeling. Günday et al. (2011) examined
the relationship between innovation performance and market performance and product
performance in 184 manufacturing companies operating in Turkey. According to the research
findings, innovation performance significantly affects both market and product innovation
performance. In the same study, it is also seen that organizational and product innovation have a
significant positive effect on innovation performance, but process innovation does not
significantly affect innovation. The relationship between innovativeness and firm performance
was examined in the study Peng (2007) conducted on a sample of 333 entrepreneurial firms.
Within the scope of the study, it was concluded that organizational structures have a regulatory
effect on the relationship between innovativeness and business performance. Anning-Dorson et
al. (2018) found the finding that innovativeness has a significant effect on firm performance in
the light of data tagged from 344 executives working in service companies.
Studies in the literature clearly point out that there is a significant relationship between
innovativeness and performance. Based on this point, our hypothesis, which was created to
reveal the significance level of the relationship between logistic innovativeness and perceived
organizational performance, and our hypotheses created to reveal the differentiation levels of the
answers given according to the demographic variables of the sample area are as follows:
H1: Logistics innovativeness has a significant positive effect on perceived organizational
performance.
H2: There is a significant difference in logistics innovativeness and perceived organizational
performance responses according to demographic variables.

103
H2a: There is a significant difference in logistics innovativeness responses according to gender
demographic variable.
H2b: There is a significant difference in perceived organizational performance responses
according to gender demographic variable.
H2c: There is a significant difference in logistics innovativeness responses according to marital
status demographic variable.
H2d: There is a significant difference in perceived organizational performance responses
according to marital status demographic variable.
H2e: There is a significant difference in logistics innovativeness responses according to age
demographic variable.
H2f: There is a significant difference in perceived organizational performance responses
according to age demographic variable.
H2g: There is a significant difference in logistics innovativeness responses according to the
educational level demographic variable.
H2h: There is a significant difference in perceived organizational performance responses
according to educational status demographic variable.
H2ı: There is a significant difference in logistics innovativeness responses according to tenure
demographic variable.
H2j: There is a significant difference in perceived organizational performance responses
according to tenure demographic variable.
H2k: There is a significant difference in logistics innovativeness responses according to the
logistics sector.
H2l: There is a significant difference in perceived organizational performance responses by
logistics sector.

Research Scales and Sampling

Within the scope of the research, logistics service provider companies operating in the Eastern
Black Sea region of Turkey were selected as the sample area to examine the relationship between
logistics innovativeness and perceived organizational performance. Simple random sampling
method was used to determine the sampling area. The simple random sampling method was
chosen considering the benefits it provides in terms of cost and time (Eygü & Kılınç, 2019). The
questionnaire form for the variables was sent to 58 logistics companies, totally 437. A total of
351 questionnaires were obtained from the scale forms collected in 2021 at a level that can be
used in analysis applications. In structural equation modeling studies, it is considered sufficient
that the sample size is more than 10 times the sum of the expressions of the variables (Kline,
1994). Therefore, our sample size is considered to be capable of representing the population.
Within the scope of the survey application, gender, age, marital status, educational status,
working time and the main logistics field of the company they work for were asked from the
personnel working in logistics companies. In addition, scale expressions of logistics
innovativeness and perceived organizational performance variables were directed to employees.
The scale expressions of the logistics innovativeness variable were selected from the team
climate inventory expressions developed by Anderson and West (1998) and made applicable in

104
the logistics sector by Ralston et al (2013) (α = 0.96). The logistics innovativeness scale is
unidimensional and consists of 5 statements in total. The perceived organizational performance
scale was created by harmonizing the organizational performance scale developed by
Khandwalla (1977) and used by Moideenkutty et al. (2011) by Kitchot et al. (2020) (α = 0.81).
Perceived organizational performance scale is unidimensional and consists of 5 statements in
total. In order for the scale expressions of the variables to be evaluated by the participants, the 5-
Likert ("1" strongly disagree, "2" disagree, "3" undecided, "4" agree, "5" strongly agree) scale
was used.
FINDINGS
Demographic Findings
The findings of the logistics employees participating in the survey application are shown in
Table 1. When Table 1 is examined, approximately 60% of the participants working in the
logistics sector are male employees. Considering the age variable, the density of participants
working in the logistics sector is between the ages of 31-40 (43%). In marital status, it is
observed that approximately 70% of logistics employees are married. When the education levels
of the participants are examined, it is seen that approximately 50% of associate degree graduates
and those who have worked in the first 10 years are more dominant than those working for 10 or
more years. When the basic logistics areas of the companies where the participants work are
examined, it is observed that the companies serving in the field of transportation constitute
almost half of the total sampling area.
Table 1: Demographic Findings
Gender Number % Marital status Number %
Woman 129 36,8 Married 245 69,8
Man 222 63,2 Single 106 30,2
Total 351 100 Total 351 100
Education
Age Number % Number %
Status
High school and
18-30 94 26,8 74 21,1
pre high school
31-40 151 43,0 Associate degree 175 49,9
41-50 87 24,8 Undergraduate 92 26,2
50 + 19 5,4 Postgraduate 10 2,8
Total 351 100 Total 351 100
Tenure Number % Logistics Area Number %
0-5 106 30,2 Transportation 172 49,0
5-10 128 36,5 Warehouse 75 21,4
10-20 89 25,3 Packaging 53 15,1
20 + 28 8,0 Customs 34 9,7
Others 17 4,8
Total 351 100
Total 351 100

105
Findings Regarding the Scales
The results of normality tests of the data obtained from logistics innovativeness and perceived
organizational performance scales are presented in Table 2. Considering the Kolmogorov-
Smirnov normality test results, it was observed that the collected data were not distributed
normally. For this reason, kurtosis and skewness values of the expressions of the scales were
examined. Tabachnick and Fidell et al. (2013) stated that when the kurtosis and skewness values
of the scale expressions are between -1.5 and 1.5, the data show a normal distribution. Since
kurtosis and skewness values of our scale expressions were between -1.5 and 1.5, it was accepted
that the data showed a normal distribution.
Table 2: Kolmogorov-Smirnov Normality Test Results
Scales N Mean SS Test Value P
Logistics Innovativeness Scale 351 3.55 0.76 3.749 0.000
Perceived Organizational
351 3.76 1.02 4.288 0.000
Performance Scale

Since the logistics innovativeness and perceived organizational performance scales were
translated into Turkish and applied, the Exploratory Factor Analyzes (EFA) of the scales were
obtained with the help of the SPSS package program. The factor loads of the expressions
according to the KFA results are shown in Table 3 and Table 4. Factor loadings are expected to
have a load value of at least 0.32 for the scale to be explanatory (Büyüköztürk et al., 2017).
When the factor loads of both scales are examined, it is understood that all factor loads are
greater than 0.32 and they are sufficient to explain the scale.
Table 3: DFA Results of Logistics Innovativeness Scale
Logistics
Items
Innovativeness
LI2- Our logistics employees have enough time to develop new ideas. 0.951
LI5- Our logistics staff provide practical support for ideas and applications. 0.942
LI1- Our logistics employees are open and sensitive to change. 0.909
LI3- Our logistics employees collaborate to help develop and implement new
0.809
ideas.
LI4- Our logistics people find and share resources to help implement new
0.641
ideas.
Eigenvalues 3.684
Percentage of Variance Explained (%) 73.68
Total Variance Percentage (%) 73.68

106
 Table 4: Perceived Organizational Performance Scale DFA Results
Perceived
Items Organizational
Performance
POP2- How would you evaluate your company's performance in terms of
0.938
product or service quality compared to your entire industry?
POP1- How would you rate your firm's performance in terms of sales or
0.931
revenue growth rate compared to your entire industry?
POP3- How would you rate your company's performance in terms of
0.929
employee productivity compared to your entire industry?
POP4- How would you evaluate your company's performance in terms of on-
0.901
time delivery compared to your entire industry?
POP5- How would you rate your company's performance in terms of price or
0.896
cost compared to your entire industry?
Eigenvalues 4.224
Percentage of Variance Explained (%) 84.48
Total Variance Percentage (%) 84.48

Confirmatory factor analyzes (CFA) of the scales were calculated with the help of AMOS
program. DFA results obtained are presented in Table 5 and Table 6. When the fit indices of
CFA analyzes are examined, it is observed that it is within the threshold values presented by
Tabachnick and Fidell et al. (2013), and although the RMSA fit value is below 0.10, it is at a
mediocre acceptable level of fit (MacCallum et al., 1996).
Table 5: CFA Results of Logistics Innovation Scale
Parameter Estimates Unstandardized (β) Standardize (β)
Measuring Model
Logistics Innovativeness > LI1 1.000* 0.870*
Logistics Innovativeness > LI2 1.113* 0.992*
*
Logistics Innovativeness > LI3 0.904 0.690*
*
Logistics Innovativeness > LI4 0.562 0.494*
*
Logistics Innovativeness > LI5 1.000 0.968*
2 ** *** *** ***
X [4, N=351] = 14, CMIN/df (3.493) , CFI (0.994) , RFI (0.976) , IFI (0.980) , TLI (0.985) NFI
(0.992)***, RMSA (0.084)****
*
p<0.01
**
3 < CMIN/df < 5 (Acceptable fit)
***
CFI, NFI, RFI, IFI, TLI > 0.95 (Good fit)
****
RMSA < 0.10 (Mediocre acceptable fit)

107
 Table 6: CFA Results of the Perceived Organizational Performance Scale
Parameter Estimates Unstandardized (β) Standardize (β)
Measuring Model
Perceived Organizational Performance > POP1 1.000* 0.905*
*
Perceived Organizational Performance > POP2 0.993 0.920*
*
Perceived Organizational Performance > POP3 1.023 0.930*
*
Perceived Organizational Performance > POP4 0.957 0.848*
Perceived Organizational Performance > POP5 1.004* 0.843*
2 ** *** *** ***
X [3, N=351] = 2.2, CMIN/df (0.733) , CFI (1.000) , RFI (0.996) , IFI (1.000) , TLI (1.001) NFI
(0.999)***, RMSA (0.000)****
*
p<0.01
**
CMIN/df < 3 (Good fit)
***
CFI, NFI, RFI, IFI, TLI > 0.95 (Good fit)
****
RMSA < 0.05 (Good fit)

Kaiser ‐ Meyer ‐ Olkin (KMO) and Bartlett Tests Results of logistics innovativeness and
perceived organizational performance scales were calculated with the help of SPSS package
program and are presented in Table 7. As explained in Table 7, the KMO sampling adequacy
levels are 0.829 (p <0.01) for the logistics innovativeness scale and 0.892 (p <0.01) for the
perceived organizational performance scale. These values are seen to be sufficient for KFA
analysis (Kalaycı, 2005).
Table 7: Kaiser Meyer Olkin (KMO) and Bartlett Tests Results of the Scales
LI POP
Kaiser Meyer Olkin Sampling Adequacy Test 0.829 0.892
Bartlett Approximate Chi-square 1714.512 1892.778
Sphericity Degrees of Freedom (df) 10 10
Test P 0.000 0.000

Reliability analyzes have been made with the SPSS package program in order to determine the
reliability analysis of the scale and are presented in Table 8. When Table 8 is examined, it is seen
that the reliability levels of both logistics innovativeness (α = 0.906) and perceived
organizational performance (α = 0.954) scales are at a high level. These values indicate a very
high level of reliability (Kalaycı, 2005).
Table 8: Reliability Analysis Results of the Scales
Scales Items Cronbach’s Alfa (α)
Logistics Innovation Scale 5 0.906
Perceived Organizational Performance Scale 5 0.954

Correlation relationships between scale expressions are shown in Table 9 and Table 10. The
highest correlation relationship between the expressions of logistics innovativeness scales is
between LI2 and LI_Ort (0.932, p <0.01), and the highest correlation relationship between

108
perceived organizational performance scale expressions is between POP2 and POP_Mean
(0.934, p <0.01).
Table 9: Correlation Relationship of Logistics Innovation Scale Expressions
Items Mean S.D LI1 LI2 LI3 LI4 LI5 LI_Mean
LI1 3.686 0.896 1
LI2 3.666 0.875 0.864* 1
LI3 3.333 1.022 0.650* 0.680* 1
LI4 3.492 0.887 0.478* 0.488* 0.461* 1
LI5 3.609 0.806 0.834* 0.961* 0.678* 0.477* 1
LI_Mean 3.557 0.762 0.896* 0.932* 0.828* 0.682* 0.921*
*
p < 0.01

Table 10: Correlation of Perceived Organizational Performance Scale Expressions

Items Mean S.D POP1 POP2 POP3 POP4 POP5 POP_Mean
POP1 3.809 1.095 1
POP2 3.831 1.070 0.896* 1
POP3 3.834 1.090 0.839* 0.857* 1
POP4 3.663 1.119 0.777* 0.780* 0.783* 1
POP5 3.683 1.180 0.759* 0.769* 0.789* 0.808* 1
POP_Mean 3.764 1.020 0.928* 0.934* 0.928* 0.904* 0.901* 1
*
p < 0.01

Findings Regarding the Hypotheses

Within the scope of the H1 hypothesis, which was created by determining the relationship
between logistics innovativeness and perceived organizational performance, it was aimed to
measure the effect of logistics innovativeness on perceived performance. In this context,
primarily the mean, standard deviation of the variables and the correlation relationship between
the variables were examined (Table 11). The existence of a statistically significant relationship
between variables is predicted to indicate the presence of significant effects between variables.
When Table 11 is examined, it is seen that the correlation between logistics innovativeness and
perceived organizational performance is very weak (0.20-0.39) (Evans, 1996)
Table 11: Correlation Relations of the Variables
Logistics Perceived Organizational
Variables Mean S.D.
Innovativeness Performance
Logistics Innovativeness 3.557 0.762 1
Perceived Organizational Performance 3.764 1.020 0.261* 1
*
p < 0.01

Within the scope of the H1 hypothesis, the structural equation model was created and the model
analysis was examined with the help of the AMOS package program. The non-standardized

109
model structure of the path analysis results of the structural equation model is shown in Figure 2,
and the standardized model structure is shown in Figure 3.

Figure 2: Path Analysis Results of the Structural Equation Model (Unstandardized)

Figure 3: Path Analysis Results of the Structural Equation Model (Standardized)

In the standardized structural equation model, a correlation relationship between e3 and e4 error
coefficients of the logistic innovativeness scale and between e6 and e7, e9 and e10 in the
perceived organizational performance scale was applied to the model in order to strengthen the
level of fit of the model. The values of fit indices and the effect of logistics innovativeness on
perceived organizational performance are summarized in Table 12.

110
 Table 12: Path Analysis Results of the Structural Equation Model
Parameter Estimates Unstandardized Standardize
(β) (β)
Structural Model
Logistics Innovativeness > Perceived Organizational Performance 0.36* 0.28*
X2 [31, N=351] = 61.5, CMIN/df (1.985)**, CFI (0.992)***, RFI (0.976)***, IFI (0.992)***, TLI (0.988)
NFI (0.983)***, RMSA (0.053)****
*
p<0.01
**
CMIN/df < 3 (Good fit)
***
CFI, NFI, RFI, IFI, TLI > 0.95 (Good fit)
****
RMSA < 0.08 (Acceptable fit)

As presented in Table 12, it is seen that logistics innovativeness has a significant positive effect
on perceived organizational performance (β = 0.28, p <0.01). As a result of this finding, our H1
hypothesis was accepted.
Within the scope of the H2 hypothesis, it was established to determine whether there is a
difference in the responses of demographic variables to logistic innovativeness and perceived
organizational performance. Independent sample T test results for the gender variable are shown
in Table 13. When Table 13 is examined, it is seen that logistic innovativeness data do not differ
according to gender differences (p> 0.05), but perceived organizational performance data differ
(p <0.05). This finding explains that our H2a hypothesis is rejected and our H2b hypothesis is
accepted.
Table 13: Independent Sample T-test Results for Gender Variable
Gender N Mean SS SD t p
Logistics
Man 222 3.579 0.773
Innovativeness 276 0.698 0.669
Woman 129 3.520 1.078
Perceived Gender N Mean SS SD t p
Organizational Man 222 3.727 1.078
304 - 0.924 0.008
Performance Woman 129 3.827 0.913

Independent sample T test results for marital status are shown in Table 14. When Table 14 is
examined, it is seen that both logistic innovativeness and perceived organizational performance
data do not differ according to marital status differences (p> 0.05). This finding explains that our
H2c and H2d hypothesis was rejected.

111
 Table 14: Independent Sample T Test Results for the Marital Status Variable
Marital
N Mean SS SD t p
Logistics status
Innovativeness Married 245 3.549 0.783
217 - 0.327 0.073
Single 106 3.577 0.715
Marital
Perceived N Mean SS SD t p
status
Organizational
Married 245 3.764 1.015
Performance 196 0.006 0.472
Single 106 3.764 1.037

ANOVA test results for the age variable are shown in Table 15. When Table 15 is examined, it is
seen that logistic innovativeness data do not differ (p> 0.05), but perceived organizational
performance data differ (p <0.05) according to the differences in age variables. This finding
explains that our H2e hypothesis is rejected and our H2f hypothesis is accepted. According to
Tukey analysis results, the 50+ age group differs according to both 31-40 and 41-50 age groups.
Table 15: ANOVA Test Results for Age Variable
Age N Mean SS SD t p
18-30 94 3.705 0.633
Logistics 31-40 151 3.539 0.696
Innovativeness 41-50 87 3.591 0.859 3,347 0.391 0.760
50 + 19 3.521 0.794
Total 351 3.557 0.762
Age N Mean SS SD t p
18-30 94 3.918 0.964
Perceived
31-40 151 3.879 1.027
Organizational
41-50 87 3.957 0.838 3,347 7.712 0.000
Performance
50 + 19 3.324 1.086
Total 351 3.764 1.020

ANOVA test results for education level variable are shown in Table 16. When Table 16 is
examined, it is seen that both logistics innovativeness and organizational performance data do
not differ (p> 0.05) according to differences in educational status. This finding explains that our
H2g and H2h hypothesis was rejected.

112
 Table 16: ANOVA Test Results for Education Status Variable
Education
N Mean SS SD t p
Status
High school
and pre high 74 3.356 0.865
Logistics
school
Innovativeness
Associate
175 3.618 0.706 3,347 2.270 0.062
degree
Undergraduate 92 3.578 0.769
Postgraduate 10 3.800 0.659
Total 351 3.557 0.762
Education
N Mean SS SD t p
Status
High school 1.246
and pre high 74 3.581
Perceived
school
Organizational
Associate 0.887
Performance 175 3.874 3,347 2.249 0.082
degree
Undergraduate 92 3.663 1.078
Postgraduate 10 4.140 0.422
Total 351 3.764 1.020

ANOVA test results for the tenure variable are shown in Table 17. When Table 17 is examined,
it is seen that logistic innovativeness data do not differ (p> 0.05) but perceived organizational
performance data differ (p <0.05) according to the differences in working year variable. This
finding explains that our H2i hypothesis is rejected and our H2j hypothesis is accepted.
According to the results of Tukey analysis, it is seen that there is a difference between those who
have a working period of 0-5 years and those who work for 5-10 years, those who have a
working period of 5-10 years and those who have a working period of 10-20 years.
Table 17: ANOVA Test Results for Tenure Variable
Tenure N Mean SS SD t p
0-5 106 3.652 0.646
Logistics 5-10 128 3.590 0.772
Innovativeness 10-20 89 3.438 0.852 3,347 1.634 0.181
20 + 28 3.428 0.793
Total 351 3.557 0.762
Tenure N Mean SS SD t p
0-5 106 3.907 0.940
Perceived
5-10 128 3.515 1.120
Organizational
10-20 89 3.988 0.902 3,347 4.940 0.002
Performance
20 + 28 3.650 0.980
Total 351 3.764 1.020

ANOVA test results for the logistics sector variable are shown in Table 18. When Table 18 is
examined, it is seen that both logistics innovativeness and organizational performance data do

113
not differ (p> 0.05) according to the differences in the logistics sector. This finding explains that
our H2k and H2l hypothesis was rejected.
Table 18: ANOVA Test Results for Logistics Sector Variable
Logistics
N Mean SS SD t p
Sector
Transportation 172 3.541 0.754
Logistics Warehouse 75 3.637 0.694
Innovativeness Packaging 53 3.452 0.837
4,346 0.599 0.663
Customs 34 3.517 0.717
Others 17 3.647 0.859
Total 351 3.557 0.762
Logistics
N Mean SS SD t p
Sector
Transportation 172 3.844 0.938
Perceived
Warehouse 75 3.744 0.865
Organizational
Packaging 53 3.400 1.306
Performance 4,346 2.248 0.064
Customs 34 3.858 1.220
Others 17 3.929 1.058
Total 351 3.764 1.020

The acceptance and rejection status of the hypotheses created within the scope of the research are
shown in Table 19.

114
 Table 19: Acceptance / rejection status of hypotheses
Acceptance
Hypothesis Explanation
/ Rejection
Logistics innovativeness has a significant positive effect on perceived
H1 Acceptance
organizational performance.
There is no significant difference in logistics innovativeness responses
H2a Rejection
according to gender demographic variable.
There is a significant difference in perceived organizational performance
H2b Acceptance
responses according to gender demographic variable.
There is no significant difference in logistics innovativeness responses
H2c Rejection
according to marital status demographic variable.
There is no significant difference in perceived organizational performance
H2d Rejection
responses according to the marital status variable.
There is no significant difference in logistics innovativeness responses
H2e Rejection
according to age demographic variable.
There is a significant difference in the perceived organizational performance
H2f Acceptance
responses according to the age variable.
There is no significant difference in logistics innovativeness responses
H2g Rejection
according to the educational level variable.
There is no significant difference in perceived organizational performance
H2h Rejection
responses according to educational status.
There is no significant difference in logistics innovativeness responses
H2ı Rejection
according to the tenure variable.
There is a significant difference in perceived organizational performance
H2j Acceptance
responses according to tenure.
There is no significant difference in logistics innovativeness responses
H2k Rejection
according to the logistics sector variable.
There is no significant difference in perceived organizational performance
H2l Rejection
responses according to the logistics sector.

DISCUSSION AND CONCLUSION

In the logistics and procurement activities within the value chain, the necessity of companies that
want to gain competitive advantage over rival companies to implement innovation-oriented
competitive strategies is discussed in a conceptual framework. In order to fulfill this requirement,
the relationship between logistics innovativeness and organizational performance has been
revealed in this study. In addition, accepting logistics service providers as sampling companies
enabled the logistics innovation approach to be evaluated from the perspective of logistics
employees. On the other hand, the perceived organizational performance scale, which is accepted
as a subjective organizational performance evaluation method, and the organizational
performance status of logistics companies were determined from the perspective of employees.
Considering the findings obtained within the scope of the research, it is seen that Logistics
innovativeness has a significant positive effect on perceived organizational performance. In the
study by Şen and Bolat (2015) examining the relationship between innovation and firm
performance of logistics companies, it is understood that the finding that innovation performance
has a significant effect on firm performance is in parallel with the findings obtained within this

115
research. Findings obtained from studies examining the relationship between logistics
innovativeness and performance explain that logistics innovativeness has a significant effect on
performance (Raltson et al., 2013; Rahim et al., 2020). It is understood that the findings obtained
from studies examining the relationship between logistics innovation and performance in reverse
logistics processes are in the same way as the findings obtained within the scope of our research
(Richey et al., 2005; Huang & Yang, 2014; Piyachat, 2017; Hutomo et al., 2018). In studies on
the calculation of sustainable logistics innovation values of companies and their relationship with
sustainable company performance, the relationship between innovation and performance has a
significant relationship (Andersson & Forslund, 2018). Looking at the relationship between
different innovation types and performance, it is observed that the innovation-performance
relationship supports the meaningfulness (Shaukat et al., 2013). Kabadurmuş (2015) examines
the relationship between logistics performance and innovation, research and development, new
product, new process and new organization variables are the dependent variables, transportation
costs are the independent variables, firm age, firm size, export status and other costs are the
control variables. As a result of analyzing the wide model structure, it has been concluded that
the transportation costs affect the R&D activities at the highest level. Although there are many
studies in the literature dealing with the relationship between innovation and performance, it is
seen that the studies on logistics innovativeness are not at a sufficient level. It is also found in the
literature where the relationship between innovation and performance is examined through
logistics companies (Acar & Günsel, 2010; Kalmuk & Acar, 2015; Shou et al., 2017; Ardito et
al., 2018; Acar, 2020).
With this research, the concept of logistics innovativeness has been discussed and its effect on
perceived organizational performance has been revealed. It is one of the main points reached as a
result of the research that logistics capabilities, which constitute the main competitive power of
logistics companies, can play an active role in creating a competitive power by adopting the
understanding of logistics innovation by exhibiting innovation-oriented approaches. In addition,
considering the gender, age and tenure of the participants, it is seen that they affect the perceived
organizational performance and that performance evaluation differences occur according to the
group differences. The lack of differentiation in the logistics innovativeness responses of the
logistics companies and the demographic characteristic of the participants also explains that the
logistics innovativeness understanding is not affected by the differentiation of logistics service
areas.
References

Acar, A. Z. (2020). The mediating role of value innovation between market orientation and business
performance: evidence from the logistics industry. International Journal of Business Innovation
and Research, 21(4), 540-563. doi:10.1504/ijbir.2020.106012
Acar, Z., & Günsel, A. (2010, September). The effects of process innovation in logistics service.
In Proceedings of the 5th European Conference on Innovation and Entrepreneurship (pp. 1-1).
Allen, R.S., Dawson, G., Wheatley, K. and White, C.S. (2007), Perceived diversity and organizational
performance, Employee Relations, Vol. 30 No. 1, pp. 20-33. doi:10.1108/01425450810835392
Anderson, N. R., & West, M. A. (1998). Measuring climate for work group innovation: development and
validation of the team climate inventory. Journal of Organizational Behavior: The International

116
 Journal of Industrial, Occupational and Organizational Psychology and Behavior, 19(3), 235-
258. doi:10.1002/(sici)1099-1379(199805)19:3<235::aid-job837>3.0.co;2-c
Andersson, P., & Forslund, H. (2018). Developing an indicator framework for measuring sustainable
logistics innovation in retail. Measuring Business Excellence, 22(1), 1-13. doi:10.1108/mbe-04-
2017-0017
Anning-Dorson, T. (2018). Customer involvement capability and service firm performance: The
mediating role of innovation. Journal of Business Research, 86, 269-280.
doi:10.1016/j.jbusres.2017.07.015
Ardito, L., Messeni Petruzzelli, A., Dezi, L., & Castellano, S. (2018). The influence of inbound open
innovation on ambidexterity performance: Does it pay to source knowledge from supply chain
stakeholders? Journal of Business Research. doi:10.1016/j.jbusres.2018.12.043
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of management, 17(1),
99-120.
Bass, F. M. (1969). A new product growth for model consumer durables. Management science, 15(5),
215-227.
Bekmezci, M., Orçanlı, K., ve Eroğluer, K. (2020). Entelektüel Sermayenin Örgütsel Performansa
Etkisinde Bilgi Paylaşımının Düzenleyici Rolü, İşletme Araştırmaları Dergisi, 12 (2), 1585-1604.
doi.:10.20491/isarder.2020.931
Berberoglu, A. (2018). Impact of organizational climate on organizational commitment and perceived
organizational performance: empirical evidence from public hospitals. BMC health services
research, 18(1), 1-9. doi.:10.1186/s12913-018-3149-z
Büyüköztürk, Ş., Çakmak, E. K., Akgün, Ö. E., Karadeniz, Ş., & Demirel, F. (2017). Bilimsel araştırma
yöntemleri. Pegem Atıf İndeksi, 1-360.
Carmeli, A., & Schaubroeck, J. (2008). Organisational crisis-preparedness: The importance of learning
from failures. Long range planning, 41(2), 177-196. doi:10.1108/meq-01-2018-0003
Delaney, J. T., & Huselid, M. A. (1996). The impact of human resource management practices on
perceptions of organizational performance. Academy of Management journal, 39(4), 949-969.
doi:10.2307/256718
Dess, G. G., & Robinson Jr, R. B. (1984). Measuring organizational performance in the absence of
objective measures: the case of the privately‐held firm and conglomerate business unit. Strategic
management journal, 5(3), 265-273.
Drucker, P.F. (2000). 21. Yüzyıl için Yönetim Tartışmaları. (Çev. İrfan Bahçıvangil, Gülenay Gorbon),
2.b., Epsilon Yayınları, İstanbul.
Eris, E. D., & Ozmen, O. N. T. (2012). The effect of market orientation, learning orientation and
innovativeness on firm performance: A research from Turkish logistics sector. International
Journal of Economic Sciences & Applied Research, 5(1).
Evans, J.D. (1996). Straightforward Statistics for the Behavioral Sciences. Pacific Grove, Brooks/Cole
Publishing, USA.
Eygü H. ve Kılınç A. (2019). Sosyo–Ekonomik Gelişmişlik Algısı Üzerinde Etkili Olan Faktörlerin Sıralı
Logit Model Yardımıyla Araştırılması: Erzurum–Kayseri Örneği, Journal of Academic Value
Studies, 5, 1023-1040. doi: 10.29228/javs.38807
Fawcett, S.E., Birou, L. and Cofield-Taylor, B. (1993), Supporting global operations through logistics and
purchasing, International Journal of Physical Distribution and Logistics Management, Vol. 23
No. 4, 3-11. doi:10.1108/09600039310041464
Ferraresi, A. A., Quandt, C. O., Santos, S. A. D., & Frega, J. R. (2012). Knowledge management and
strategic orientation: leveraging innovativeness and performance. Journal of Knowledge
Management, 16(5), 688-701. Doi.:10.1108/13673271211262754
Gebauer, H. (2011). Exploring the contribution of management innovation to the evolution of dynamic
capabilities. Industrial Marketing Management, 40(8), 1238-1250.
doi.:10.1016/j.indmarman.2011.10.003

117
Gunday, G., Ulusoy, G., Kilic, K., & Alpkan, L. (2011). Effects of innovation types on firm
performance. International Journal of production economics, 133(2), 662-676.
doi.:10.1016/j.ijpe.2011.05.014
Huang, Y. C., & Yang, M. L. (2014). Reverse logistics innovation, institutional pressures and
performance. Management Research Review, 37(7), 615-641. doi.:10.1108/MRR-03-2013-0069
Hutomo, A., Saudi, M. M., & Sinaga, H. O. (2018). The mediating effect of sustainability control system
on reverse logistics innovation and customer environmental collaboration towards sustainability
performance. Journal of Fundamental and Applied Sciences, 10(1S), 705-721.
doi.:10.4314/jfas.v10i1s.53
Kabadurmuş, F. N. K. (2019). The Relationship Between Logistics Performance and Innovation: An
Empirical Study of Turkish Firms. Alphanumeric Journal, 7(2), 157-172.
Kalaycı, Ş. (2005), Uygulamalı Çok Değişkenli İstatistik Teknikleri, Asil Yayın Dağıtım, Ankara, s.322
Kalmuk, G., & Acar, A. Z. (2015). The mediating role of organizational learning capability on the
relationship between innovation and firm's performance: A conceptual framework. Procedia-
Social and Behavioral Sciences, 210, 164-169. doi: 10.1016/j.sbspro.2015.11.355
Kemppainen, K. and Vespalainen, A.P.J. (2007), Logistical and technological differentiation as a
precondition of supply networking, International Journal of Logistics Management, Vol. 18 No.
1, 81-101. doi.:10.1108/09574090710748180
Khandwalla, P. (1977). The design of organizations. New York: Harcourt Brace Jovanovich.
Kim, D., Cavusgil, S. T., & Calantone, R. J. (2006). Information system innovations and supply chain
management: channel relationships and firm performance. Journal of the academy of marketing
science, 34(1), 40-54.
Kitchot, S., Siengthai, S. and Sukhotu, V. (2020), The mediating effects of HRM practices on the
relationship between SCM and SMEs firm performance in Thailand, Supply Chain Management:
An International Journal, Vol. 26 No. 1, doi: 10.1108/SCM-05-2019-0177.
Kline, P. (1994). An easy guide to factor analysis. Routledge.
Lebas, M. J. (1995). Performance measurement and performance management. International journal of
production economics, 41(1-3), 23-35.
Lee, E. S., & Song, D. W. (2010). Knowledge management for maritime logistics value: discussing
conceptual issues. Marit. Pol. Mgmt., 37(6), 563-583. Doi: 10.1080/03088839.2010.514959
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of
sample size for covariance structure modeling. Psychological methods, 1(2), 130.
Midgley, D. F., & Dowling, G. R. (1978). Innovativeness: The concept and its measurement. Journal of
consumer research, 4(4), 229-242.
Moideenkutty, U., Al‐Lamki, A., & Murthy, Y. S. R. (2011). HRM practices and organizational
performance in Oman. Personnel Review.
Panayides, P. (2006). Enhancing innovation capability through relationship management and implications
for performance. European Journal of Innovation Management, 9(4), 466-483.
doi:10.1108/14601060610707876
Peng, B., Zheng, C., Wei, G., & Elahi, E. (2020). The cultivation mechanism of green technology
innovation in manufacturing industry: From the perspective of ecological niche. Journal of
Cleaner Production, 252, 119711. doi.:10.1016/j.jclepro.2019.119711
Piyachat, B. (2017). The relationships among resources’ commitment reverse logistics innovation reverse
logistics performance and reverse logistics cost savings: Manufacturing vs service
industry. Journal of Administrative and Business Studies, 3(3), 122-135.
Rahim, M. F., Asim, M., & Manzoor, S. (2020). The Effect of Logistics Salience on Logistical
Capabilities and Firm Performances. European Journal of Business and Management
Research, 5(1).
Ralston, P. M., Grawe, S. J., & Daugherty, P. J. (2013). Logistics salience impact on logistics capabilities
and performance. The International Journal of Logistics Management, 24(2), 136-152.
doi.:10.1108/IJLM-10-2012-0113

118
Richey, R. G., Genchev, S. E., & Daugherty, P. J. (2005). The role of resource commitment and
innovation in reverse logistics performance. International Journal of Physical Distribution &
Logistics Management, 35(4), 233-257. doi.:10.1108/09600030510599913
Salunke, S., Weerawardena, J., & McColl-Kennedy, J. R. (2011). Towards a model of dynamic
capabilities in innovation-based competitive strategy: Insights from project-oriented service
firms. Industrial Marketing Management, 40(8), 1251-1263.
doi.:10.1016/j.indmarman.2011.10.009
Seo, Y. J., Dinwoodie, J., & Kwak, D. W. (2014). The impact of innovativeness on supply chain
performance: is supply chain integration a missing link?. Supply Chain Management: An
International Journal. doi.:10.1108/SCM-02-2014-0058
Shaukat, S., Nawaz, M. S., & Naz, S. (2013). Effects of innovation types on firm performance: An
empirical study on Pakistan's manufacturing sector. Pakistan Journal of Commerce and Social
Sciences (PJCSS), 7(2), 243-262.
Shou, Y., Shao, J., & Chen, A. (2017). Relational resources and performance of Chinese third-party
logistics providers: the mediating role of innovation capability. International Journal of Physical
Distribution & Logistics Management, 47(9), 864-883. doi.:10.1108/IJPDLM-09-2016-0271
Subramanian, A., & Nilakanta, S. (1996). Organizational innovativeness: Exploring the relationship
between organizational determinants of innovation, types of innovations, and measures of
organizational performance. Omega, 24(6), 631-647. doi.:10.1016/S0305-0483(96)00031-X
Şen, E. ve Bolat, M. (2015). İşletmelerde Demokratik Yönetim Anlayışının İnovasyon ve Firma
Performansı Üzerine Etkisi: İstanbul Avrupa Yakası Lojistik İşletmeleri Üzerine Bir Uygulama,
İstanbul Ticaret Üniversitesi Sosyal Bilimler Dergisi, (27), ss. 149-172.
Tabachnick, B., & Fidell, L. (2013). BG Tabachnick. LS fidell using multivariate statistics (sixth ed.)
Pearson, Boston.
Weerawardena, J., & Mavondo, F. T. (2011). Capabilities, innovation and competitive
advantage. Industrial Marketing Management, 40(8), 1220-1223.
doi.:10.1016/j.indmarman.2011.10.012
Zeffane, R. and S.J. Bani Melhem, 2017. Trust, job satisfaction, perceived organizational performance
and turnover intention: A public-private sector comparison in the United Arab Emirates.
Employee Relat., 39: 1148-1167. doi.:10.1108/ER-06-2017-0135

119
 DEMAND FORECASTING IN RETAIL: A MACHINE LEARNING BASED
APPROACH FOR ACCURATE SHORT-TERM FORECASTS

Mehmet Mustafa Özalp1, Dr. Özgür Akarsu2

Abstract ⎯ Demand planning is still the core process of supply chain management in which accurate forecasts
of customer demand are the basis of all critical supply chain processes such as inventory management, planning
& scheduling of production, and management of replenishment. Since the retail sector highly relies on end-
customer expectations and dynamics, it is crucial to foresee future customer behavior to improve the processes,
decrease costs, and ensure product readiness. However, due to the rapidly changing environments, complex
competition dynamics, and unexpected, rare events such as fluctuations in the economy or lockdowns due to a
pandemic, it is hard to forecast future demand with conventional methods. In this study, we propose a multi-
layered machine learning approach to forecast future demand regarding historical sales, intra-organizational
factors as well as macro socio-economic factors. Our results show that it is possible to create significant
business value in supply chain processes by forecasting the short-term demand accurately in retail.

Keywords ⎯ demand forecasting, demand planning, machine learning, retail, supply chain optimization

1. INTRODUCTION
Demand forecasting is one of the most popular topics in many sectors such as retail, production, service, energy,
and finance. Increasing competition and rapidly changing environments escalated the importance of demand
forecasting, since knowing the future demand of internal or external customers enables organizations to make
more effective plans for resource allocation and ensure that the supply of the required products is served at the
right time and right place.
Therefore, demand forecasting has become one of the key activities in the supply chain since most decisions in
practice aim to make supply meet demand (Zotteri & Kalchschmidt, 2007). Having an accurate estimation of
the future demand of customers, enables organizations to make better and more rational arrangements based on
data both at the operational and strategic levels. In this sense, demand forecasting as the voice of the customer
or the sense of the market exposed by the data, provides a baseline for purchasing, storage, transfer, pricing, and
positioning decisions on the operational level; long-term resource plans, investment and growth commitments,
determination of product and market strategies on the strategic level.
In this manner, for decades researchers from different backgrounds have been working to build models that
explain past trends and foresee the future. Despite the importance of having accurate demands is vital for many
organizations, the complex nature of the phenomena makes it very difficult to achieve. The difficulty not only
comes from the internal dynamics which are shaped by changes in product hierarchy, marketing, and sales
strategies, even warehouse planning, but also countless external factors like competition dynamics, socio-
economic transformations, government regulations, etc. which affect the patterns of demand.
In the retail sector, in which planning, maintaining, and executing the supply chain processes are located as the
main part of the core business processes, the potential benefits of having accurate forecasts are more visible and
tangible than most of the other sectors. During Covid 19 pandemic which hit the world in 2020, the effect of
having a smart and adaptive demand forecasting process has become more crucial, because of the inexhaustible
turbulence in the markets related to the pandemic situation and government measures. Therefore, in today’s
rapidly changing markets where sales trends have high volatility, product lifetimes are relatively short and
consumer behavior is unstable, demand forecasting models should have the ability to inherit all potential internal
and external factors dynamically to have accurate forecasts about the future.

1
Mehmet Mustafa Özalp, Istanbul Technical University, Faculty of Management, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]
2
Dr. Özgür Akarsu, KocDigital Çözümler A.Ş., [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

120
Recent advances in big data technologies and machine learning methods have facilitated the tools and techniques
that provide researchers and practitioners to build models that handle the complexities of contemporary markets.
Correspondingly, this prospective study was designed to investigate the use of machine learning methods in
demand forecasting in retail using an extensive list of internal and external factors. By utilizing a holistic
approach that aims to cover all of the quantifiable current factors and building an extendable model
infrastructure that enables future modifications, having accurate forecasts continuously has been guaranteed in
the long-term.

2. LITERATURE REVIEW
Demand forecasting modeling approaches can be categorized as traditional methods based largely on time-series
methods which aim to cover mostly historical trends and patterns, and machine learning methods that
additionally match nonlinear relationships between inputs and outputs (Fildes et al, 2019).
Traditional time series analysis methods constitute the largest part of demand forecasting literature. Models
such as exponential smoothing (Brown, 1959), Holt Winters (Winters, 1960), Box & Jenkins (Box and Jenkins,
1969), regression models (Papalexopoulos and Hesterberg, 1990), and Auto-regressive integrated moving
average (ARIMA) have been used frequently and achieved adequate results in many different fields (Kuo and
Xue, 1999).
However, recent research especially in the field of retail, assert that standard time series models are sometimes
insufficient in predicting sales because of the non-linearity and volatility which exist in historical retail sales
(Alon et al, 2001; Chu and Zhang, 2003; Kuvulmaz et al, 2005. Zhang and Qi, 2005).
On the other hand, as an alternative to the time-series approach which has been widely used for decades, the
machine learning techniques offer a novel methodology that yields better results for large data sets where
assumptions of linearity, stationarity, and conformity to normal distribution are not easily assured (Fildes et al.,
2019; Ren et al, 2020). Nevertheless, as a prospective state-of-the-art solution for forecasting problems, machine
learning gained great interest both from academia and practitioners in the field. There were plenty of studies in
which machine learning models such as SVM (Al-Musaylh et al, (2018), Ghalehkhondabi et al, (2017), Ricardo
et al, (2018), Lu (2014)), ANN (Tiwari and Adamowski (2015), Mouatadid and Adamowski (2017), Loureiro
et al, (2018), Bibaud-Alves et al, (2019)), Decision Trees (Thomassey (2010), Warrior et al, (2016), Castillo et
al, (2017), Ntaliakouras et al, (2019)), Extreme Learning Machine ((Wong and Guo (2010), Choi et al, (2014))
were used to forecast future demand accurately.
Machine learning approaches offer great flexibility to use intra-organizational factors such as sales history,
inventory, price changes, store sizes, campaigns etc.; inter-organizational factors such as market share, supplier
dynamics, spare-part inventory, competitive forces etc.; and lastly external factors such as inflation rate, regional
demographics, interest rate, weather etc. that potentially affect demand to take as an input to explain fluctuations
in demand (Thomassey, 2010).
Since the retail sector highly relies on end-customer expectations and dynamics, it is crucial to foresee future
customer behavior to improve the supply chain process, decrease costs and ensure product readiness. Strategic,
tactical, and operational plans of the retail sector depend on demand forecasting and accurate demand forecasts
have a big impact on retail supply chain performance. This dependency led researchers to conduct new studies
in retail demand forecasting (Fildes et al, 2019).
Ali et al. (2009) studied a demand forecasting method using regression trees based on time series data of sales
and promotions at stock-keeping unit (SKU) and store level in a European grocery retail chain. They emphasized
that time series techniques yield better results in non-promotional periods compared to regression trees, but
regression trees significantly increase prediction accuracy for promotion and such complex inputs.
Wong and Guo (2010) conducted experiments to compare hybrid Extreme Learning Machine, Neural Network,
and ARIMA methods, where they used a city-category level aggregated data of a Chinese clothing retail chain.
According to their results, the hybrid model they proposed can effectively solve the medium-term sales forecast
problem, indicating that the proposed model can provide much superior performance compared to the traditional
ARIMA models and the two newly developed sales forecast neural network models.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

121
Thomassey (2010) elaborated on industry-specific dynamics in their study on clothing retail; by examining the
studies conducted in this sector, he proposed models that include fuzzy logic, artificial neural networks, and
data mining studies with more accuracy and reliability. In this study, two different levels of the horizon were
proposed as a long-term horizon (about one year) and a short-term horizon (few weeks) according to the features
involved. As a result, forecasting models should logically be adapted to the selected horizon since constraints
are not similar. For the long-term horizon, proposed models were based on seasonality and rely on fuzzy
inference systems for the influence of explanatory variables. Sales forecasts with a short-term horizon were
performed with a neural network which enables the system to update the long-term forecasts according to the
last known sales.
Yu et al (2011) developed a model which combined the extreme learning machine method and traditional
statistical methods by using a fashion retailer’s sales data. They compared the model they developed with
polynomial regression and traditional artificial neural network methods and obtained better results in terms of
mean square error.
Choi et al (2014) created a smart prediction algorithm by combining the extreme learning machine and grey
model for a fast-fashion retailer. Even though the model was trained with a relatively small amount of data, it
created very good predictions under certain time and data constraints. Moreover, with an artificial data set, it
was indicated that the model performed well when the slope of the demand trend or the variance of the seasonal
cycles is large.
Kaya et al (2014) developed an adaptive hybrid model which generates forecasts at the product category level
for a Turkish clothing retailer. In this study, by calculating seasonal patterns and price effects, a seasonal pattern
was revealed, and it was combined with average predictions of the initial model.
In another comparative study, by Gaur et al (2015), the performance between the KNN algorithm and Bayes
Networks using Walmart sales data was compared. The results showed that Bayes networks technique passes
the Nearest Neighbor algorithm.
Di Pillo et al (2016) stated that statistical methods such as ARIMA and exponential smoothing (ES) could fail
to model nonlinear behaviors when there is a high irregularity in historical sales data. In their research, support
vector machines (SVM) were used to build a forecasting model which also considers promotional effects. Their
results showed that SVM model performed better than time-series models such as ARIMA and ES.
Kharfan and Chan (2018) proposed two frameworks using machine learning techniques using point-of-sale
(POS) data from a leading US-based shoe retailer. As a result, prediction models based on machine learning
techniques achieved better prediction accuracy compared to the current performance of the company.
Additionally, in their study, key factors that are influencing demand were identified.
Krishna et al (2018) conducted a study on the application of different machine learning models for demand
forecasting for retail sales and compared the performance results. They applied both regression algorithms such
as multiple regression, polynomial regression, ridge regression, lasso regression, and boosting algorithms like
AdaBoost, XGBoost, and GradientBoosting to the same data set. In their study, among all of the algorithms
which were applied, Gradient Boosting method displayed the best performance.
Deep learning models which have attracted great attention in the field of machine learning in recent years were
also used in demand forecasting in retail. Kaneko and Yada (2016) used three-year point-of-sale data from a
retail store to build a sales forecast using deep learning model via H2O framework. By using L1 regulation,
prediction accuracy of their model achieved to 86% and this achievement is 13% greater than the model they
made for comparison with logistic regression. Along with this result, they stated that deep learning is more
suitable for creating models with multi-specific variables.
Slimani et al (2016) applied a multi-layer perceptron structure to build a daily demand forecasting model for a
retail store. On the other hand, Vhatkar and Dias (2016), trained an artificial neural network (ANN) for the sales
of oral care products and stated that the choice of model for estimation depends entirely on the type of data set.
Another study using deep learning models was conducted by Loureiro et al. (2018) for clothing retail. In their
study, the performance of the deep learning models was compared with other machine learning algorithms such
as decision trees, random forest, support vector machine. The results implied that even though deep learning

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

122
models performed well for predicting sales in the fashion retail market, some of the other machine learning
techniques like random forest produced better results.
Lastly, Wang et al. (2019) proposed a hybrid approach for retail companies to choose the correct prediction
model for certain product types instead of using a single model. In their work, they compared classical time-
series models and the most recent forecasting technologies according to durable and non-durable products
respectively. The results implied that SVM was more suitable for perishable products, whereas LSTM is the for
non-perishable items due to its low cost and excellent forecast performance.
Huber and Stuckenschmidt (2020) specified that one key challenge is to forecast demand on special days that
are subject to vastly different demand patterns than on regular days and presented the case of a bakery chain
with an emphasis on special calendar days. They discussed the possibility of formulating a classification
approach instead of a regression problem. They also specified that an empirical comparison with established
approaches reveals the superiority of machine learning methods, while classification based approaches
outperform regression-based approaches. Moreover, they found that machine learning methods not only provide
more accurate forecasts but are also more suitable for applications in a large-scale demand forecasting scenario
that often occurs in the retail industry.
Güven and Şimşir (2020) established an artificial neural-network (ANN) and support vector machine (SVM)
model that considered color parameters for the retail apparel industry to forecast demand. As a result of their
study, it has been determined that ANN outperformed SVM on seven datasets out of ten for the datasets without
color and their performances were even for the datasets with color.
Tian et al. (2021) proposed a Markov-combined method (MCM) for forecasting intermittent retail demands by
using historical sales and inventory status. They applied the MCM method to two large datasets of two different
e-commerce companies in China. As a result, they stated that the MCM method was more accurate to forecast
intermittent demand than the Single Exponential Smoothing (SES), Syntetos-Boylan Approximation (SBA),
and Croston (CR) methods.
Wolters and Huchzermeier (2021) proposed a framework to forecast demand accurately under the combination
of seasonal sales variations and promotion-included sales peaks over the forecasting horizon. Firstly, they
forecast the seasonal sales cycle by fitting a harmonic regression model to a decomposed training set, which
excludes promotional and holiday sales, and then extrapolate that model to a testing set. Secondly, they integrate
the resulting seasonal forecast into a multiplicative demand function that accounts for consumer stockpiling and
captures promotional and holiday sales uplifts. The final model is then fitted using ridge regression. They use
sales data from a grocery retailing chain to compare the forecasting accuracy and they stated that the
significantly more accurate forecasts generated by their model attest to the merit of the approach developed
here.
To sum up, the literature review conducted presented a rewarding ground for further applications of machine
learning to demand forecasting problems. In terms of, forming more accurate results which capture the nonlinear
trends and effect of factors that create irregularity in historic sales data, machine learning offers a promising
approach.

3. METHODOLOGY
In this study, a machine learning-based forecasting model was applied in a retail company to forecast future
sales of products with different sales patterns and volumes over time. The product portfolio of the company
consisted of durable nonperishable products which were supplied in the dealer, wholesale, and online sales
channels. Since the dealer channel covered %87 of all sales, it was selected as the main scope of the forecasting
model.
During the business understanding step of the research, several workshops were conducted with the process and
data owners from the company’s supply chain department to identify the forecast horizon and possible indicators
that affect the sales trends of different product categories. Information gathered during this step was used while
building the modeling framework to determine the possible features to be included in the model and granularity
level that could match with requirements of the supply chain process.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

123
The general solution framework of the demand forecasting model was summarized in Figure 1. Since the
requirements of the demand forecasting model are related to the supply chain process characteristics, possible
future changes could trigger rework in all of the solution steps such as data ingestion, feature engineering, or
model development. To minimize such possible future development efforts, the solution framework was built
in a modular and extendable framework.

Figure 1. Overview of Solution Framework.

3.1. Determination of Granularity Level

During the data understanding step, historical data that covers the sales transactions between 2015 to 2019 was
collected as the primary source of modeling. Exploratory analysis of the collected data showed us that the
distribution of sales among products is not uniformly distributed and there are fluctuations between product
sales trends over years. Therefore, the sales distribution of product categories was analyzed and products to be
included in the model were determined. The final coverage of the case study included 13 distinct products with
59 unique categories.
Determining the demand forecasting unit (DFU) and the optimum level of granularity in terms of product,
geographical, customer/channel hierarchy is one of the main issues in demand forecasting in the supply chain.
Using more granular data lowers the specification error (Synetos, et.al, 2017) but on the other hand, makes it
difficult to collect continuous transactional data that capture the fluctuations patterns over time. Since the
assumptions of stationarity and non-correlated demand (Zotteri and Kalchschmidt, 2007) between different
levels of product, geographic, and customer/channel hierarchies mostly are not valid in real-life cases, lower
levels of granularity could be preferred to ensure the data availability which is essential especially for machine
learning-focused forecasting approaches.
Thus, granularity in this case study was determined at a relatively higher level as ‘the product category’ for
product hierarchy, ‘geographical regions’ for geographical hierarchy, and ‘brand level’ for customer/channel
hierarchy, to increase the historical life span of the demand forecasting unit.

Table 1. Granularity & DFU Levels

Product Categories Regions Brands # of DFU

59 5 2 59*5*2=590

Historical sales transaction data were aggregated according to the stated granularity level on weekly basis. By
merging this weekly historical sales data with externally collected data, the basis for the demand forecasting
model was constructed which will be used to create features and train machine learning models in the later
stages.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

124
3.2. Feature Engineering
The feature engineering process which can be described as formulating and extracting the most appropriate
features given the data (Zheng and Casari, 2018) consists of two main sections: Feature creation and feature
selection. In the demand forecasting framework, the former includes combining and enrichment of different
features to describe the trends in historic sales while the latter, covers the selection of features to be included in
the machine learning algorithm which provides the best model performance.

3.2.1. Feature Creation

During the problem definition and business understanding step of the study, several data sources that could
explain the fluctuation of the demand were determined and historical data were collected. Careful examination
of the past data showed possible signs of relations between the targeted value and its possible predictors. The
collected data categories and their content is presented in Figure 2. The first category “historic sales data”, aims
to capture the historical sales trends and intra-organizational dynamics that affect sales. Price changes, product
attributes, inventory level, campaigns, and promotions were all grouped under this category. Secondly,
“Calendar data” enables the model to learn the effect of seasonality, holidays, special days, and other related
events on sales. Thirdly, “external dynamics” include all possible socio-economic indicators that affect sales.
Changes in the tax regulations, currency exchange rates, and other economic indicators as well as weather
conditions were included in the model training to provide the model necessary data to explain the fluctuations
of sales over time. Last but not least, during the implementation phase, the emergence of the global pandemic
of Covid-19 virus led us to quantify the status of the pandemic and add pandemic-related features to the model
to understand and relate the pandemic-related sales fluctuations. Global pandemic data provided by “Our World
in Data” (https://ourworldindata.org/) and mobility data provided by google
(https://www.google.com/covid19/mobility/) were collected and transformed into inputs for the model training
phase.

Figure 2. Collected Data for Demand Forecasting Model

Using the collected data, explicit features were created using three different groups of calculations which can
be grouped as statistical transformations, creation of auto-regressive features, and creation of calendar features.
The first one aimed to transform the discrete observations of several variables which were present for unique
timestamps, into periodical metrics by applying basic statistical calculations. Hence, by calculating the weekly
average, standard deviation, percent change, maximum or minimum of variables such as sales price, inventory
level, currency rate. etc. several features were created which represent the weekly status of the related variable.
The second group of transformations focused on creating auto-regressive features which would add the
sequential nature of time-series data to the machine learning model. Since demand forecasting is a problem that
is based on sequential data collected over time, the target value at time t, is dependent on the target values

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

125
collected at timestamps such as t-1, t-2, t-3, etc. Therefore, we created shifted sales for t-1, t-2, … t-52 weeks
to cover both short-term, seasonality, and year effects of historical sales while forecasting the future demand.
Similarly, to avoid possible data leakages, shifted transformations were also applied to collected sequential data
such as currency rate, customer confidence index, inflation rate, etc. Since it wouldn’t possible to know the
future values of these variables while forecasting the future demand, their shifted t-n features were included in
the model training. Lastly, calendar dates that were marked with special events were created as days after or
days past features to quantify their effect on consumer demand and buying habits.
After all of these transformation and calculation steps, 257 distinct features were created as inputs to the machine
learning model.

3.2.2. Feature Selection

The main goal of the feature selection process is to reduce the number of input features which leads to a decrease
in computational costs and improvement in the model performance. Two common methods for applying feature
selection are univariate filter methods (such as Chi-square test, Fisher’s score, Correlation Coefficient, etc.) and
multivariate wrapper methods (such as forward feature selection, backward feature elimination, etc.). Both two
methods have their advantages and disadvantages. The former does not consider the interaction between features
whereas the latter increases the cost of computing extensively especially when there are many potential input
features (Zheng and Casari, 2018).
In this study instead of selecting only one of the two feature selection approaches, we combined these methods
to reduce computational cost and improve model performance. Firstly, input features were grouped according
to their source of the domain (economic features, price/promotion features, calendar features, etc.), and
univariate analysis was conducted to filter out the features with inconsistency and poor data quality before the
model development phase.
Secondly, a tree-based algorithm was trained, and input features were sorted by importance produced by the
trained machine learning model. The first n most important features were held, and the rest of the features were
added or dropped iteratively according to their contribution to model forecasting accuracy. Finally, a subset of
features that increase model performance was selected.

3.3. Model Development

Demand forecasting problem is based on sequential time series data and traditionally solved with time series
prediction methods. However, by creating auto-regressive lagged features and a target column which is the
historical sales for the selected time unit, algorithms used for regression problems can be easily applied. By
using regression models, it is possible to build a robust modeling framework that gives outputs for large numbers
of forecasting units at the same time. Especially, when the different forecasting units have similar trends and
sales behaviors, the model can transfer information from one forecasting unit to another and provides a flexible
and cost-efficient modeling framework for demand forecasting problems.
Before starting to build the forecasting model, the main problem was to explore the forecasting units with similar
patterns and group them to be modeled together. To do so, a clustering algorithm was used to determine the
granularity clusters according to patterns (sales pattern, sum and average sales, deviations etc.) of the forecasting
units.
The main modeling framework of this study consist of 2 main sections which can be seen in Figure 3. The first
section focuses on the short-term demand 4 weeks forward for each model run. The second section focuses on
the long-term forecast for 20 weeks. For the former, a machine learning approach and for the latter time-series
model which aims to cover the trend and seasonality of the sales in long run were applied. Since the future
values of external features are impossible to predict and their lagged effects are meaningless in the long run,
time-series models which capture the trend and seasonality of the historical sales are the only option for long-
term demand forecasting.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

126
To obtain accurate forecasts for the short term, a tree-based regression model was applied to the historical data
set to forecast the future demand of each forecasting unit. Tree-based regression models can be used efficiently
in regression problems when the number of variables and the number of cases is high (Torgo, 1999). Unlike
linear models, they map non-linear relationships quite well. They are adaptable at solving any kind of problem
at hand (classification or regression) (https://www.analyticsvidhya.com/blog/2016/04/tree-based-algorithms-
complete-tutorial-scratch-in-python/, access: 25.06.2021). Ferreira et al (2015) stated that the benefit of using
regression trees is that they do not require the specification of a certain functional, parametric form between
features and demand; the model is more general in this sense. Moreover, tree-based algorithms are more
interpretable for business owners (Ferreira et. al, 2015).
Additionally, tree-based models are more responsive to market demand fluctuations, when geospatial or
temporal variations of certain independent variables affect the demand dramatically. In this use case since the
number of forecasting units (590) and the number of the possible features (257) is relatively high and the size
of the collected historical data is small (3 years of historical sales) tree-based regression models were selected
to build the short-term accurate forecasting model. Among tree-based regression models, LightGBM with its
high speed, low computational requirements, and low memory usage (Ke et al, 2017) was selected.

Figure 3. Forecasting Model Framework

To ensure the generalizability of the forecasting model, the last 52 weeks of the collected data were categorized
as test period and was used to evaluate model performance. A weighted percentage error metric defined in
Equation 1 was used to evaluate the model performance. For each product category, geographical granularity
level, and model run the sum of the difference between model prediction and actual values were calculated. The
absolute value of the deviation is divided into the sum of actual values to calculate a percentage error and lastly,
subtraction of weighted absolute percentage error from 1 provides the ‘accuracy’ value of the model. This
accuracy formulation not only enabled us to calculate and monitor model performance in the lowest granularity
level, but also provided a weighted aggregation to check the model performance at any desired level. Since in
our retail case, the distribution of sales quantity among product categories is long tailed, balancing the effect of
products with low sales volume on overall accuracy calculation is crucial.
WAPE formulation automatically increases the effect of the top sales products while calculating the overall
accuracy. Therefore, using the WAPE metric while optimizing the forecasting model, top-selling products
which are also more important for reducing inventory costs in the company’s supply chain process have more
effect than the low sales volume products.

∑𝑖,𝑗,𝑘,𝑡 𝑦
^−∑𝑖,𝑗,𝑘,𝑡 𝑦
𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑𝐴𝑏𝑠𝑜𝑙𝑢𝑡𝑒𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒𝐸𝑟𝑟𝑜𝑟(𝑊𝐴𝑃𝐸) = ∑𝑖,𝑗,𝑘,𝑡 𝑦
*100 (1)

𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 = (1 − 𝑊𝐴𝑃𝐸) (2)

© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

127
Where;
i: Product category
j: Geographical level
k: Model run date
t: Forecast horizon
y^: Sales prediction value
y: Actual sales value

Finally, since the forecasting models were built in a relatively higher level of granularity, the outputs should be
merged and disaggregated to SKU and dealer level to provide the required inputs for the supply chain operations
of the company. The disaggregation step is based on business rules determined by analysis of historical sales,
stocks, purchase orders, and other related business rules.

4. FINDINGS
The forecasting model was built using a train data set which consists of 3 years of historical sales starting from
2016 January and tested for the last 52 weeks of available data (March 2019 to March 2020).
The forecast outputs generated by Lightgbm model gave an overall accuracy score of %84. Accuracy values
for top-selling 5 products can be seen in Table 2. The forecast accuracy values of top-selling 5 product categories
are all above %75. This finding also proves that the model produces good forecasting outputs for every critical
product and the distribution of the deviation is balanced.

Table 2. Accuracies of the 5 top-selling products

Sales Percent Accuracy

Pr-A 26% 80.2%

Pr-B 15% 79.3%
Pr-C 12% 84.0%
Pr-D 12% 75.0%
Pr-E 9% 82.4%

To understand the behavior of the model for different product categories, accuracy values were compared with
the sales volume and percentage for each category. As stated in Figure 4, sales volume and model accuracy
score is positively correlated (r= 0.40). Except for the several outliers, as the sales volume and sales percentages
of product categories increase, we notice an increase in forecasting accuracy as well.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

128
 Figure 4. Accuracy Scores and Sales Volume for Product Categories

The comparison of total actual sales with forecasted values for each week during the test period in Figure li5
shows that the model has captured the fluctuations of sales with great success except certain weeks where there
is a huge incline or decline in the realized sales.

Figure 5. Total forecast and actual comparison for the test period

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

129
 5. DISCUSSION OF RESULTS AND IMPLICATIONS FOR MANAGEMENT
The results of this study provide new insights into the demand forecasting literature by indicating a methodology
that covers three fundamental issues which contribute to critical challenges in the field.
Firstly, in many business cases due to the product portfolio with many SKUs or dispersed geographical locations
in which the companies are operating, the number of forecast units increases dramatically. Thus, building and
sustaining robust models for each forecasting unit becomes difficult. Especially, machine learning based
approaches which require intense effort to select features, tune hyperparameters and monitor the performance
metrics, are very time-consuming when each forecasting unit is modeled by itself. In this sense, the methodology
proposed in this study solves this issue by modeling forecast units with similar historical patterns in the same
machine-learning model. This approach not only reduces the number of models to be built and monitored but
also lets different forecasting units learn from each other’s patterns by transferring knowledge between
forecasting units. Moreover, in cases where the average lifetime of SKUs is short and there is not enough
historical data, the approach applied in this use case enables transfer knowledge between forecasting units and
respectively enhances the forecasting power of the model. As we have seen in the findings section, the %84
overall accuracy score obtained in this study is promising for similar cases which contain products with similar
patterns.
Secondly, the sequential nature of data in demand forecasting problems enables us to make future forecasts by
identifying the trend and seasonality in the past data. However, in contemporary rapidly changing markets where
numerous macro and micro factors create a complex eco-system that affects the customer behavior, relying only
on on-trend and seasonality causes us to miss short-term fluctuations in demand. Hence, any model which aims
to explain the past fluctuations in demand and accurately forecast the future should include quantifiable external
factors such as inflation, currency rates, market shares, or weather condition which possibly affect the customer
demand. In this use case, adding external features to explain the past fluctuations in demand enabled us to build
a high-performance forecasting model. The comparison of actuals and predictions for the test period showed
that apart from the outliers, the model was able to capture the short-term fluctuations in the demand accurately.
Lastly, one of the major problems of demand forecasting solutions requirements for modification occurs due to
the changes in product portfolio or geographical dispersion. Forecasting models should be easily modified to
add new SKUs/dealers or exclude dying SKUs/closed dealers. On the other hand, models should be responsive
to changes in internal and external dynamics and unforeseen repetitive events should be able to be included in
the model if they affect demand. In this use case, handling these two essential requirements provided us to obtain
a high accuracy value. Factors such as currency fluctuations, covid cases, and country measures were quantified
to create a more robust model. The overall architecture was designed in a modular manner to enable future
modification requirements.

6. CONCLUSION
The importance of demand forecasting in the retail sector increased significantly due to the rising competition
in global economies and the appearance of business models which locate customers as the primary trigger of
operations. The ones who understand customer behavior were able to serve the required products and services
to their customers on time with high quality and at the same time optimize their costs in their supply chain
processes. However, the same dynamics which increased the importance of demand forecasting also
transformed traditional forecasting into a complex multivariate problem in which numerous factors cause
unexpected fluctuations in customer demand. Therefore, in today’s world, any approach which aims to forecast
customer demand has to capture these current and potential challenges.
So, it is possible to say that the machine learning based demand forecasting which was applied in this study in
a retail company offers a generalizable framework to solve these issues. By providing an approach that covers
huge numbers of products and geographical levels and enabling fast modifications for future cases, machine
learning based demand forecasting has very critical implications in the field. Our results proved that even in a
very complex ecosystem with huge numbers of forecasting units, machine learning based demand forecasting
offers an accurate short-term forecast.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

130
Secondly, one of the major implications of this study is the stated importance of understanding the domain
characteristics for demand forecasting. While determining the granularity level, transforming the model outputs
to business outcomes, and building the modeling framework deep-dive analysis of components produced
important insights. Additionally, using unsupervised clustering algorithms to determine which granularity
groups should be modeled together enabled a solid base to build an end-to-end solution. Without relying on
intuitive business decisions and human intervention, we build a responsive and adaptable model which learns
from experience and is generalizable for future cases.
For further research, this approach could be tested in different sectors other than retail, and also other machine
learning techniques could be applied and compared with the outputs of the current model. Moreover,
disaggregation rules which were designed to transform model outputs to low-level operational requirements
could be autonomous instead of relying solely on business rules.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

131
REFERENCES
A., Babai, Z., Boylan, J., Kolassa S., Nikolopoulos, K., (2017), Supply Chain Forecasting: Theory, Practice,
their Gap and the Future. European Journal of Operational Research.
Ali, Ö.G., Sayin, S., van Woensel, T., Fransoo, J., (2009), SKU demand forecasting in the presence of
promotions. Expert Systems with Applications.
Al-Musaylh, M.S., Deo, R.C., Adamowski, J.F., Li, Y., (2018), Short-term electricity demand forecasting with
MARS, SVR and ARIMA models using aggregated demand data in Queensland, Australia. Advanced
Engineering Informatics.
Alon, I., Qi, M., & Sadowski, R. J., (2001), Forecasting aggregate retail sales: A comparison of artificial neural
networks and traditional methods. Journal of Retailing and Consumer Services.
Bibaud-Alves, J., Thomas, P., El Haouzi, H.B., (2019), Demand forecasting using artificial neuronal networks
and time series: Application to a French furniture manufacturer case study. IJCCI 2019 - Proceedings of the
11th International Joint Conference on Computational Intelligence.
Box, G.E.P., Jenkins, G.M., (1969), Time Series Analysis Forecasting and Control, Prentice Hall, Englewood
Cliffs.
Brown, R.G., (1959), Smoothing Forecasting and Prediction of Discrete Time Series. Prentice Hall, Englewood
Cliffs.
Castillo, P.A., Mora, A.M., Faris, H., Merelo, J.J., Snchez, P.G., Fernandez-Ares, A.J., De las Cuevas, P.,
Garcia-Arenas, M.I., (2017), Applying computational intelligence methods for predicting the sales of newly
published books in a real editorial business management environment.
Choi, T., Hui, C., Liu, N., Ng, S., & Yu, Y., (2014), Fast fashion sales forecasting with limited data and time.
Chu, C.W., & Zhang, G.P., (2003), A comparative study of linear and nonlinear models for aggregate retail
sales forecasting. International Journal of Production Economics.
Di Pillo, G., Latorre, V., Lucidi, S., Procacci, E., (2016), An application of support vector machines to sales
forecasting under promotions.
Ferreira, K.J., Lee, B.H.A., Simchi-Levi, D., (2015), Analytics for an Online Retailer: Demand Forecasting and
Price Optimization.
Fildes, R., Ma, S., Kolassac, S., (2019), Retail Forecasting: Research and Practice. International Journal of
Forecasting.
Gaur, M., Goel, S., Jain, A., 2015, Comparison between Nearest Neighbours and Bayesian Network for Demand
Forecasting in Supply Chain Management.
Ghalehkhondabi, I., Ardjmand, E., Young, W.A., Weckman, G.R., (2017), Water demand forecasting: review
of soft computing methods. Environmental Monitoring and Assessment.
Güven, İ., Şimşir, F., (2020), Demand forecasting with color parameter in retail apparel industry using artificial
neural networks (ANN) and support vector machines (SVM) methods. Computers & Industrial Engineering.
Huber, J., Stuckenschmidt, H., (2020), Daily retail demand forecasting using machine learning with emphasis
on calendric special days. International Journal of Forecasting.
https://www.analyticsvidhya.com/blog/2016/04/tree-based-algorithms-complete-tutorial-scratch-in-python/,
access: 25.06.2021.
Kaneko, Y., Yada, K., (2016), A Deep Learning Approach for the Prediction of Retail Store Sales. IEEE
International Conference on Data Mining Workshops, ICDMW.
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

132
Kaya M., Yeşil, E., Dodurka, M.F., Sıradağ, S., (2014), Fuzzy Forecast Combining for Apparel Demand
Forecasting.
Ke, J., Zheng, H., Yang, H., Chen, X.M., (2017), Short-term forecasting of passenger demand under on-demand
ride services: A spatio-temporal deep learning approach. Transportation Research Part C: Emerging
Technologies.
Kharfan, M., Chan, V.W.K., (2018), Forecasting Seasonal Footwear Demand Using Machine Learning.
Krishna, A., Akhilesh, V., Aich, A., Hegde, C., (2018), Sales-forecasting of Retail Stores using Machine
Learning Techniques.
Kuo, R.J., Xue, K.C., (1999), Fuzzy neural networks with application to sales forecasting. Fuzzy Sets and
Systems.
Kuvulmaz, J., Usanmaz, S., Engin, S.N., (2005), Time-series forecasting by means of linear and nonlinear
models. MICAI 2005: Advance in artificial intelligence.
Loureiro, A.L.D., Miguéis, V.L., da Silva, L.F.M., (2018), Exploring the use of deep neural networks for sales
forecasting in fashion retail. Decision Support Systems.
Loureiro, A.L.D., Miguéis, V.L., da Silva, L.F.M., (2018), Exploring the use of deep neural networks for sales
forecasting in fashion retail.
Lu, C., (2014), Sales forecasting of computer products based on variable selection scheme and support vector
regression.
Mouatadid, S., Adamowski, J., (2017), Using extreme learning machines for short-term urban water demand
forecasting. Urban Water Journal.
Ntaliakouras, N., Vonitsanos, G., Kanavos, A., Dritsas, E., (2019), An Apache Spark Methodology for
Forecasting Tourism Demand in Greece. 10th International Conference on Information, Intelligence, Systems
and Applications.
Papalexopoulos, A.D., Hesterberg, T.C., (1990), A regression-based approach to short-term system load
forecasting, IEEE Transactions on Power Systems.
Ren, S., Chan, H., Siqin, T., (2020), Demand forecasting in retail operations for fashionable products: methods,
practices, and real case study. Annals of Operations Research.
Ricardo, H., Gonçalves, I., Costa, A.C., (2018), Forecasting tourism demand for Lisbon’s region through a data
mining approach. Proceedings of the 11th IADIS International Conference Information Systems.
Slimani, I., El Farissi, I., Achchab, S., (2016), Configuration of Daily Demand Predicting System Based on
Neural Networks.
Thomassey, S., (2010), Sales forecasts in clothing industry: The key success factor of the supply chain
management.
Tian, X., Wang, H., Erjiang, E., (2021), Forecasting intermittent demand for inventory management by retailers:
A new approach. Journal of Retailing and Consumer Services.
Tiwari, M.K., Adamowski, J.F., (2015), Medium-term urban water demand forecasting with limited data using
an ensemble wavelet-bootstrap machine-learning approach. Journal of Water Resources Planning and
Management.
Torgo, L., (1999), Tree-based regression models are known for their simplicity and efficiency when dealing
with domains with large number of variables and cases. https://www.dcc.fc.up.pt/~ltorgo/PhD/

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

133
Vhatkar, S., Dias, J., (2016), Oral-Care Goods Sales Forecasting Using Artificial Neural Network Model.
Wang, J., Liu, G.Q., Liu, L., (2019), A Selection of Advanced Technologies for Demand Forecasting in the
Retail Industry.
Warrior, K.P., Shrenik, M., Soni, N., (2016), Short-term electrical load forecasting using predictive machine
learning models. 2016 IEEE Annual India Conference.
Winters, P.R., (1960), Forecasting sales by exponential weighed moving averages. Management Science.
Wolters, J., Huchzermeier, A., (2021), Joint In-Season and Out-of-Season Promotion Demand Forecasting in a
Retail Environment. Journal of Retailing.
Wong, W., Guo, Z., (2010), A hybrid intelligent model for medium-term sales forecasting in fashion retail
supply chains using extreme learning machine and harmony search algorithm.
Yu, Y., Choi, T.M., Hui, C.L., (2011), An intelligent fast sales forecasting model for fashion products.
Zhang, G.P., & Qi, M., (2005), Neural network forecasting for seasonal and trend time series. European Journal
of Operational Research.
Zheng, A., Casari A., (2018), Feature Engineering for Machine Learning, O'Reilly Media.
Zotteri, G., Kalchschmidt, M., (2007), A model for selecting the appropriate level of aggregation in forecasting
processes. Int. J. Production Economics.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

134
 SUBORBITAL SPACE TOURISM

Enes Demiralay1, Engin Hasan Çopur2, Turan Paksoy3

Abstract ⎯ Due to the recent technological advances, the space tourism is widely believed to be possible in
near future. When the revolution in the tourism takes place, the economic growth rate of the countries where
the investment in this newly emerging tourism sector is made will tend to increase in long term. For, more
private companies will be involved in the space tourism industry. Suborbital space tourism appears to be
ahead of other types of space tourism due to its relatively low cost required to perform a suborbital
spaceflight in which a space vehicle reaches the outer space but completes its flight without making a full
orbit around a self-gravitating body where it is launched. Virgin Galactic and Blue Origin are the most
famous companies in a tight race to become the leading company in the realm of the suborbital space tourism
industry. According to the recent surveys, it is predicted that more than 2 million people will enjoy this
memorable and fascinating trip to the outer space in the coming years. In this study, the current state of the
art is presented, and a suborbital space tourism matrix is proposed.

Keywords ⎯ Airports, Space, Spaceports, Suborbital space tourism

1. INTRODUCTION
In recent decades, one of the most asked questions is whether the space could be a new touristic environment
for the people living on Earth. This is very interesting question, yet it is not surprisingly brand new. The
expression “space tourism” has been used in the discussions about the future of the military and civil space
technologies since the mid-1960s (Rogers, 2001). Today, a 50 years long dream of humans came true thanks
to the successful launches to the edge of spaces that were managed by the privately funded companies.
Although one can count on one hand the number of the companies doing business in the space tourism
industry, their goals to pave the way for opening the space frontier to everyone are incredibly valuable and
capable of changing the future of the tourism fundamentally. However, only two launches to suborbital space
with some civilian crew has been completed successfully until July 2021. Henceforth, it is likely that more
suborbital space launches will be broadcasted around the world live in near future if the cost for suborbital
space flight is reduced to a more affordable amount. According to the news, some tickets have already been
sold at a price of around $200,000 to experience how great it feels to be in a non-gravitational environment
and to admire the beauty of the universe while watching Earth from the outer space (Amos, 2021; Sheetz,
2021).
Turkey is one of the countries in which the incomes from the tourism sector are highly important to their
economic growth rate. Therefore, it is very vital to analyze the international tourism demand and identify the
proper actions that should be taken for realizing the economic growth strategy. As a country with a fascinating
history and incredible natural beauties, thereby one of the most beautiful touristic centers, Turkey will have to
review its international tourism strategy, considering the recent trends in the tourism industry. Now it is the
time to consider whether the developments of the space tourism can affect the sustainability of Turkey’s
international tourism strategy or not. At first glance the answer seems that there is no direct relationship
between them and thereby no significant influences could be observed on the economic growth rate in Turkey.
Indeed, it is highly possible. However, this study approaches the question from a different perspective based
on the case of Cappadocia, a historical region in the middle of Anatolia. Since Cappadocia has got a
spectacular landscape for hot air balloon tours, great number of foreign and domestic tourists visit the region
every year according to the tourism statistics (Aslaner, 2019; Karakuş, 2019; Özel & Kozak, 2017). The

1
Enes Demiralay, Konya Technical University, Faculty of Engineering and Natural Sciences, Department of Industrial Engineering,
Konya, Turkey, [email protected]
2
Engin Hasan Çopur, Necmettin Erbakan University, Faculty of Aeronautics and Astronautics, Department of Astronautical
Engineering, Konya, Turkey, [email protected]
3
Turan Paksoy, Necmettin Erbakan University, Faculty of Aeronautics and Astronautics, Department of Aviation Management,
Konya, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
135
density of the tourist circulation is also very vital for other neighboring tourist facilities such as located near
the seaside in Antalya since many tourists who intend to join a hot air balloon trip in Cappadocia are also
planning to visit a seaside resort located at the south or west of Turkey. With the benefit of hindsight from the
Cappadocia case, this study gives an answer to the question of the impacts of the developments in space
tourism on the Turkey’s tourism strategy.
It is forecasted that the space tourism will be at the heart of the tourism industry in the future and the countries
capable of sustaining the space tourism industry will be able to receive a large amount of incomes. Therefore,
the private or public investments of Turkey in the space technologies are likely to cause to lead to the
development of space tourism in Turkey. If this vision becomes reality, Turkey will increase not only its share
of the pie in the international tourism market but also increase the number of the foreign visitors in the future,
enabling a way to support the current tourism facilities in the different regions. Like Cappadocia case, foreign
tourists will not visit Turkey for only one suborbital space journey. On the contrary, they will probably visit
other beautiful regions in Turkey after or before the launch to the edge of the space. Therefore, the space
tourism will influence the whole tourism industry in Turkey more positively than expected.
Now, it is worth noting that Turkey will have to assess its current infrastructure systems and then the
necessary infrastructure improvements will need to be made according to a national strategy. In this year, the
first step was taken by Turkish Space Agency with the announcement of the national space programme. In this
programme, the important strategies, policies and missions are identified to improve the technical skills of
Turkey in the developments of space technologies. This programme will inevitably help develop the space
tourism industry in Turkey. However, despite all these great progress, Turkey will suffer from the lack of its
own launch site if not planned properly. Therefore, this study focuses on this issue and discusses whether
some of the existing airports are likely to be expanded to enable suborbital space launches.

2. CURRENT STATE OF THE ART FOR SPACE TOURISM

Space tourism is not a brand-new concept, but its popularity has been rapidly growing, especially in the last
decade. The fame of the space tourism spreading quickly around the world owes much, of course, to the
cutting-edge developments in space technology. In the literature, there are different definitions of space
tourism but all the definitions have one central meaning, i.e. space tourism is a commercial space travel for
non-scientific or non-military activities (Duval & Hall, 2015; Yazıcı & Tiwari, 2021). Space tourism can be
divided into four main categories: i) Earth-based and cyberspace tourism, ii) atmospheric and suborbital
tourism, iii) orbital space tourism, and iv) space trip beyond Earth orbit (Cater et al., 2015; Weibel, 2020).
The first subcategory includes the Earth-based activities such as visits to the space centers around the world as
well as simulation-based activities such as virtual gaming environments and virtual reality travel. In addition
to those, travels to see the sun eclipse and the Northern lights are also other examples of Earth-based activities
that do not need to leave the Earth surface (Toivonen, 2020). The second category is completely different
since in the activities of this category, the tourists can experience a feeling of weightlessness in an airplane or
a spacecraft launched to reach mostly an altitude higher than 100 km and then return to the location where it is
launched without completing a full orbit around the Earth. In the third category, the main aim is to establish a
low-orbit hotel that is able to offer an extraordinary accommodation to the guests or provide the fascinating
view of the Earth from a spacecraft orbiting around the Earth. The last category includes the tourism activities
in which, for example, tourists can stay in a hotel built on Moon or Mars surface. Another option would be to
make an orbital spaceflight around Moon or Mars.
Table 1 is used to present the possible space tourism destinations. The first row indicates the journey from
Earth to the tourism destinations. As mentioned above, the popular tourism destinations can be listed as Earth
orbit, Moon, Moon orbit, planet such as Mars and planet orbit. Besides the destinations located in Earth’s
atmosphere or outer space, there are some attractive terrestrial facilities where tourists can observe sky with
telescopes, gain a shuttle launch experience or obtain some useful information about the flight to Moon. The
first column of Table 1 indicates the return ride from the tourism destinations to Earth. On the other hand, it is
supposed that there will be some other journeys, for example from Moon to Mars, in future, especially after
human beings are capable of colonizing planets in outer space. When this happens, the blank cells of Table 1
will be indicating the new space tourism trips.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

136
 Table 1. The space tourism matrix
Earth (E) Earth Orbit (EO) Moon Orbit (MO) Planet Orbit (PO) Planet (P)
Earth (E) E2E E2EO E2MO E2PO E2P
Earth Orbit (EO) EO2E - - - -
Moon Orbit (MO) MO2E - - - -
Planet Orbit (PO) PO2E - - - -
Planet (P) P2E - - - -

3. CONCLUSION
With the help of the developing space technologies, a new dimension has been opened for the tourism sector.
The theoretical and practical framework is still taking shape in this new sector called space tourism. In the
literature, there is a few studies on suborbital space tourism. Therefore, it is believed that this study can be
regarded as one of the leading studies in literature for this newly developing field. In this study, the current
practices in the world on space tourism are summarized and a matrix modelling the framework of space
tourism applications is proposed.

4. REFERENCES
Amos, J. (2021). Virgin Galactic: Sir Richard Branson rockets to the edge of space.
https://www.bbc.com/news/science-environment-57797297
Aslaner, M. A. (2019). Hot-air balloon tour accidents in the Cappadocia region. Aerospace Medicine and
Human Performance, 90(2), 123–127. https://doi.org/10.3357/AMHP.5121.2019
Cater, C., Garrod, B., & Low, T. (2015). The Encyclopedia of Sustainable Tourism. CABI.
https://books.google.com.tr/books?id=txbGCgAAQBAJ
Duval, D., & Hall, C. (2015). Sustainable space tourism: new destinations, new challenges. In The Routledge
Handbook of Tourism and Sustainability (pp. 468–478). Routledge.
https://doi.org/10.4324/9780203072332-50
Karakuş, Y. (2019). Identifying the Best Alternative Tourism Product for a Destination: the case of
Cappadocia. Anais Brasileiros de Estudos Turísticos - ABET, 9, 1–9. https://doi.org/10.34019/2238-
2925.2019.v9.16929
Özel, Ç. H., & Kozak, N. (2017). An exploratory study of resident perceptions toward the tourism industry in
Cappadocia: a Social Exchange Theory approach. Asia Pacific Journal of Tourism Research, 22(3),
284–300. https://doi.org/10.1080/10941665.2016.1236826
Rogers, T. F. (2001). “Space tourism”-- its importance, its history, and a recent extraordinary development.
Acta Astronautica, 49, 537–549. https://doi.org/10.1016/S0094-5765(01)00125-4
Sheetz, M. (2021). Jeff Bezos reaches space on Blue Origin’s first crewed launch.
https://www.cnbc.com/2021/07/20/jeff-bezos-reaches-space-on-blue-origins-first-crewed-launch.html
Toivonen, A. (2020). Sustainable Space Tourism: An Introduction. Channel View Publication.
https://books.google.com.tr/books?id=mSkGEAAAQBAJ
Weibel, D. L. (2020). Space Tourism: The Elusive Dream. Journal of Tourism and Cultural Change, 18(6),
743–746. https://doi.org/10.1080/14766825.2020.1777644
Yazıcı, A. M., & Tiwari, S. (2021). Space Tourism: An Initiative Pushing Limits. Journal of Tourism Leisure
and Hospitality, 3(1), 38–46.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

137
 CHANGES IN SUPPLY CHAINS DURING THE COVID-19 PANDEMIC

Ahmet ÇETİNDAŞ1

Abstract ⎯ The COVID-19 epidemic affected the whole world economically and sociologically which lead to
disruptions in supply chains. Due to the lockdowns, production interruptions and demand uncertainty occurred.
The demand boom for some products caused supply issues. Supply chains of live-saving products such as drugs,
medical equipment, and food were protected from the beginning but other supply chains faced significant
disruptions. The reasons for that were unpredictable changes in the demand and the origin of the outbreak which
is China the most exporting country in the world. Companies that designed their supply chains more agile and
flexible and made improvements by adopting the innovative attitude overcame the impact of the epidemic with
minimum damage. It is important for success that during the pandemic, companies maintain their activities and
adopt some strategies such as sustainability in the supply chain, dynamic response strategies, technology usage,
the culture of collaboration, diversified supply bases, and synchronized strategic processes.

Keywords ⎯ COVID-19 Pandemic, Supply Chain, Change, Disruption

INTRODUCTION
It is clear that the economic and sociological effects of the Covid-19 epidemic are enormous. The pandemic
has exposed vulnerabilities and fragilities in most of the global supply chains. A survey applied and reported by
Fortune (2020) showed that 94% of Fortune 1000 listed businesses are experiencing supply chain issues due to
COVID-19. Vidya and Prabheesh (2020) conducted a network analysis for the period (Q1) regarding the effects of
Covid-19 on global networks and countries connected to the global trade network such as China, South Korea, Italy,
Japan, USA and Germany. It was concluded that the supply chain was interrupted and the total trade intensity
decreased from 0.833 to 0.429. Although these problems decreased in the following months, demand uncertainties
caused by the epidemic, which continued to show its effect in most countries, continued to strain supply chains
throughout 2020 too. The exposure of the current global supply chain's weaknesses during the COVID-19 era has
caused problems such as loss of revenue, unfulfilled demand and supply (Linton & Vakil, 2020). It should also be
seen as a learning process for embracing resilience and robustness in the supply chain in a shrinking economy
(Currie, 2020:87).
This study has researched the effect of the epidemic on supply chains. The unstable demand and restrictions of
states lead to supply chain disruptions. China, where the virus spread plays an important role in the global economy.
Many companies in many countries are using in their production Chinese materials or finished products. After the
government applied lockdowns in china, the production decreased which affected its export too. These lead to
supply chain disruptions in the whole World which was also the focus of this paper. The reasons for supply chain
disruptions during the epidemic were explained and possible solutions have been proposed. Finally, some
suggestions to companies which faced supply chain disruptions are made.

SUPPLY CHAINS IN THE COVID-19 PANDEMIC

Although all industries are interconnected through complex supply chains and logistics networks, operations
have been severely disrupted during the COVID-19 pandemic. Worldwide, economic activities reached their lowest
level with an unprecedented decline, and the World Economic Forum (WEF) predicted that this would cause even
more economic recessions and global financial crises (Lucchese & Pianta, 2020:99). The COVID-19 pandemic,
which affected the global economy and shook supply and trade significantly throughout the period, also caused

1
Ahmet ÇETİNDAŞ, Hasan Kalyoncu University, Faculty of Economics, Administrative and Social Sciences, Department of International
Trade and Logistics, Gaziantep, Turkey, [email protected]

138
long-term disruptions and high uncertainties in supply chains. Supply chain disruptions as a result of such a crisis
can be characterized as unpredictable long-term risks. Disruptions in the supply chain that occur in connection with
the spread of the epidemic cause simultaneous disruptions between supply, demand and logistics. The disruptions
in the supply chains of firms caused a decrease in their performance, which increased the importance of addressing
such interruptions in order to effectively and successfully ensure the operations of procurement (Aldrighetti et al,
2019:1).
One of the most important risks facing supply chains during the COVID-19 pandemic is the difficulty of
forecasting the demand for services and goods (Grida et al, 2020:2). Due to partial or full curfews, demand for most
goods and services has decreased in most countries. These prohibitions have disrupted both national and
international supply chains. Another reason for the decreased demand is that the epidemic caused job losses in some
sectors and thus economic difficulties. (De Vito & Gomez, 2020:1).
It is important to understand in which direction global supply chains can evolve and which possible solutions
are considered by companies. Solving this problem is not easy, as today's supply chains are incredibly complex with
partners from multiple geographies connected to the global trade ecosystem (Sharma et al, 2020:444). Due to the
origin of the virus and the role of Chine in global production, the impact was most felt in China. There are more
than 200 Fortune Global 500 companies only in the city of Wuhan, where the pandemic started and spread to the
World (IFC, 2020:2). China, the largest exporter country has experienced supply shortages in its main export goods
such as metals, textiles, plastics, auto parts, electronics, pharmaceuticals etc. After mid-March, lockdowns were
applied in many countries. These caused to a decrease in production in these countries and thus a decrease in demand
for parts produced in China. The sudden outbreak led to significant mismatches between supply and demand for
special goods such as masks, medical gowns, disinfectant products, chemicals, PCR tests, respirators, etc. These
goods were normally required in limited quantities but after the epidemic, a demand boom came up (Dente &
Hashimoto, 2020:1).
The amount of inventory that companies generally hold to meet the demand varies between two and five weeks
depending on the supply chain strategy. Shipping from China to the USA or Europe takes an average of 30 days
and to Turkey takes 25 days. Due to the lead times, if material supply is interrupted for for more than a certain
period, it means that many companies will stop production. For example, Fiat Chrysler Automobiles announced on
February 14 that it was temporarily suspending production at its automobile factory in Serbia, as it could not get
parts from China. Similarly, Hyundai has announced that it has decided to suspend production activities at its
factories in Korea due to disruptions in the supply of parts from China. The pandemic has also had significant effects
on the high-tech industry. Apple announced that its earnings in the first three months of 2020 were much lower than
expected (Haren & Simchi-Levi, 2020).
The Covid-19 pandemic has demonstrated the vulnerability of extensive and complex value chains around the
world to production disruptions. From a supply chain perspective, disruptions in China have created problems in
many countries. In response, many of the countries trying to minimize the negative effects of the virus on their
businesses and supply chains by choosing their supply chain members from closer alternative partners or by bringing
their strategic value chains to their homeland. Shortening supply chains has benefited countries with skilled
manufacturing sectors and fair export policies (eg Colombia, India and Mexico) to partially replace China in the
medium term. There has also been a trend towards placing additional storage areas near demand centers to reduce
time to market for goods (IFC, 2020:5).
As a result of the restrictions applied due to the epidemic, the inability of truck drivers to reach the containers
arriving in China and the cancellation of ship voyages for outgoing cargo caused containers to accumulate in the
ports. These problems experienced in China, which supplies raw materials and parts to the whole world, brought
the production operations in countries to a standstill. Thus, major industries around the world such as automotive,
electronics, pharmaceuticals and medical equipment, as well as consumer goods have been affected (IFC, 2020:2).
Regarding material stocks, it is known that unlike earthquakes, floods, volcanic eruptions and other natural
disasters, the epidemic did not destroy any capital assets, but only slowed down the flow of people and the
processing of materials within the economy. In other words, although the use of capital assets remains low as long

139
as lockdowns are implemented, it is expected to return to normal levels at the end of the epidemic crisis. Further
demand shrank was caused as a result of the unemployment and loss of income during the period. The epidemic
risked turning into an economic crisis that led to the obsolescence of capital assets and the deterioration of material
stocks (Dente & Hashimoto, 2020:2).
Increasing pressure to reduce supply chain costs is pushing companies to implement strategies such as lean
manufacturing, offshoring and outsourcing. These cost cutting strategies brings risk of supply disruptions, especially
in times of demand instability as in the epidemic (Haren ve Simchi-Levi, 2020). In order to reduce the effects of the
epidemic on supply chains, companies need to design their supply chains more agile and flexible and make
improvements by adopting an innovative attitude according to the course of the epidemic (Remko. 2020). The
logistics system is crucial in managing disruptions and recovery of supply chains (Choi 2020:5). The effects of
Covid-19 on manufacturing, distribution and retailing and the global supply chain of many disrupted products and
services require forward-looking strategies. Companies trying to maintain their business during the Covid-19
pandemic have realized that strategies such as sustainable supply chain, dynamic response strategies, technology
use, culture of collaboration, diversified supply base and synchronized strategic processes play a vital role for
success (Sharma et al, 2020:447).
In infectious diseases such as Covid-19 that pose a risk to public health medical logistics and food logistics
should be protected. Therefore, also in this epidemic, it was tried to eliminate the disruptions in the supply of drugs,
medical equipment and food at the first place (Chen et al, 2020:2).
The Supply of Drugs and Medical Equipment
The main strategies to prevent human-to-human transmission of COVID-19 infection are social distancing and
the use of disinfectants along with masks. Due to the public's need for masks and disinfectants, significant pressures
have arisen in the supply chains of health products. On the other hand, health personnel needed more personal
protective equipment. For example, since the tremendous increase in demand for N-95 masks cannot be met, it has
become very difficult to reach it in many countries (Singh et al., 2020:6). The reason for the high infection and
death rates in Italy, where the epidemic has spread the most, is thought to be the lack of personal protective
equipment. In the United States, it has been observed that the demand for respirators and masks is much higher than
the supply (Jacobs et al, 2020)
China, which produced about half of the world's mask needs before Covid-19 (Bradsher & Alderman, 2020), could
not export either masks or other health equipment after the epidemic first spread. With the decrease in the number
of cases in China, although it started to export necessary health products to other countries, it was never possible to
meet the booming demand (Ivanov & Dolgui, 2020). Since the producers in the health industry planed their capacity
according to the expected demand, they could not meet the demand boom in such a fast time. Hereby, besides new
investors in the health industry, many manufacturers from different sectors turned their productions into health
products. For example, automotive manufacturers in India seized the opportunity and have turned to produce
medical equipment such as respirators, test kits, and oxygen cylinders (Francisco & Heinzmann 2020). There have
been many similar transformations in Turkey as well. Under the leadership of the Ministry of Health, the first
domestic respirator was produced with the cooperation of the companies Baykar, ASELSAN, Arçelik and Biosis
(Yavuz et al, 2020). Further, many textile and fashion companies started to produce masks, gloves and hospital
textiles and many chemical factories started to produce disinfectants and cleaning products.
CONCLUSION
The COVID-19 virus which has affected the whole world economically and sociologically has also exposed
vulnerabilities and fragilities in many global supply chains. In the first quarter of 2020, due to the COVID-19
pandemic, 94% of the enterprises included in the Fortune 1000 list experienced supply chain problems. The supply
chains of countries that lead the global trade network such as China, South Korea, Italy, Japan, Germany and the
USA were interrupted, and the total trade volume decreased from 0.833 to 0.429. Although these problems
decreased in the following months, the demand uncertainties caused by the epidemic, which continued to show its
effect in many countries, continued to strain and disrupt the supply chains throughout 2020. From the moment the

140
epidemic occurred, drug, medical equipment and food supply chains were protected and isolated to prevent
disruptions. In other industries, it has been seen that disruptions in supply chains are caused by two important
reasons. The first reason is the unpredictable changes in the demand for many goods and services due to the
economic difficulties caused by the lockdowns and the loss of jobs in some sectors. The second reason is that the
epidemic has emerged in China which plays a very important role in the global supply chains. In order to reduce
the effects of COVID-19 on their supply chains, companies need to design their supply chains more agile and
flexible and make improvements by adopting an innovative attitude according to the course of the epidemic. It is
important for success that during the pandemic, companies maintain their activities and adopt some strategies such
as sustainability in the supply chain, dynamic response strategies, technology usage, culture of collaboration,
diversified supply bases, and synchronized strategic processes.

REFERENCES
Aldrighetti, R., Zennaro, I., Finco, S. & Battini, D. (2019). Healthcare supply chain simulation with disruption
considerations: a case study from northern Italy,” Glob. J. Flex. Syst. Manag. 20 (1), 81–102, 2019
Bradsher, K. ve Alderma, L. (2020). https://www.nytimes.com/2020/03/13/business/masks-china-
coronavirus.html
Choi, T.M. (2020). Innovative ‘Bring-Service-Near-Your-Home’OperationsUnder Corona-Virus (COVID-
19/SARSCoV-2) the Outbreak: Can Logistics Become the Messiah? Transportation Research Part E: Logistics and
Transportation Review, 101961,
Currie, C.S.M., Fowler, J.W., Kotiadis, K., Monks, T., Onggo, B.S., Robertson, D.A. & Tako, A.A. (2020). How
Simulation Modelling Can Help Reduce the Impact of COVID-19. Journal of Simulation 14 (2): 83–97,
De Vito, A. & Gomez, J.P. (2020). Estimating the COVID-19 cash crunch: global evidence and Policy. J. Account.
Public Policy 106741,
Dente, S.M.R & Hashimoto, S. (2020). COVID-19: A pandemic with positive and negative outcomes on resource
and waste flows and stocks. Resources, Conservation and Recycling, 161, 104979,
Fortune (2020). 94% of the Fortune 1000 are seeing coronavirus supply chain disruptions: Report. Retrieved March
30, 2020, from https://fortune.com/2020/02/21/fortune-1000-coronavirus-china-supply-chain-impact/. (Accessed
Date: 09.09.2020).
Francisco, B. & Heinzmann, T. (2020). From Perfume toHand Sanitiser, TVs to FaceMasks: How Companies are
Changing Track to Fight COVID-19. Advanced Manufacturing and Production. World Economic Forum, March
24. https://www.weforum.org/agenda/archive/covid-19?page=28. (Accessed Date: 20.02.2020).
Grida, M., Mohamed, R. & Zaied, A.N.H. (2020). Evaluate the impact of COVID-19 prevention policies on
supply chain aspects under uncertainty. Transportation Research Interdisciplinary Perspectives, volume 8,
100240,
Haren, P. & Simchi-Levi, D. (2020). How Coronavirus Could Impact the Global Supply Chain by Mid-March.
Harvard Business Review https://hbr.org/2020/02/how-coronavirus-could-impact-the-globalsupply-chain-by-mid-
march. (Accessed Date: 15.09.2020).
IFC (2020). The Impact of COVID-19 on Logistics. International Finance Corporation,
https://www.ifc.org/wps/wcm/connect/industry_ext_content/ifc_external_corporate_site/infrastructure/resources/t
he+impact+of+covid-19+on+logistics (Accessed Date: 05.02.2021).
Ivanov, D. & Dolgui, A. (2020). Viability of intertwined supply networks: Extending the supply chain resilience
angles towards survivability. A position paper motivated by COVID-19 outbreak. International Journal of
Production Research, 1–12,

141
Linton, T. & Vakil, B. (2020). Coronavirus Is ProvingWe Need More Resilient Supply Chains. Harvard Business
Review, https://hbr.org/2020/03/coronavirus-is-proving-that-we-need-more-resilient-supply-chains (Accessed
Date: 15.01.2021).
Lucchese, M. & Pianta, M. (2020). The Coming Coronavirus Crisis:What CanWe Learn? Intereconomics 55 (2):
98–104,
Remko, V.H. (2020). Research Opportunities for a More Resilient Post-COVID-19 Supply Chain – Closing the
Gap between Research Findings and Industry Practice. International Journal of Operations & Production
Management 40 (4): 341–355,
Sharma, A., Adhikary, A. & Borah, S.B. (2020). Covid-19′s impact on supply chain decisions: Strategic insights
from NASDAQ 100 firms using Twitter data. Journal of Business Research 117, 443–449,
Sharma, A., Adhikary, A. & Borah, S.B. (2020). Covid-19′s impact on supply chain decisions: Strategic insights
from NASDAQ 100 firms using Twitter data. Journal of Business Research 117, 443–449,
Singh, S., Kumar, R., Panchal, S. & Tiwari, M.K. (2020). Impact of COVID-19 on logistics systems and
disruptions in food supply chain, International Journal of Production Research,
Vidya, C.T. & Prabheesh, K.P. (2020). Implications of COVID-19 Pandemic on the Global Trade Networks.
Emerging Markets Finance and Trade. 56(10). 2409-2413-2414, 2020
Yavuz, E., Açıkgöz, F. & Maşe, İ. (2020). Selçuk Bayraktar yerli solunum cihazı üretim sürecini anlattı.
https://www.hurriyet.com.tr/gundem/selcuk-bayraktar-yerli-solunum-cihazi-uretim-surecini-anlatti-41499156
(Accessed Date: 05.12.2020).

142
 A SCOPING REVIEW OF THE LOT SIZING PROBLEM WITH
MULTIPLE ITEMS, MULTIPLE PERIODS AND MULTIPLE SUPPLIERS
Matheus Allgaier1, Luiz Felipe Scavarda2, Andréa Regina Nunes de Carvalho3, Fernanda Ruffo
Tercero4

Abstract- Since its first appearance in the literature, the lot sizing problem has been extensively studied and
has undergone several changes. However, most articles still handle the lot sizing decisions with their
simplest settings, not covering scenarios with multiple items, multiple supplier and multiple periods, despite
this being a common setting in today’s enterprises. Therefore, this paper presents a scoping review of
articles that address the lot sizing problem while considering multiple items, multiple suppliers, and
multiple periods, with the purpose of organizing and integrating the fragmented findings, while providing
a taxonomy of the most common and relevant attributes and proposing a research agenda. The 45 articles
retrieved for the study were analyzed and a taxonomy of the 23 most frequent or relevant attributes was
proposed. This taxonomy groups the attributes into four categories (costs, discounts, capacity restriction,
and other incentives and restrictions), and may lay a foundation for further explorations. This paper also
provides insights of the existing gaps in the literature and poses questions for future research. Practitioners
can take stock of the findings of this paper with a better knowledge of how those attributes can influence
the lot sizing problem in their real-life settings.

Keywords: Economic Order Quantity, Inventory management, Order allocation, Procurement planning,
Production management, Supplier selection

INTRODUCTION
The lot sizing problem was presented by Harris (1913) in his article "How many parts to make at once".
Its objective is to determine the quantity to be produced (EPQ - Economic Production Quantity) or to be
purchased (EOQ - Economic Order Quantity) of a given product that will meet a given demand, ensuring
the lowest total cost. According to Alfares and Turnadi (2018), in its simplest form, the EOQ is presented
with a single material, a single supplier, a single period, the demand is constant, and no other restrictions or
incentives are considered.
Since its first appearance, the lot sizing problem has been extensively studied and has undergone
several changes. However, it is still possible to verify the main trends and academic gaps (Glock et al.,
2014). According to a recent analysis of existing literature reviews, few studies are related to problems that
have, simultaneously, multiple items, multiple suppliers, and multiple periods (Aissaoui et al., 2007; Glock
et al., 2014).
This configuration, although little addressed, is the one that best reflects the current reality of many
companies. The multi-item restriction is necessary due the increasing number of SKU (Stock Keeping Units)
in companies and distribution centers (Gagliardi et al., 2008). The multi-supplier restriction is based on
companies' growing dependence on their external supply chain (Ploetner and Ehret, 2006) and facilitated
access to suppliers worldwide. Finally, the multi-period restriction stems from situations where demand has
predictable variations (e.g. seasonality) (Andrade et al., 2021) or when there may be a the possible cost
reduction on the planning horizon (Aissaoui et al., 2007), by choosing, for example, not to buy a product in
a given period to save on inventory costs and get better prices through discounts for larger quantities.
The purpose of this paper is to organize and integrate the fragmented findings on the lot sizing problem
with concurrent constraints of multiple suppliers, products and periods, through a scoping review, while
providing a taxonomy of the most common attributes and a research agenda.

1
Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio), Rio de Janeiro, RJ, 22453-900, Brazil,
[email protected]
2
Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio), Rio de Janeiro, RJ, 22453-900, Brazil, [email protected]
3
Instituto Nacional de Tecnologia (INT), Rio de Janeiro, RJ, 20081-312, Brazil - [email protected]
4
Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio), Rio de Janeiro, RJ, 22453-900, Brazil,
[email protected]

143
 The rest of the article is organized as follows: Section 2 presents the classical modeling of the lot sizing
problem and the three variations of the problem that are considered in this article. Section 3 offers the
scoping review methodology. Section 4 lists the main attributes found in the selected studies and proposes
a taxonomy. Section 5 presents the result of cataloging the articles according to the attributes defined in the
previous sections. Finally, Section 6 offers the conclusions and limitations of this research.

PROBLEM DEFINITION
According to Glock et al. (2014), in its simplest form, the lot sizing problem seeks to define the
economically optimal amount that a company should request from its supplier. For a given demand, the
optimal quantity will be calculated considering the ordering cost and holding cost of the materials. In this
scenario, the company has a single product to be ordered, only one supplier, and a single period. The
suppliers’ lead time is not considered and a stockout is not allowed.
Despite its simplicity, the model proposed by Harris (1913) was confirmed several times as a robust
and efficient model to avoid unnecessary accumulation of stock in companies (Glock et al., 2014). However,
the reality of companies has changed several times since its first appearance, revealing therefore the need
for more complex and specific models. Innovation, globalization of markets and more active customers
brought trends that manufacturing companies could not escape (Perona and Miragliotta, 2004). Lamba et
al. (2019) claim that globalization has increased the number of elements in a supply chain, which now have
more buyers, suppliers, and products. In addition, the products offered have smaller differences in price and
quality, increasing the number of options for the buyer companies.
The increased number of products and raw materials is already a reality in several companies. One
example refers to industrial sites that use numerous raw materials to manufacture a single final product in
their portfolio (Kirschstein and Meisel, 2019). Additionally, these materials usually share physical and
budgetary space (Franco and Alfonso-Lizarazo, 2020). Therefore, lot sizing models should consider more
realistic scenarios.
The communication technologies advance has also contributed to easier access to a larger number of
suppliers. Christopher (2016) identified some advantages in using IT as it represents a strong connection in
the supply chain and provides a window of opportunities for cost reduction. Reverse auctions, for example,
allow suppliers to compete for the best selling price for the buyer (Goossens et al.,2007). These
circumstances generate flexibility for companies, who can divide their orders among several suppliers,
reducing their risk and building a competitive environment (Jolai et al., 2011). From the moment a company
can choose more than one vendor to buy its materials, the problem of lot sizing becomes an Order Allocation
problem (Nazari-Shirkouhi, 2013).
Rezaei and Davodi (2008) explain that by defining the problem with multiple periods, the company can
choose different suppliers in each of the periods. Additionally, more cost-efficient management can be
carried out by grouping small orders into just one, saving on ordering costs, shipping costs and getting better
discounts, given a low holding cost. Aissaoui et al. (2007) highlight that multi-period problems are more
difficult and still represent the minority of the articles presented.
The first work to simultaneously consider multiple items, multiple suppliers and multiple periods was
that of Basnet and Leung (2005). Until then, the articles considered only some of these restrictions. These
authors defined the problem as follows: For a given planning horizon, the demand for each period of a given
product list is provided. Each product can be purchased from a set of suppliers, and each vendor offers a set
of products, with different prices from other suppliers. Each period, the company must define whether to
buy any products, in what quantity and from which suppliers. For each purchase order issued to a vendor,
the company pays an ordering cost, specific to each vendor. Additionally, for each product unit in inventory
at the end of the period, the company bears a holding cost, a product-dependent variable.
The authors sought to minimize the total cost of supplies of a company, composed by: (i) the purchasing
cost, given by multiplying the quantities purchased by the price per unit, (ii) the ordering cost, resulting
from the multiplication between the number of orders and the cost per transaction, and (iii) the holding cost,
given by the quantity of products in stock in each period multiplied by the storage cost of a product for a
period. The authors proposed, in addition to mathematical modeling, a heuristic to solve the problem.

144
 METHODOLOGY
This work adopted the scoping review methodology which is used to identify the key concepts and
state of the art of a given theme (Arksey and O'Malley, 2005). The scoping review is an adequate tool to
examine the extent, nature, and scope of a research activity, identify gaps in the literature and summarize
and disseminate research results (Levac et al., 2010). The methodological steps adopted within this research
followed the ones proposed by Arksey and O'Malley (2005) which include: (i) Definition of the research
question; (ii) Identification of relevant studies; (iii) Selection of studies; (iv) Mapping of the data; (v)
Grouping and summarization of results.
This work was guided by the question "What are, and what characteristics share, the lot sizing studies
that simultaneously consider multiple items, multiple suppliers and multiple periods?". This scope was
chosen because it is an adequate representation of the business reality and because it is a theme still rarely
studied (Aissaoui et al., 2007), as contextualized in the introduction of this paper. The interest in the main
shared characteristics arose from the lack of studies that cataloged them and because it is an effective way
to identify the gaps in the literature and to stimulate future research.
The identification of relevant studies occurred in two ways: Direct search in electronic databases and
search in the bibliographic references of other selected articles. The Science Direct and Google Scholar
databases were used for the initial search of the literature. Four sets of keywords were defined: {lot-sizing,
EOQ}, {multi product, multi item, multi commodity}, {multi supplier, multi vendor, "supplier selection"}
and {multi period}, so that various combinations of keywords were used in the databases. For Science
Direct, all abstracts were selected, while in Google Scholar, because of the high volume of results, the
articles were ordered by relevance and the first 100 results went through an abstract reading and pre-
selection. Afterwards, all references were saved and managed through the Zotero software. This step
resulted in 163 articles to be classified.
Since in the previous step some articles irrelevant to the proposed theme were collected, a refining and
selection step of the article set became necessary. The exclusion criteria "single product", "single supplier"
and "single period" were adopted, since these articles would not meet the research question. Thus, the
presence of these words in the title or abstract of an article indicated that it should be removed from the
study base. For the remaining articles, it was necessary to read the titles, abstracts and, in some cases, the
entire manuscript, to make the correct selection. At the end of this stage, there were only 26 studies left to
be analyzed and catalogued. Subsequently, 19 other articles were added through snow balling by reading
the bibliographic reviews of the selected articles and systematic literature reviews, especially the reviews
conducted by Aissaoui et al. (2007), Jans and Degraeve (2007), and Glock et al. (2014). As mentioned by
Mirzaee et al. (2018), the problem of lot sizing, or supplier selection, is unique for each company, which
results in a large number of variations for the classic modeling of the problem. Therefore, to properly chart
the articles, the restrictions and incentives found in these papers were defined as a cataloging parameter.
This post hoc analysis enabled the identification of 23 relevant restrictions and incentives in the literature,
that were grouped into 4 categories: costs, discounts, capacity constraints, and other incentives and
restrictions. It is noteworthy that these were not the only characteristics found. On the other hand, they were
selected because they are either frequently used in studies or cited in systemic literature reviews, or represent
innovative approaches or are commonly found in the industry.
Finally, all articles were classified according to the 23 attributes defined in the previous step and this
result is exposed next in the next section.

PROBLEM VARIATIONS CLASSIFICATION

The problem presented by Basnet and Leung (2005) has undergone several adaptations in recent years,
with different restrictions and incentives being added. Restrictions are characteristics of the problem that
limit or reduce the amount to be purchased each period, while incentives have the opposite effect.
In this paper, the most recurrent restrictions and incentives were selected and catalogued to assist in the
classification of articles. Some restrictions and incentives were later added to catalog because of their
innovative approach or because they were cited as gaps in the literature by other papers. It is noteworthy
that, although two articles consider the same restriction or incentive, they can treat it in different ways, by
developing different formulations or dependency variables. Four categories of restrictions and incentives
were defined. The first category encompasses the different costs considered in the problems. The second

145
category deals with the different discount policies offered by suppliers. The third covers the limitations
related to the capacity of the buyer, the supplier or even the means of transport. Finally, the latter category
explores other restrictions and incentives.
Costs
Supply-related costs can reach up to 70% of the total cost of finished products (Ghodsypour and
O'Brien, 2001). These costs can involve the price of raw materials, ordering, inventory, and shipping costs.
In this way, companies are obliged to minimize their costs through efficient supply chain management
(Mirzaee et al., 2019). Shipping costs should be highlighted, since they have gained relevance with increased
globalization and can change purchasing decisions, opting for suppliers with higher prices, but lower
transportation costs (Darvishi et al., 2020).
The Holding Cost (HOC), which represents the financial value to stock a unit of product for one unit of
time, stands out as the most common variable considered among the analysed articles (i.e., present in 88%
of the selected papers). When calculating HOC, the physical costs of storing the product in inventory,
depreciation and opportunity cost of invested capital are considered (Glock et al., 2014). In most cases,
HOC depends solely on the product, but, on more complex models, it may depend simultaneously on the
product, warehouse, and period (Moheb-Alizadeh and Handfield, 2019).
The Procurement Cost (PRC), also known as transaction cost or ordering cost, is the amount of money
spent to issue a purchase order, regardless of the quantity of products ordered (Basnet and Leung, 2005). It
can be presented as a specific value per supplier (Cunha et al., 2018), as a vendor and product dependent
variable (Alfares and Turnadi, 2018) or as a variable dependent simultaneously on the supplier, warehouse,
and period (Moheb-Alizadeh and Handfield, 2019).
The Fixed Transportation Cost (FTC) is the amount to be paid to the carrier regardless of the quantity
transported. Alfares and Turnadi (2018) consider a fixed cost per vehicle, but a different value for each
supplier. Entezaminia et al. (2016) adopt a cost dependent on the transport modal, supplier, and warehouse.
The Variable Transportation Cost (VTC) is the amount to be paid for each unit of product transported.
Thanh et al. (2008) consider a variable cost for product, supplier, and warehouse. Megahed and
Goetschalckx (2019) and Salema et al. (2010) present it as a variable dependent on the product, supplier,
warehouse, and period.
The Shortage Cost (SHC) occurs when there is no inventory to meet a given demand. This event may
result in a discount being offered to the customer when this demand is met or may result in a missed sale,
therefore an opportunity cost. Franco and Alfonso-Lizarazo (2020) present this cost as an emergency
purchase, for a higher price, to avoid the lack of stock. Megahed and Goetschalckx (2019) apply a penalty
for each missing unit, without distinguishing products or periods. Nourmohamadi Shalke et al. (2018)
consider a cost of lack of stock for each product. Mirzaee et al. (2019) propose a distinct cost for each
product and period. Moheb-Alizadeh and Handfield (2019) simultaneously consider product, period, and
customer in defining the cost of lack of inventory.
Discounts
Discounts have long been considered in lot sizing problems (Kirschstein and Meisel, 2019). There are
several reasons why a supplier can offer discounts. For Glock et al. (2014), discounts encourage the buyer
to place larger orders more often. By stimulating larger orders, suppliers also reduce their cost of stocking
finished products (Cunha et al., 2018).
In Incremental Quantity Discount (IQD), the vendor defines several price levels based on the quantity
requested for a single product. Thus, the corresponding price will be applied for the units within each level
(Mirzaee et al., 2019).
In the All Units Quantity Discount (AUD), the vendor also defines several price levels. However, when
one reaches a certain level, that price will be applied to all units of that product.
The Total Quantity Discount (TQD) considers the total units requested of all products in one order
(Kirschstein and Meisel, 2019) and all products will benefit from the same percentual discount.
In the discount by financial volume (BVD - Business Volume Discount) the supplier will apply a
percentage discount on the total value of the order after it reaches a certain level. In this way, all products
will suffer the same percentage reduction in price (Razmi and Maghool, 2010).

146
 Capacity constraints
Capacity constraints may occur because of the production process, supplier's packaging size, transport
limitations, warehouse capacity or even by the budgetary limitation of the buyer.
The Minimum Order Quantity (MIQ) is a vendor related restriction, which requires that the requested
quantity exceeds a certain quota. This requirement may arise to enable an economy of scale for the supplier
(Gorji et al., 2014). For example, in industrial suppliers, certain machines are only economically operable
with higher production quantities, given their high setup time and cost. Additionally, the MIQ may be related
to the cost and capacity of transportation or packaging. Gorji et al. (2014) and Mota et al. (2018) use a
minimum order quantity for each product and supplier. Megahed and Goetschalckx (2019) consider a
minimum monthly order from each supplier, making it impossible for a supplier not to be considered in
certain periods.
The Maximum Order Quantity (MAQ) is also a supplier related restriction, usually limited to its
production capacity or when the buyer wants to avoid supply dependency. It is the most common type of
capacity constraint in the articles considered in this study. Govindan et al. (2017) and Salema et al. (2010)
treat MAQ as a supplier dependent variable. Thanh et al. (2008) and Stecca et al. (2016) present MAQ as a
supplier, product and period dependent variable.
Standard Lot Size (SLZ) occurs when the vendor requires the requested quantity to be a multiple of its
standard batch. This can occur because of the production system or the size of the packages. Franco and
Alfonso-Lizarazo (2020) and Stecca et al. (2016) formulate the problem of hospitals that buy medicines in
boxes and must distribute them in units to their patients, therefore, SLZ becomes a product dependent
variable.
The Transport Capacity (TPC) restriction considers the maximum number of units to be loaded by the
transport mode used in the problem. This restraint can consider only the number of units, as well as the
dimensions and weight of the loaded products, or their packaging. In companies that use transport service
providers, capacity may decrease in periods that the provider already has demand from other clients. Lamba
et al. (2019) consider a fixed capacity for the transportation mode of the problem. Mota et al. (2018) study
a specific capacity for each transport mode. Moheb-Alizadeh and Handfield (2019) consider transport
capacity dependent simultaneously on product, supplier, means of transport, warehouse, and period.
Warehouse Capacity (WHC) limits the amount of products the buyers’ company can keep at any time
due to the physical limitations of its warehouses. Similar to the TPC, it can be measured by units, or consider
the dimensions and specifications of each product, or its packaging. For companies that hire storage service
providers, the capacity of each warehouse may vary by period (Peidro et al., 2009). There are also scenarios
where capacity is defined by material and period (Mohammadi et al., 2020). When the problem has multiple
warehouses, the storage capacity of each warehouse should be considered (Balcázar-Camacho, 2016).
Finally, the budget limitation (BUD) establishes that the financial volume of all purchase orders for a
given period cannot exceed a limit established by the company (Razmi and Maghool, 2010).
Other incentives and restrictions
This category includes restrictions and incentives that do not fall into the other categories. Here, the
authors took the liberty of including attributes that, although uncommon within the studies, add complexity
to the problem, such as backordering, perishability and payment methods. Additionally, attributes that
merge the problem of order allocation with other problems have been added, such as multiple warehouses
(location of facilities), multiple means of transport (modal selection), and the production system (EPQ -
Economic Production Quantity).
Lead Time (LTI) is the number of units of time for the supplier to deliver the requested products to the
buyer company. It increases the complexity of lot sizing problems because it requires a certain prediction
of how much will be consumed over the replenishment cycle. This attribute becomes even more relevant
for companies that have global suppliers (Darvishi et al., 2020). Authors who do not consider this attribute
justify that the delivery time is less than one day, therefore, can be neglected when modeling the problem
(Basnet and Leung, 2005). Alfonso-Lizarazo (2019) and Cunha et al. (2018) use LTI as a fixed value for
the whole model. Kirschstein and Meisel (2019) studied LTI as a supplier and warehouse dependent
variable.
The Backordering (BKO) occurs when there is a lack of inventory, also known as shortage, and the
customer agrees to wait until the resupply period (Glock et al., 2014). In this case, the company is left with

147
an unfulfilled demand that should be added to its regular demand. Backordering, besides generating a
shortage cost, can aggregate daily costs by delivery delay.
The Defective Goods Rate (DGR) states that every batch of supplier product may have some defective
products, that cannot be consumed. In this case, defective products can be returned, discarded, or even sold
at a reduced price (Rezaei and Davodi, 2008).
The Perishability of the items (PER) implies that, after a certain time, the product will deteriorate and
can no longer be used, sold, or has its value drastically reduced. Franco and Alfonso-Lizarazo (2019)
modeled a problem that considered the perishability of hospital products. Other sectors that may suffer this
type of problem are food, fashion, technology, and periodical magazines.
Payment methods (PME) consist of the different forms of down and installment payment, including
differing interest rates, that the buyer can use to pay the vendor. According to Razmi and Maghool (2010),
this encourages the supplier to place larger orders, as it will not be fully restricted by the period budget and
can achieve better discount levels.
Finally, the authors considered the models that contain Multiple Warehouses (MWH), Multiple
Transport Modes (MTM) and those that jointly solve Production Decisions (PRD). In the problems of
multiple warehouses, the buyer company has access to several storage facilities, owned or rented, where
each will have different capacities, delivery times, storage costs, fixed and variable transportation costs and
even different order costs. In the problems of multiple transport modes, the buyer has access to different
means of transport, which may be insourced, owned by the supplier or outsourced with third party service
providers. Each means of transport will have a different capacity, fixed and variable transportation cost, and
lead times. On the other hand, the problems integrated with the production seek to optimize the total cost of
the company, when planning production according to the best discounts obtained from suppliers (Cunha et
al., 2018).

RESULTS AND DISCUSSION

The result of the classification of the 45 articles into 23 attributes is shown in Table 1. It can be noted
that there are still few articles that simultaneously meet the configurations of multiple products, multiple
suppliers, and multiple periods. It is also noteworthy that, although there are more common attributes among
the works, such as HOC (Holding Cost), MAQ (Maximum Order Quantity) and LTI (Lead Time), these are
not mandatory components in the formulation of the problem.
The analysis in Table 1 reveals that there are still many gaps and opportunities among the attributes
selected for this review. The standard lot size (SLZ) was only analyzed in two articles, both of which deal
with supplies of goods in the healthcare area. The shortage cost (SHC) and backordering (BKO), although
conceptually connected, still appear separately in some articles. Perishability (PER) and payment methods
(PME) were each analyzed by only one article, and both neglected the main cost variables, showing that
there is still room for development in this area. Discount policies, although popular in simpler versions of
the lot sizing models, are still scarce in these articles. It is also noteworthy that only 4 articles dealt with
discount policies that consider multiple products in an integrated way (TQD and BVD). Lead Time (LTI)
is present in most real-life settings, but it’s still underexplored (40%). No paper simultaneously considered
all costs attributes.
Multiple transport modes (MTM) problems could be better explored if attributes like minimum
quantity (MIQ), standard lot sizes (SLZ) and budget limitation (BUD) were added, as those could shift or
limit the decisions. It should be noted that, although LTI is a decision criterion for modal selection, it was
considered only in 66% of the articles.
There is a lack of studies of production decisions (PRD) and restraints like standard lot sizes (SLZ),
budget limitation (BUD), defective good rates (DGR) and perishability (PER). Those should be jointly
addressed, because these attributes will create restrictions in the raw materials supply and add real-life
complexity to the problems.

148
 Table 1. Articles Classification

Capacity Other Incentives and

Costs Discounts
Constraints Restrictions
Authors H P F V S I A T B M M S T W B L B D P P M M P
O R T T H Q U Q V I A L P H U T K G E M W T R
C C C C C D D D D Q Q Z C C D I O R R E H M D
Ahmad and Modal (2016) X X X X X
Alfares and Turnadi (2018) X X X X X X X X X X X X
Badri et al. (2013) X X X X X X X
Balcázar-Camacho et al. (2016) X X X X X X X X
Basnet and Leung (2005) X X
Cárdenas-Barron et al. (2015) X X
Cheraghalipour and Farsad
X X X X X X X X
(2018)
Cunha et al. (2018) X X X X X X X
Darvishi et al. (2020) X X X X X X X X X
Díaz-Madroñero et al. (2017) X X X X X X X X X
El Beihery et al. (2019) X X X X X
Entezaminia et al. (2016) X X X X X X X X X X
Franco and Alfonso-Lizarazo
X X X X X
(2020)
Gholizadeh et al. (2020) X X X X X X X X X
Govindan et al. (2017) X X X X X X X X X
Goossens et al. (2007) X X
Gorji et al. (2014) X X X X X X X X X
Jolai et al. (2011) X X X X X X
Kaur and Singh (2016) X X X X X X X
Kaur and Singh (2017) X X X X X X
Kaur and Singh (2018) X X X X X X X
Kaur and Singh (2021) X X X X X
Kirschstein and Meisel (2019) X X X X X X X X X X
Lamba and Singh (2019) X X X X X
Lamba et al. (2019) X X X X X
Lu et al. (2020) X X X X X
Megahed and Goetschalckx
X X X X X X X
(2019)
Mirzaee et al. (2018) X X X X X X X X X
Mohammadi et al. (2020) X X X X X X X X X
Moheb-Alizadeh and Handfield
X X X X X X X X X X X X
(2019)
Mota et al. (2018) X X X X X X X X X X
Nourmohamadi Shalke et al.
X X X X X X X X
(2018)
Peidro et al. (2009) X X X X X X X X X
Razmi and Maghool (2010) X X X X X X X
Rezaei and Davoodi (2008) X X X X X
Sadeghi Moghadam et al. (2008) X X

149
 Salema et al. (2010) X X X X X X X X X X
Stecca et al. (2016) X X X X X
Thanh et al. (2008) X X X X X X X
Tosarkani and Amin (2018) X X X X
Ware et al. (2014) X X X X
Woarawichai and Naenna (2017) X X X X X
Woarawichai and Naenna (2018) X X X X X
Yaghin et al. (2020) X X X X X X X X X
Zeballos et al. (2014) X X X X X X X X X X X

CONCLUSION
In this study, the authors did a scoping review of the lot sizing problems, that considered,
simultaneously, multiple products, multiple suppliers and multiple periods. The articles in this configuration
of the problem are still scarce, despite the practical importance of the theme, and to our knowledge, this is
the literature review that brought the largest number of articles on the subject, being this the main
contribution of the article. The result of this study allows researchers and practitioners to find gaps to be
studied and solved. It also serves as a repository to help address real life scenarios, speeding the process of
finding similar settings and attributes.
A taxonomy was developed, providing a detailed explanation of several aspects of the problems,
allowing easy access to articles of interest and a clear identification of literature gaps. Those attributes were
also associated with real life situations, in order to raise scenarios where they could be used. The attributes
were separated in 4 groups: Costs, Discounts, Capacity restrictions, and other incentives and restrictions.
At least in the references consulted in this study, there was no cataloging with so many attributes.
For future research, some attributes could be better explored, like perishability of goods and payment
methods, as the literature is still scarce. New combinations of attributes could also be presented, as stated
in this paper’s discussion, e.g. DGR and discounts were jointly covered by only one paper. Additionally,
most attributes should be studied in new settings, for example, PER and SLZ are still limited to healthcare
scenarios, although they are common in several sectors, like food and apparel.
There are several attributes that were not highlighted in the study, given the immensity of variations of
the original problem, and a suggestion for future research is to expand, improve and exemplify the list
proposed here. Another future research topic is to refine the taxonomy by adding the probabilistic behavior
of variables, e.g. demand and lead times could be deterministic or stochastic, and how they behave time-
wisely, e.g. perishability and holding costs can be linear or exponential. This will provide better search
parameters for practitioners that want to implement those models in real life scenarios. The taxonomy could
be further improved with the addition of new dimensions, such as number of echelons, performance and
green criteria for supplier selection, among others.
Several articles solved the lot sizing problem through mathematical modelling. We encourage
researchers and practitioners to make use of heuristics and other techniques to address and further optimize
the solutions methods. Practitioners can take stock of the findings of this paper with a better knowledge of
how the discussed attributes can influence their lot sizing problem in real life settings.
The limitations of the work are the methodology used and the databases consulted. There is room for
better exploration with snowball search (back and forth) from all the articles mentioned here. The research
terms used in this work can be used in other databases to expand the set of articles to be studied.

REFERENCES
Ahmad, M. T., and S. Mondal. 2016. “Dynamic Supplier Selection Model under Two-Echelon Supply Network.”
Expert Systems with Applications 65:255–70.
Aissaoui, Najla, Mohamed Haouari, and Elkafi Hassini. 2007. “Supplier Selection and Order Lot Sizing Modeling: A
Review.” Operations Research and Outsourcing 34(12):3516–40.
Alfares, Hesham K., and Rio Turnadi. 2018. “Lot Sizing and Supplier Selection with Multiple Items, Multiple Periods,
Quantity Discounts, and Backordering.” Computers & Industrial Engineering 116:59–71.

150
Andrade, X., Guimarães, L., & Figueira, G. 2021. “Product line selection of fast-moving consumer goods”. Omega,
102, 102389.
Arksey, H., & O'Malley, L. 2005. “Scoping studies: towards a methodological framework”. International journal of
social research methodology, 8(1), 19-32.
Badri, Hossein, Mahdi Bashiri, and Taha Hossein Hejazi. 2013. “Integrated Strategic and Tactical Planning in a Supply
Chain Network Design with a Heuristic Solution Method.” Computers & Operations Research 40(4):1143–54.
Balcázar-Camacho, Delio Alexander, Cesar Amílcar López-Bello, and Wilson Adarme-Jaimes. 2016. “Strategic
Guidelines for Supply Chain Coordination in Healthcare and a Mathematical Model as a Proposed Mechanism for the
Measurement of Coordination Effects.” Dyna 83(197):203–11.
Basnet, Chuda, and Janny M. Y. Leung. 2005. “Inventory Lot-Sizing with Supplier Selection.” Computers &
Operations Research 32(1):1–14.
Cárdenas-Barrón, Leopoldo Eduardo, José Luis González-Velarde, and Gerardo Treviño-Garza. 2015. “A New
Approach to Solve the Multi-Product Multi-Period Inventory Lot Sizing with Supplier Selection Problem.” Computers
& Operations Research 64:225–32.
Cheraghalipour, Armin, and Saba Farsad. 2018. “A Bi-Objective Sustainable Supplier Selection and Order Allocation
Considering Quantity Discounts under Disruption Risks: A Case Study in Plastic Industry.” Computers & Industrial
Engineering 118:237–50.
Christopher, M. 2016. Logistics & supply chain management. Pearson Uk.
Cunha, Artur Lovato, Maristela Oliveira Santos, Reinaldo Morabito, and Ana Barbosa-Póvoa. 2018. “An Integrated
Approach for Production Lot Sizing and Raw Material Purchasing.” European Journal of Operational Research
269(3):923–38.
Darvishi, Fateme, R. Ghasemy Yaghin, and Abdolhossein Sadeghi. 2020. “Integrated Fabric Procurement and Multi-
Site Apparel Production Planning with Cross-Docking: A Hybrid Fuzzy-Robust Stochastic Programming Approach.”
Applied Soft Computing 92:106267.
Díaz-Madroñero, Manuel, Josefa Mula, and David Peidro. 2017. “A Mathematical Programming Model for Integrating
Production and Procurement Transport Decisions.” Applied Mathematical Modelling 52:527–43.
El Beheiry, Mohammed M. and others. 2019. “Joint Lot Sizing and Scheduling of a Multi-Product Multi-Period Supply
Chain.” Independent Journal of Management & Production 10(5):1516–35.
Entezaminia, Arezoo, Mahdi Heydari, and Donya Rahmani. 2016. “A Multi-Objective Model for Multi-Product Multi-
Site Aggregate Production Planning in a Green Supply Chain: Considering Collection and Recycling Centers.” Journal
of Manufacturing Systems 40:63–75.
Franco, Carlos, and Edgar Alfonso-Lizarazo. 2020. “Optimization under Uncertainty of the Pharmaceutical Supply
Chain in Hospitals.” Computers & Chemical Engineering 135:106689.
Ghodsypour, S. H., & O’brien, C. 2001. “The total cost of logistics in supplier selection, under conditions of multiple
sourcing, multiple criteria and capacity constraint”. International journal of production economics, 73(1), 15-27.
Gholizadeh, Hadi, Hamed Fazlollahtabar, and Mohammad Khalilzadeh. 2020. “A Robust Fuzzy Stochastic
Programming for Sustainable Procurement and Logistics under Hybrid Uncertainty Using Big Data.” Journal of
Cleaner Production 258:120640.
Glock, Christoph H., Eric H. Grosse, and Jörg M. Ries. 2014. “The Lot Sizing Problem: A Tertiary Study.”
International Journal of Production Economics 155:39–51.
Goossens, D. R., A. J. T. Maas, F. C. R. Spieksma, and J. J. van de Klundert. 2007. “Exact Algorithms for Procurement
Problems under a Total Quantity Discount Structure.” European Journal of Operational Research 178(2):603–26.
Gorji, Mohammad Hossein, Mostafa Setak, and Hossein Karimi. 2014. “Optimizing Inventory Decisions in a Two-
Level Supply Chain with Order Quantity Constraints.” Applied Mathematical Modelling 38(3):814–27.
Govindan, Kannan, Jyoti Dhingra Darbari, Vernika Agarwal, and P. C. Jha. 2017. “Fuzzy Multi-Objective Approach
for Optimal Selection of Suppliers and Transportation Decisions in an Eco-Efficient Closed Loop Supply Chain
Network.” Journal of Cleaner Production 165:1598–1619.
Jans, Raf, and Zeger Degraeve. 2007. “Meta-Heuristics for Dynamic Lot Sizing: A Review and Comparison of
Solution Approaches.” European Journal of Operational Research 177(3):1855–75.
Jolai, Fariborz, Seyed Ahmad Yazdian, Kamran Shahanaghi, and Mohammad Azari Khojasteh. 2011. “Integrating
Fuzzy TOPSIS and Multi-Period Goal Programming for Purchasing Multiple Products from Multiple Suppliers.”
Journal of Purchasing and Supply Management 17(1):42–53.
Kaur, Harpreet, and Surya Prakash Singh. 2016. “Modeling Flexible Procurement Problem.” Pp. 147–70 in Managing
Flexibility: People, Process, Technology and Business, edited by Sushil, K. T. Bhal, and S. P. Singh. New Delhi:
Springer India.
Kaur, Harpreet, and Surya Prakash Singh. 2017. “Modeling Low Carbon Procurement and Logistics in Supply Chain:
A Key towards Sustainable Production.” Sustainable Production and Consumption 11:5–17.
Kaur, Harpreet, and Surya Prakash Singh. 2018. “Heuristic Modeling for Sustainable Procurement and Logistics in a
Supply Chain Using Big Data.” Computers & Operations Research 98:301–21.

151
Kaur, Harpreet, and Surya Prakash Singh. 2021. “Multi-Stage Hybrid Model for Supplier Selection and Order
Allocation Considering Disruption Risks and Disruptive Technologies.” International Journal of Production
Economics 231:107830.
Kirschstein, Thomas, and Frank Meisel. 2019. “A Multi-Period Multi-Commodity Lot-Sizing Problem with Supplier
Selection, Storage Selection and Discounts for the Process Industry.” European Journal of Operational Research
279(2):393–406.
Lamba, Kuldeep, and Surya Prakash Singh. 2019. “Dynamic Supplier Selection and Lot-Sizing Problem Considering
Carbon Emissions in a Big Data Environment.” Technological Forecasting and Social Change 144:573–84.
Lamba, Kuldeep, Surya Prakash Singh, and Nishikant Mishra. 2019. “Integrated Decisions for Supplier Selection and
Lot-Sizing Considering Different Carbon Emission Regulations in Big Data Environment.” Computers & Industrial
Engineering 128:1052–62.
Levac, D., Colquhoun, H., & O'Brien, K. K. 2010. “Scoping studies: advancing the methodology”. Implementation
science, 5(1), 1-9.
Lu, Shan, Li Zhu, Yue Wang, Lei Xie, and Hongye Su. 2020. “Integrated Forward and Reverse Logistics Network
Design for a Hybrid Assembly-Recycling System under Uncertain Return and Waste Flows: A Fuzzy Multi-Objective
Programming.” Journal of Cleaner Production 243:118591.
Megahed, Aly, and Marc Goetschalckx. 2019. “A Modeling Framework and Local Search Solution Methodology for
a Production-Distribution Problem with Supplier Selection and Time-Aggregated Quantity Discounts.” Applied
Mathematical Modelling 68:198–218.
Mirzaee, Hossein, B. Naderi, and S. H. R. Pasandideh. 2018. “A Preemptive Fuzzy Goal Programming Model for
Generalized Supplier Selection and Order Allocation with Incremental Discount.” Computers & Industrial
Engineering 122:292–302.
Mohammadi, Milad, Majid Esmaelian, and Arezoo Atighehchian. 2020. “Design of Mathematical Models for the
Integration of Purchase and Production Lot-Sizing and Scheduling Problems under Demand Uncertainty.” Applied
Mathematical Modelling 84:1–18.
Moheb-Alizadeh, Hadi, and Robert Handfield. 2019. “Sustainable Supplier Selection and Order Allocation: A Novel
Multi-Objective Programming Model with a Hybrid Solution Approach.” Computers & Industrial Engineering
129:192–209.
Mota, Bruna, Maria Isabel Gomes, Ana Carvalho, and Ana Paula Barbosa-Povoa. 2018. “Sustainable Supply Chains:
An Integrated Modeling Approach under Uncertainty.” Omega 77:32–57.
Nazari-Shirkouhi, S., Shakouri, H., Javadi, B., & Keramati, A. 2013. “Supplier selection and order allocation problem
using a two-phase fuzzy multi-objective linear programming”. Applied Mathematical Modelling, 37(22), 9308-9323.
Nourmohamadi Shalke, Parvin, Mohammad Mahdi Paydar, and Mostafa Hajiaghaei-Keshteli. 2018. “Sustainable
Supplier Selection and Order Allocation through Quantity Discounts.” International Journal of Management Science
and Engineering Management 13(1):20–32.
Peidro, David, Josefa Mula, Raúl Poler, and José-Luis Verdegay. 2009. “Fuzzy Optimization for Supply Chain
Planning under Supply, Demand and Process Uncertainties.” Fuzzy sets and systems 160(18):2640–57.
Perona, M., & Miragliotta, G. 2004. “Complexity management and supply chain performance assessment. A field
study and a conceptual framework”. International Journal of Production Economics, 90(1), 103-115.
Razmi, Jafar, and Elham Maghool. 2010. “Multi-Item Supplier Selection and Lot-Sizing Planning under Multiple Price
Discounts Using Augmented ε-Constraint and Tchebycheff Method.” The International Journal of Advanced
Manufacturing Technology 49(1):379–92.
Rezaei, Jafar, and Mansoor Davoodi. 2008. “A Deterministic, Multi-Item Inventory Model with Supplier Selection
and Imperfect Quality.” Applied Mathematical Modelling 32(10):2106–16.
Sadeghi Moghadam, Mohammad Reza, Amir Afsar, and Babak Sohrabi. 2008. “Inventory Lot-Sizing with Supplier
Selection Using Hybrid Intelligent Algorithm.” Soft Computing for Dynamic Data Mining 8(4):1523–29.
Salema, Maria Isabel Gomes, Ana Paula Barbosa-Povoa, and Augusto Q. Novais. 2010. “Simultaneous Design and
Planning of Supply Chains with Reverse Flows: A Generic Modelling Framework.” European Journal of Operational
Research 203(2):336–49.
Stecca, Giuseppe, Ilaria Baffo, and Toshiya Kaihara. 2016. “Design and Operation of Strategic Inventory Control
System for Drug Delivery in Healthcare Industry.” 8th IFAC Conference on Manufacturing Modelling, Management
and Control MIM 2016 49(12):904–9.
Thanh, Phuong Nga, Nathalie Bostel, and Olivier Péton. 2008. “A Dynamic Model for Facility Location in the Design
of Complex Supply Chains.” International journal of production economics 113(2):678–93.
Tosarkani, Babak Mohamadpour, and Saman Hassanzadeh Amin. 2018. “A Possibilistic Solution to Configure a
Battery Closed-Loop Supply Chain: Multi-Objective Approach.” Expert Systems with Applications 92:12–26.
Ware, Nilesh R., S. P. Singh, and D. K. Banwet. 2014. “A Mixed-Integer Non-Linear Program to Model Dynamic
Supplier Selection Problem.” Expert Systems with Applications 41(2):671–78.

152
Woarawichai, Chirawat, and Thanakorn Naenna. 2018. “Solving Inventory Lot-Sizing with Supplier Selection under
Alternative Quantity Discounts and Vehicle Capacity.” International Journal of Logistics Systems and Management
30(2):179–94.
Woarawichai, Chirawat, and Thanakorn Naenna. 2017. “Multi-Product and Multi-Period Inventory Lot-Sizing with
Supplier Selection under Quantity Discount.” International Journal of Services and Operations Management
28(2):264–77.
Yaghin, R. Ghasemy, P. Sarlak, and A. A. Ghareaghaji. 2020. “Robust Master Planning of a Socially Responsible
Supply Chain under Fuzzy-Stochastic Uncertainty (A Case Study of Clothing Industry).” Engineering Applications of
Artificial Intelligence 94:103715.
Zeballos, Luis J., Carlos A. Méndez, Ana P. Barbosa-Povoa, and Augusto Q. Novais. 2014. “Multi-Period Design and
Planning of Closed-Loop Supply Chains with Uncertain Supply and Demand.” Computers & Chemical Engineering
66:151–64.

153
 EXPLORING THE SUSTAINABLE SUPPLY CHAIN INTEGRATION: A
CASE STUDY OF DOGUS GROUP
Yasemin Ülker1, Banu Demirel2

Abstract ⎯ Limited studies focus on the linkage of a company's sustainability strategy and its integration with
its supply chain management. Sustainable supply chain integration aims to achieve sustainable flow of
products, services, information, and money to maximize value for all corporate stakeholders. A well-
integrated supply chain should improve both customer service and company success. Coupled with the
environmental and social facets, supply chain also ensures the societal welfare. The main purpose of the study
is to understand how the company integrates sustainability perspectives and practices to the overall supply
chain and to uncover the challenges. The majority of the supply chain integration studies focus on
manufacturing companies. There have been much fewer studies focused on service companies. This study
intends to fill this gap by investigating the sustainable integration into a group company that has
manufacturing and service together. Considering that Dogus Group is a prominent company that considers
sustainability as a core value in Turkey, a case study is conducted to have a better insight of the sustainable
supply chain integration. Content analysis conducted on data collected indicate the key aspects that support
and hinder supply chain sustainability integration. Thus, the findings of the study are expected to pave the
way for the integration in small and medium scaled businesses in Turkey by providing a road map.

Keywords ⎯ case study, supply chain integration, sustainable supply chain

INTRODUCTION
Sustainability studies put forward that contemporary irresponsible consumption and production patterns
have to be reconsidered for the well-being of the future generations. The jeopardy in the current economic
model embraced globally is unfolded as climate change, inequality of income distribution, poverty and
many other ways that endanger the future welfare of the planet. Energy, water and many other natural
resources are diminishing Besides, most of the countries suffer the consequences of climate crisis with
extreme weather condition, droughts or sudden floods with devastating impacts for the eco systems and
economies. Sustainability is the key word for the prosperity of the future of the planet earth which is also
known as ‘triple bottom line (3BL) in the literature (Brockhaus et al, 2013) to highlight the social,
environmental and economic pillars. Sustainability is one of the crucial features of Corporate Social
Responsibility (CSR) which takes into consideration social aspects while decision making in management
(Mosteanu, 2020). Many companies are restructuring their processes to comply with the sustainable
development goals introduced by UN. Governments, NGOs and many other communities enforce and set
barriers to accelerate the sustainability practices of businesses. Businesses, whose main purpose was to
make profit and grow, now need to set not only economic but also environmental and social targets in
accordance with the sustainable development goals throughout their supply chains by integrating their
sustainability philosophy to the whole supply chain.
The main purpose of the study is to understand how the company integrates sustainability perspectives
and practices to the overall supply chain and to reveal the challenges for implementation. The following
research questions are aimed to be answered via a single case study methodology.
RQ1: How does company perceive sustainability?
RQ2: What kind of practices company does for sustainability?
RQ3: What are the efforts they put in order to integrate sustainability into the supply chain?
RQ4: Do they spread the sustainability perception to all stakeholders?
RQ5: What are the challenges for implementation?

1
Yasemin Ülker, Bahçeşehir University, Faculty of Economics, Administrative and Social Sciences, Department of Logistics
Management, İstanbul, Turkey, [email protected]
2
Banu Demirel, Dokuz Eylul University, Faculty of Business, Department of Business Administration, İzmir, Turkey,
[email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

154
The layout of this article is as follows. Initially, there is a literature review on sustainability, sustainable
supply chain management and its integration. After explaining the research and data gathering process, a
brief overview of the Doguş Group as a case study's subject is provided in the research methodology
section. Findings part illustrates the tables of the codes and categorization which are created via
MAXQDA software program. Finally, brief summary is presented, followed by the conclusion section.

LITERATURE REVIEW
SUSTAINABLE SUPPLY CHAIN MANAGEMENT (SSCM)
Vice President at the Boston Consulting Group, noted that, ‘As the economy changes, as competition
becomes more global, it’s no longer company vs company but supply chain vs supply chain’ (Henkoff,
1994). SCM is ‘collaborative management across organizational boundaries of seamless value-added
processes designed to meet the needs of the supply chain’s end customer’ (Fawcett et al, 2006). In simple
explanation, supply chain management aims to boost efficiency by integrating all aspects of corporate
processes (Bowersox et al, 2002). ‘All value-creating activities work together to offer the highest level of
client value,’in well-conducted supply chains’ (Pagell, 2004). The duty of Supply chain management is
not only to handle the products or materials, money and information flow; collaborating of the partners
and specifying the targets are also their tasks. Adding sustainability to the supply chain means thinking the
whole chain's social, environmental, and economic aspects. Seuring and Lüller (2018) stated that
environmental and social requirements must be completed by members of sustainable supply chains in
order for them to remain in the supply chain. Customers' expectations should be combined with economic
sustainability criteria to enhance the competitiveness.
Sustainable Supply Chain Management (SSCM) refers to the vital, clear unification, and accomplishment
of the economic, social, and environmental objectives of an institution through methodical designation of
strategic business procedures within the organisation to improve the overall commercial efficiency in
respect of individuals and their supply chain (Carter & Rogers, 2008). SSCM also connotes the array of
management methods including ‘effect on the environment’ as a crucial recognition for all phases within
each product’s value chain as well as an interdisciplinary outlook that include the product life-cycle in
entirety (Laurin & Fantazy, 2017).
There are substantial growing concerns of sustainability resulting from the increased knowledge of
partners within the global supply chains (Wiese et al, 2015).
Just as businesses are for-profit organizations and their main purpose is to survive, the economic purpose
of the supply chain is to support this. therefore, while the economic aspect of sustainable supply chain
management remains the same, 2 more dimensions have been added to it: social and environmental. Zhou
et al. (2000) has stated that for social sustainability, it is necessary to produce according to the needs of
the society. environmental sustainability is possible by saving on the use of resources and reducing waste.
Mollenkopf et al. (2010) opines that, initiatives for the environment; boost adeptness and product output,
help in the reduction of negative exposures and costs in results in increased business profits. The
environmental initiatives encompass; minimising carbon emissions which accounts for energy efficiency,
reduced packaging, and utilisation of renewable resources. Others are; equality, increasing employee
participation, diversity within the workplace, and the practice of corporate social responsibility especially
providing for the community or society the organisation is located. Increasing these drives
Gopalakrishnan et al. (2012) have listed the drivers of sustainability as; stakeholder concerns, government
regulation, depletion of natural resources, economy with a low carbon footprint, social responsibility and
environmental requirements. The key forces that were found to be behind acceptance and usage of
sustainability strategies for organisations are mostly government and non-governmental organisations
(Gonza´lez-Benito et al, 2005b). Strength of supply chain is able to advance environmental efficiency and
set up corporate environment methods through ISO 14001 (Vachon and Mao, 2008).
Carter & Rogers (2008) demonstrated the intersections of 3BL and the broad view of SSCM in Figure 1.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

155
 FIGURE 1: SUSTAINABLE SUPPLY CHAIN MANAGEMENT (CARTER & ROGERS, 2008).

SUSTAINABLE SUPPLY CHAIN INTEGRATION

According to Vasileiou and Morris (2006), it is essential that firms figure out influence in the promotion
of sustainability within supply chain. These are relative influences which advocate for practices of
sustainability to be accepted. This suggests that, the rapport within those who partake in supply chain is a
pivotal factor of sustainability. Gupta & Palsule-Desai (2011) identified three unified aspects of SSCM.
Firstly, the effect of organisational actions on the environment must be keenly considered as an elemental
component in their decision-making and not as pressure from external sources including government, civil
societies, communities and the likes. Also, it is important to closely keep an eye on impacts of all stages
of the whole value chain on the environment. Finally, component is relative to the corporate blueprint
which is above functional viewpoint to represent a universal standpoint of sustainability that merges
challenges and solutions through all functional perimeters.
The unification of 3BL raises prestige among patrons, suppliers, workers and stakeholders thus,
facilitating economic gains (Capaldi, 2005). Sustainable supply chain integration has an ultimate aim of
attaining tenable stream of information, goods and services, and investments for the provision of increased
value to corporate (Wolf, 2011). Institutions are admonished to be anticipatory in respect of sustainable
operations by Carter and Dresner (2001). This, in turn, will better competitive edge of organisations since
it would be tough to imitate their initiatives. Accordingly, on successfully integrating three facets of the
‘triple bottom line’, there could be economic gains through enhancement of communal standards and
conservation of the environment for the sake of generations to come. Basically, sustainable supply chains
take into account the use of correlation existing within the partakers (that is players of supply chain),
resources and actions. Also, it considers relationships such as interplay, organisation, rivalry, and
collaboration (Svensson, 2007). The principal factor of supply chain setting up sustainable procedures is
the link between suppliers and key companies. Many benefits could be attributed to a strong supply
relationship including; cost effectiveness, functional efficiency, adaptability to dynamic demands,
minimising carbon emissions, high-tech novelty among others (Simpson and Power, 2005).
Notable challenges in the implementation of SSCM methods as suggested by Morali & Searcy (2013)
include:
• Inability to understand the elaborate interaction between sustainability pillars and its effect on the
economic bottom line.
• Financial commitments
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

156
 • Management of risk and supervision of suppliers
• Measurement
• Clarity of knowledge and information
• Adjusting initiatives of SSCM with corporate strategy
• Organisational culture

Other obstacles are; the rising intricacy of designation, inadequate information sharing between
associates, lack of proficiency on the part of manager, and expensive management (Seuring and Müller,
2008).
Walker and Brammer (2009) suggest that it would be beneficial to check some of the general situations of
the company before applying the sustainability practices. These points are;
• Investigate the sustainability approach, regulations and policies
• Organizational culture fit including openness to the change
• Catch suppliers who have environmentally friendly practices and process
• Ascertain that they have the ability to implement sustainability according to cost, quality, and
culture

RESEARCH METHODOLOGY
This is an exploratory research built on a case study. To present comprehensive view, to enhance the
objectivity and to increase the construct validity multiple sources of data is used. , Firstly, online interview
was conducted with the company’s corporate sustainability expert and purchasing manager. Semi-
structured interview technique was used to be more spontaneous and make participants feel comfortable.
Before the meeting started, interviewer asked the permission for recording then the meeting is recorded.
The online meeting has taken 63 minutes. The 15 open-ended questions were asked. The recordings were
transcribed verbatim into an excel file and each author independently reviewed the verbatims to enhance
the validity. In the second stage, data is collected from Dogus Automotive Corporate Sustainability Report
2020, Dogus Group Corporate Sustainability Report 2015 and Dogus Global Compact Report 2020.
To interpret the data, content analysis has been applied. The importance of content analysis is to
determine definite statements i.e. statements that contain expressions (Gillham, 2000). Case study
research gears towards exploratory more than confirmatory in that, there is the identification of themes
and classifications of phenomena and not proving corelations or testing hypotheses. It as well involves
collection and the analyses of data from diverse (Hancock, 2017). Interview transcription and the Dogus
Global Compact Report 2020 were entered into the MAXQDA software program. Codes were created,
then they were transferred to an Excel file. To ensure reliability, particular codes are further categorized
under second order themes separately by the two researchers. The differences between the coding of the
authors were discussed and the coding protocol was revised until a full agreement was reached. The inter-
coder reliability demonstrated a satisfactory level of agreement (92%).
Dogus Group as a Case Study
In this part, first of all, the general company information of the Dogus Group is given. Afterwards, the
sustainability practices of the company and the problems it faces in its integration into the sustainable
supply chain are mentioned. The main reason why Doğuş group was chosen for the case study was that
the group has many companies in different fields, the number of suppliers is high, and the sustainability
report is considered to be good example for the sector. Doğuş Group operates in six main sectors:
automotive, construction, media, food & beverage, tourism & retail, real estate and energy. The Group has
more than 300 companies and 19 thousand employees. The reason to choose Dogus Group as a case study
is; one of their sub-group ‘Dogus Automotive’ has been sharing sustainability report for 10 years. The
authors' goal is to publish their experiences in an academic setting to inspire and constitute a road map for
other organizations in all industries looking for sustainability integration into their supply chains.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

157
 FINDINGS
RQ1: How does company perceive sustainability?

As presented in Table 1, sustainability perception of Dogus Group is comprised of four themes which are:
Sustainability as a Corporate Value (f=6), Sustainability as a Civil Liability (3), Cooperation with Supply
Chain (SC) partners (f=1) and Sustainability as a Philosophy (f=6). Hence, group This can be confirmed
by the corporate sustainability manager of the company with upcoming sentences:
‘The most striking issue for us is the adoption of this definition by all our stakeholders, especially our
employees, its impact on our lifestyles and making it a permanent element’.
TABLE 1: SUSTAINABILITY PERCEPTION OF THE COMPANY
Second Order Theme First Order Theme f
Sustainability as a Philosophy -Saving awareness 6
-life and business style
-holistic view
-consumption awareness orientation
-customer employee satisfaction
-Triple Bottom Line
Sustainability as a Civil Liability primary duty 3
everyone's responsibility service to the country

Cooperation with SC partners involving all stakeholders into sust. 1

Sustainability as a Corporate Value DNA 6
Not a new terminology
Company reputation
Permanent
‘Recover (evaluate), improve, transform’ motto
Adapting the processes, mission to susta.

RQ2: What kind of practices company does for sustainability?

Dogus Group embraces various sustainable practices which are categorized in seven categories.
Sustainability report (f=1), economic sustainability practices (f=2), developing sustainable business model
(f=2), corporate sustainability (f=6), environmental sustainability practices (f=2), social sustainability
practices (f=13), sustainability criteria for supplier selection (f=5) emerged as the second order themes
obtained.social sustainability practices is the mostly repeated category which implies that the current
practices focus mainly on social pillar of sustainability.. The corporate sustainability manager said that
‘When you look at our DNA, we have the principle of serving the country. The purchasing manager made
a contribution as, ‘we do not call customer, we say guest instead’.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

158
 TABLE 2: SUSTAINABILITY PRACTICES OF THE COMPANY
Second Order Theme First Order Theme f
Sustainability report Sustainability report publication 1
Economic sustainability practices Special auditing against corruption 1

Developing sustainable business Creation of new business model 2

model Circular economy practices
Corporate sustainability Creating teams for sust. 6
Adopting mission and processes for sust.
Governing model
Concrete targets
Whatever the subject is discussing sust. side
Environmental sustainability Increasing the number of waste bins 2
practices Afforestation projects
Auditing
Signing the Declaration of Equality at Work 13
Behaving equally to all employees
CSR practices
Social sustainability practices The construction of Soma Doğuş Houses
donations
UN Women's Empowerment Principles (WEPs)
7/24 support line for employees (health,psychology,
nutrition..)
Workshops for children
Guiding suppliers for business ethics principles
Employee Health and Safety
Code of Ethics study
Foundation support educational activities
Museum and research center
Sponsorhip
Sustainability Criteria for Supplier Human rights 5
Selection Employee rights
Employee's health and safety
Environmental impact
Have to follow regulations:

RQ3: What are the efforts they put in order to integrate sustainability into the supply chain?

Projects Development (f=2), Effective SCM Efforts (f=4), Digital Innovation (f=1) and Meetings (f=2) are
the main categories obtained under sustainable integration indicated Table 3. The group plans and
executes different projects in order to spread and improve sustainability integration not only for internal
but for the whole supply chain integration. Their initiatives include all the stakeholders such as suppliers,
customers and employees. Their great advantage is having an automotive group that implements and
becomes an expert in this field and guides other groups on sustainability issues. This can be confirmed by
the corporate sustainability manager:
‘Our automotive company's guidance on this journey was important for us in terms of know-how design’

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

159
 TABLE 3: SUSTAINABLE INTEGRATION EFFORTS OF THE FIRM
Second Order Theme First Order Theme f
Projects Development Project: regardless of the subject, discussing 2
sustainability side
Supplier selection according to sustainability
Effective SCM Efforts Forcing suppliers for sustainable activities 4
Know-how to other sub groups
Information distribution from many channels
Spreading 3BL to suppliers and customers
Digital Innovation DRIVE program in automotive group 1
Meetings Regular info meetings w/all stakeholders 2
Inter departmental works for integration

RQ4: Do they spread the sustainability perception to all stakeholders?

Even though, sustainable integration and spreading sustainability perception to all stakeholders are very
familiar concepts, it was a necessary distinction to understand the ways in which they spread the
perception to the stakeholders. In sum, digital innovation (f=4), auditing (f=1), supply chain relations
(f=3) are the way to spread sustainable perception. Corporate sustainability manager supported this result:
‘Currently, company-based data of all departments within ourselves is transferred to the dashboard. All
of our employees will be able to follow sustainability activities in digital environments’.
TABLE 4: EFFORTS OF SPREADING SUSTAINABILITY PERCEPTION TO ALL STAKEHOLDERS
Second Order Theme First Order Theme f
Digital Innovation Publishing all the sustainable data in dashboard and digital 4
platforms
New platform creation for communication
Projects/ideas sharing platform: bringing employees and
manager together
Notifications to increase awaraness
Auditing Hygen auditing of suppliers 1
Supplly Chain Relations Long-term partnership w/suppliers 3
Guiding suppliers for business ethics principles
Trying to prove sustainability is not expensive

RQ5: What are the challenges for implementation?

The most prominent challenge is actually hidden in their primary purpose. They want to spread
sustainable awareness and show sustainability is their duty as a company and it is a society’s concern. So
the mostly repeated challenge appears as Stakeholder Involvement Concerns with a frequency of four (see
Table 5). The group’s aim is to integrate this awareness to every single employee and spread it to society
with their projects and initiatives. The second challenge obtained is measurement concerns involving data
collection, setting target and lack of consolidated data; consolidated data collection. ‘If we cannot
measure, we cannot manage’ said purchasing manager.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

160
 TABLE 5: CHALLENGES OF IMPLEMENTATION SUSTAINABILITY IN FIRM
Second Order Theme First Order Theme f

Measurement Concerns Data collection 3

Setting target
Lack of consolidated data
Stakeholder Involvement Resistance to change in logistics side 4
Concerns Lack of systematization in logistics side
Misperception about implementation cost
Spreading sustainability philosophy to all stakeholders

CONCLUSION
Consciousness of customers regarding the environment and issues of social sustainability are on the
increase. Accordingly, supply chains are compelled to create and set up production and delivery systems
that are sustainable. Formation of any sustainability schedule in supply chain demands forming and
operating a unified method addressing pertinent socioeconomic issues and that of the environment. The
framework suggested in this paper has conspicuously indicated that, the comparative cooperation and
amalgamation of distinctive components of supply chain is capable of ensuring economic gains.
(Gopalakrisnan, 2012).
Silvestre (2015) suggests that, relative to the huge rivalry within the worldwide business environment,
academics, policy makers, leaders of businesses recognise the significance of suitable SCM as an
indispensable element to attain competitive edge both locally and on national level.
After reviewing key sustainable initiatives and supply chain literature, Dogus Group has successfully
integrated sustainable practices into the supply chain. All the 3BL (economic, social and environment)
factors are included in this integration. It should be noted that sustainability is a matter of consciousness
and awareness and cannot be fully integrated unless it becomes the culture of the company or the
integration will be temporary.
The limitation of the study; Although case study research is more exploratory than confirmatory that gives
detailed information, it is ad-hoc and cannot be generalized. Also because each company may have
different supply chain processes which means the integration focus and the drivers may change through
the processes and the company culture.
Future research would be conducted with each sub-group of the company members, partners, suppliers.
Which would enhance more comprehensive perspective about the company integration. In addition, it can
be supported with a quantitative study through conducting an interview with all (as much as possible)
employees.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

161
 REFERENCES
Bowersox, D.J., Closs, D.J. and Cooper, M.B. (2002), Supply Chain Logistics Management, McGraw-Hill/Irwin,
New York, NY.
Brockhaus, S., Wolfgang, K., & Knemeyer, A.M. (2013). Where do we go from here? Progressign sustainability
implementation efforts across supply chains. Journal of Business Logistics, 34, 167–182. doi: 10.1111/jbl.12017.
Carter, C.R., Dresner, M., 2001. Purchasing role in environmental management: cross-functional development of
grounded theory. Supply Chain Management 37 (3), 12–26.
Carter, C.R., & Rogers, D.S. (2008). A framework of sustainable supply chain management: Moving towards new
theory. International Journal of Physical Distribution & Logistics Management, 38, 360–387. doi:
10.1108/09600030810882816.
Capaldi, N., 2005. Corporate social responsibility and the bottom line. International Journal of Social Economics 32
(5), 408–423.
Fawcett, S. E., Ogden, J. A., Magnan, G. M., & Cooper, M. B. (2006). Organizational commitment and governance
for supply chain success. International Journal of Physical Distribution & Logistics Management.
Gillham, B. (2000). Case study research methods: Bloomsbury Publishing.
Gopalakrishnan, K., Yusuf, Y. Y., Musa, A., Abubakar, T., & Ambursa, H. M. (2012). Sustainable supply chain
management: A case study of British Aerospace (BAe) Systems. International Journal of Production Economics,
140(1), 193-203.
Gonza´lez-Benito, J. and O. Gonza´lez-Benito: 2005b, ‘Environmental Proactivity and Business Performance: An
Empirical Analysis’, Omega 33(1), 1–15.
Gupta, S., & Palsule-Desai, O.D. (2011). Sustainable supply chain management: Review and research opportunities.
IIMB Management Review, 23, 234–245. doi: 10.1016/j.iimb.2011.09.002.
Henkoff, R. (1994), ``Delivering the goods'', Fortune, 28 November, pp. 64-78.
Laurin, F., & Fantazy, K. (2017). Sustainable supply chain management: a case study at IKEA. Transnational
Corporations Review, 9(4), 309-318.
Mollenkopf, D., Stolze, H., Tate, L., Ueltschy, M., 2010. Green, lean, and global supply chains. International Journal
of Physical Distribution and Logistics Management 40 (1/2), 14–41.
Morali, O., & Searcy, C. (2013). A review of sustainable suppy chain management practices in Canada. Journal of
Business Ethics, 117, 635–658. doi: 10.1007/s10551-012-1539-4.
Mosteanu, N. R., Faccia, A., Ansari, A., Shamout, M. D., & Capitanio, F. (2020). Sustainability Integration in
Supply Chain Management through Systematic Literature Review. Calitatea, 21(176), 117-123.
Pagell, M. (2004), ‘Understanding the factors that enable and inhibit the integration of operations, purchasing and
logistics’, Journal of Operations Management, Vol. 22 No. 5, p. 459.
Silvestre, B.S. (2015). Sustainable supply chain management in emerging economics: Environmental turbulence,
institutional voids and sustainability trajectories. International Journal ofProduction Economics, 167, 156–169. doi:
10.1016/j.ijpe.2015.05.025.
Simpson, D., Power, D., 2005. Use the supply relationship to develop lean and green suppliers. Supply Chain
Management 10 (1), 60–68.
Svensson, G., 2007. Aspects of sustainable supply chain management (SSCM): conceptual framework and empirical
example. Supply Chain Management: An International Journal 12 (4), 262–266.
Seuring, S. ve Müller, M. (2008). From a literature review to a conceptual framework for sustainable supply chain
management. Journal of Cleaner Production, 16(15): 1699-1710.
Vachon, S., Mao, Z., 2008. Linking supply chain strength to sustainable development: a country-level analysis.
Journal of Cleaner Production 16, 1552–1560.
Vasileiou, K., & Morris, F. (2006). The sustainability of the supply chain for fresh potatoes in Britain. Supply Chain
Management: An International Journal, 11, 317–327. doi: 10.1108/13598540610671761.
Walker, H., & Brammer, S. (2009). Sustainable procurement in the United Kingdom public sector. Supply Chain
Management, 14(2), 128-137. doi:10.1108/13598540910941993
Wiese, B., Knecht, M. (2015). Socialization into organizations and balancing work and family. In Vuori, J., Blonk,
R., Price, R. (Eds.), Sustainable working lives. Managing work transitions and health throughout the life course (pp.
87-105). New York, NY: Springer. doi:10.1007/978-94-017-9798-6_6.
Zhou, Z., Cheng, S. ve Hua, B. (2000). Supply chain optimization of continuous process ındustries with
sustainability considerations. Computers and Chemical Engineering, 24(2-7): 1151 – 1158.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

162
 BLOCKCHAIN’S POTENTIAL FOR WASTE MANAGEMENT IN TEXTILE
SECTOR
Elif Dursun1, Yasemin Ülker2, Yavuz Günalay3

Abstract ⎯ Transparency in a supply chain makes it possible to monitor whole process. Thus, information,
products, services, and payment exchanged within chain are enabled to be observed clearly. However, most of
the firms cannot reach transparency goals due to some reasons like complexity, information asymmetry, lack of
control, etc. Blockchain technology comes as a solution to those problems where it changes supply chains in an
evolutionary way. Blockchain since it gained popularity, is opening new application areas in supply chain with
an innovative and sustainable approach. One of the areas its contribution will make difference is waste
management. Waste management is an essential topic to all supply chains since it is an important step to
sustainable world. Textile sector is one of the critical ones because as Ellen MacArthur Foundation’s 2017
report states that within 30 years, this industry will consume 15% of world’s carbon budget. Annually, 14% of
fiber is lost during production, collection, and processing phases. Blockchain as a new technology proposes
innovative solutions to such problems. In this study, the potential of blockchain technology to overcome
sustainability problems by focusing on waste management in textile sector is discussed. Main aim of the study
is to provide an alternative solution framework utilizing blockchain to the current waste management activities
in textile sector. A qualitative study approach adopting in-depth interview methodology is developed to analyze
waste management process of actors in textile industry. In depth interviews are conducted and the data acquired
is analyzed via MAXQDA software. Findings of the study show that blockchain technology is still its infancy
and needs to be communicated to the actors of the sector. The technology has low potential due to the barriers
it faces during development phase. However, it is a topic of technological development on the table of large
players for now.

Keywords ⎯ Blockchain, supply chain, sustainability, waste management, textile industry, qualitative study

INTRODUCTION
Increasing population all over the world causing consumption rates to rise up to the amounts that threaten
the limits of existing resources. It is a risk for environmental sustainability, so policies are being developed
to overcome this problem. At this point, waste management policies have high importance to control the
situation and increase awareness on the subject. Besides environmental impacts, waste management is also
important to companies to minimize resources and costs incurred in the processes. More broadly, it leads
companies to be a part of circular economy which is a system based on principles of recycling, reusing, and
regenerating products and materials during the whole process. In this sense, such system brings not only
environmental benefits, but also economical and individual benefits like profit generating, innovation, job
creation, better utility of material, decreased depreciation of products and materials, healthy environment,
etc. (Ellen MacArthur Foundation, 2017).
As world population growing fast, the waste they generate also growing where The East Asia and Pacific
region have the largest share by 23% and the Middle East and North Africa have the smallest by 6 %. In
total, municipal solid waste produced annually is 2.01 billion tons of which 34% is produced by developed
countries (European Environment Agency, 2016). In 2016, USA (258 million), China (220.4 million), and
India (168.4 million) are the top producers of municipal solid waste worldwide, while Turkey produced
31.28 million tons. Municipal Solid Waste is expected to be 2.59 billion tons in 2030 and 3.4 billion tons in
2050 worldwide (Tiseo, 2021). Industrial waste – generated in Europe -, on the other hand, has stable trend

1
Elif DURSUN, Bahçeşehir University, Faculty of Economics, Administrative and Social Sciences, Department of Logistics
Management, Istanbul, Turkey, [email protected]
2
Yasemin ÜLKER, Bahçeşehir University, Faculty of Economics, Administrative and Social Sciences, Department of Logistics
Management, Istanbul, Turkey, [email protected]
3
Yavuz GÜNALAY, Bahçeşehir University, Faculty of Economics, Administrative and Social Sciences, Department of Logistics
Management, Istanbul, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
163
over 2010-2016. Annual waste generation of EU-28 manufacturing industry in 2016 is recorded as 250, 28
million of tones (European Environment Agency, 2021).
Industrial activities are important to development of cities and countries as it increase welfare and economic
growth. However, they are giving harm to the environment and human health. European Union is one of the
regulatory bodies developing policies for pre-cautions and post-actions regarding waste management. The
Waste Framework Directive (WFD, 2008/98/EC) is one of the policies providing a general framework with
definitions of concepts. Circular Economy Package prepared in 2018 is another framework document for
waste management in a broader sense and strategic way since it covers the whole cycle from production to
consumption with the aim of closed loop lifecycle (European Environment Agency, 2021). this approach to
waste management, increasing recycling of municipal waste by 65%, decreasing landfill rates to 10% within
15 years, including textiles within separate collection obligations by 2022 are among EU’s aims (European
Commission, n.d.). In Turkey, legislations are determined and documented under Waste Management
Directive by Ministry of Environment and Urbanization (Çevre ve Şehircilik Bakanlığı, 2015). In
accordance with this directive and other policies, 1698 facilities operated for waste disposal and recycle. In
those facilities, 44 million tons of waste disposed, 310.000 tons incinerated, and 33 million tons are recycled
in 2016 (TÜİK, 2019).
Textile industry is one of the manufacturing industries that causes environmental damage as it dispose both
solid waste, water waste, and air emissions. Market share of textile industry is considerably large with 925.3
billion USD value in last year (Grand View Research, Inc., 2021). According to European Parliament
Briefing (2019), expenditure on textile products is increased by 40% within 16 years due to decreasing
prices and availability of fashion products. It is predicted that the Europeans don’t use more than 30% of
their clothes per year and mixed with household waste which almost half of them are not to be recycled (
European Union, 2019). Most of the processes of textile production from raw material to customer and
reverse are not suitable for tracking by authorities since current systems like Life Cycle Assessment (LCA)
and Environmental management systems (EMS) could not provide enough transparency through the chains.
Beton et all. (2014) in their report on environmental impact of textile production, emphasize that there is a
lack of market and flow data, environmental data, and social data which cause disconnection within supply
chain of textile industry. Due to this information gap, waste management strategies would not be properly
and efficiently applied to minimize textile waste in manufacturing phase. They also state the important role
of new technologies which can contribute to the recycling and reuse of materials (European Union, 2014).
In this sense, blockchain as a new technology which enhances the way to track data flows would be an
effective alternative to the current systems. Promising the features of secured data storage and immutability,
blockchain technology can bring a new approach to the waste management methods with a trustworthy
record of waste generated throughout chains. Recently, EU has aimed at closed loop product cycles with
respect to Circular Economy Package in which it is presented that returning products or materials into the
economy would increase recycling and reuse rates with ability of better collection and management of waste
(Koligkioni et al. 2018). Supply chains are complex structures as they have different players within.
Therefore, tracking the waste generation gets hard as proceeding through production route. For textile
industry, there are various steps till a raw material – mostly cotton or synthetic fiber – is turned into a fabric
or clothing to be marketed. In each step, raw material is treated differently, and it beclouds reaching waste
management goals.
In this study, it is aimed at discovering how the blockchain technology - as an alternative way - can be
utilized for tracking the waste generated through textile supply chains. Textile industry is chosen because it
is important to the human health as well as balance of nature. Disposed waste can be traced with various
instruments like RFID, Barcodes, and other technologies. However, they are not enough to provide
transparent and trustworthy information. A much more integrated and distributed system like blockchain is
needed for better control over waste disposals. Thereby, we specify our research questions with regards to
three aspects which are sustainability, waste management, and blockchain technology. Sustainability aspect
is limited to environmental dimension as we mainly focus on waste management. Thus, social and
governmental dimensions had to be ignored in this study. To answer the research questions and objective,
in-depth interview methodology is adopted in the study. A specific number of companies (ten) are
determined from textile sector and those that accepted our interview invitation (three) were interviewed
with open ended questions on an online platform. Analysis of interviews are carried out on a qualitative
analysis software called MAXQDA. Research questions and objective are answered after the interview

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
164
transcripts are categorized and analyzed with the help of this software. Backed by literature review and
definitions of these concepts, this study has nine research questions. Those include:
• RQ1: What is the sustainability perception of textile companies?
• RQ2: What kind of sustainability practices and methods are being implemented?
• RQ3: What are the contributions of textile companies for circular economy?
• RQ4: What are the waste management methods and practices in textile companies?
• RQ5: What kind of traceability methods are being used for waste?
• RQ6: What are the barriers that textile companies face in waste management (WM)?
• RQ7: How familiar the textile companies with blockchain technology (BCT)?
• RQ8: What is the perception of textile companies about integrating BCT in WM systems?
• RQ9: What are the barriers to the blockchain technology in waste management?

In summary, main parts of this paper is designed as (1) definitions of main concepts and literature review
(2) rationale and explanation of methodology used in the study, (3) presentation of findings, and (4)
conclusion and limitations of study with discussion from literature.

CONCEPTUAL BACKGROUND AND LITERATURE REVIEW

Sustainability
Sustainability is a term root back to the 90s where it was defined with respect to triple bottom line of people,
planet, and profits. After a couple of developments and new approaches, it is covered by the terms of
environment, society, and governance of investment decisions. Hereby, the term is started to be redefined
according to these aspects where it includes four big buckets where Hedstrom (2019) explains in his book.
These buckets consist of Environmental Stewardship, Social Responsibility, Governance, and Strategy and
Execution. Although it is confused with green and corporate social responsibility, the term is officially
defined as ‘...the pursuit of a business growth strategy that creates long-term shareholder value by seizing
opportunities and managing risks related to the company’s environmental and social impacts’.
As time pass, sustainability has become an important feature for supply chains, and it required businesses
to take critical actions. Ahi and Searcy (Ahi and Searcy 2013) define sustainability with supply chain
approach after detailed research on sustainability and supply chain literature. In their study, sustainability
is defined as ‘The creation of coordinated supply chains through the voluntary integration of economic,
environmental, and social considerations with key inter-organizational business systems designed to
efficiently and effectively manage the material, in-formation, and capital flows associated with the
procurement, production, and distribution of products or services in order to meet stakeholder requirements
and improve the profitability, competitiveness, and resilience of the organization over the short- and long-
term’. Roxana MOSTEANU et al. (Roxana MOSTEANU et al. 2020), in their paper, summarize the other
definitions made for sustainable supply chain management (SSCM) highlighting the aspects of social,
natural, and financial objectives.
Waste Management
On the DIRECTIVE 2008/98/EC (THE EUROPEAN PARLIAMENT AND THE COUNCIL OF THE
EUROPEAN UNION, 2008) ‘Waste’ means any substance or object which the holder discards or intends
or is required to discard; and ‘waste management’ means the collection, transport, recovery and disposal of
waste, including the supervision of such operations and the after-care of disposal sites, and including actions
taken as a dealer or broker. As the definitions state, waste is produced and flows within the supply chain
activities from raw material to the customer and from customer to the disposal. Therefore, we could say that
minimization of waste starts with the journey of raw material turning into a finished good to be used by
consumers. The factors of sustainability – environment, economy, and society – are considered to be critical
for the continuity, and current population with increasing consumption rates is putting these factors in
danger. Organizations and society are forced and motivated to take precautions to decrease waste
generation. However, it is hard to prevent waste to be generated, so tracking of waste gets more important.
For waste management goals, there are different initiatives being performed. EU is very strict to change the
way how waste is disposed and managed not only with the systems they build but also from the economic
perspective. They have put The Circular Economy Package into action in 2018 with the support of Horizon

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
165
2020 projects. Circular Economy stipulates value of products, materials, and resources to be sustained in a
circular way where they cycle within economy to minimize waste production and carbon emissions by
contributing to a sustainable, competitive, and resource efficient economy. Actions of the Package promote
circular economy model in each step of value chain (EUROPEAN COMMISSION, 2018). Figure 1 below
displays the circular economy diagram from value chain perspective.

FIGURE 1: CONCEPTUAL DIAGRAM ILLUSTRATING THE CIRCULAR ECONOMY IN A

SIMPLIFIED WAY (EUROPEAN COMMISSION, 2018)
Along with Circular Economy Package, European Union has other policy measures for sustainability
concerns. Among those policies, there include Waste Framework Directive, Environment Action
Programme, and Horizon 2020 of all focusing reducing waste, maximizing recycling and re-use rates,
limiting incineration to non-recyclable materials, decreasing landfilling rates of non-recyclable materials,
and ensuring full implementation of policies by the member states (EUROPEAN COMMISSION, 2018).
USA also has strict policies and regulations regarding waste generation as they produced 262.4 million tons
of municipal solid waste in 2015 of which only 91.2 million tons (34%) are recovered. The Environmental
Protection Agency (EPA) is the regulating body with 1976 Resource Conservation and Recovery Act
(RCRA) (Tiseo, 2020). EPA provides guide for industrial wastes management processes. The guide helps
firms to understand and assess risks caused by industrial wastes with some tools such as Industrial Waste
Management Evaluation Model (IWEM), Composite Model for Leachate Migration with Transformation
Products (EPACMTP) and Industrial Waste Air Model (IWAIR). EPA covers different sectors from
manufacturing to transportation. Textile manufacturing is one of the sectors giving hazardous impact to the
environment due to air and water pollution it causes. RCRA enforces any textile manufacturing facility to
manage their hazardous waste as defined in the RCRA compliance paper focusing on textile manufacturing
(Environmental Protection Agency, 2002).
In Turkey, waste management is regulated under the Environmental Law constituted in 1983 (ÇEVRE
KANUNU Kanun Numarası : 2872, 1983). At the national level, waste management strategy is assured by
the Constitution, laws and implementing regulations, international conventions, national plans, programs,
etc. Polluters are sanctioned by The Turkish Criminal Code and Law on Minor Offences. The Solid Waste
Control Regulation constituted in 1991 especially forces stakeholders to minimize waste generation, to build

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
166
a structure where they can properly classify recoverably and non-recoverable wastes for better and eco-
friendly waste management procedures to be performed. The Regulation on the Control of Hazardous
Wastes has been put into effect in 1995 and altered in accordance with the EU Acquis.
Blockchain Technology
Blockchain is a decentralized network of blocks that are chained together one after one. It is a technology
that has the characteristics of consensus, immutability, transparency, automation, and security. It enables
participant to have ownership and self-governance of data they generate. Therefore, it is called as a trust-
free system where no third party involvement is needed. That means eliminating the procedures and
lowering the workload these procedures carries as well. Intermediaries would be obstacles to the speed and
might incur extra costs for a work to be done (Scott, de Castro Neto, and Pinheiro 2021). Companies are
trying to be efficient and effective, and elimination of intermediaries is an important opportunity to
accomplish these objectives. Additionally, this ability makes them trace the procedures in a transparent and
fluent way. With respect to these opportunities blockchain technology promises, researchers are also
intensively studying this technology from various perspectives. Although BCT has got its popularity in
finance, there are many other application areas unleashed by academics and industry this technology may
change in the future. Waste Management in Supply Chain and Logistics is one of these areas.
In literature, Scott et al. (Scott et al. 2021) distinguished three different areas where blockchain is proposed
as a solution for waste management. These areas include supply chain management, providing incentives,
and waste services market / platform. For supply chain management topic, they have reported some studies
focusing on coordination of supply chains and the track of waste and recycling procedures. Track of waste
and recycled material is important in a way that it puts responsibility on producer of waste and provides
transparency for users and regulators. It is also stated that although new technologies like smart bins are
being utilized, these systems are still centralized and open to data loss and have trust issues. Blockchain
technology challenges the existing applications by providing transparent, automated, disintermediated, and
decentralized features.
Ahmad et al. (Ahmad et al. 2021) in their study focus on the medical supply chain waste with a blockchain
approach. They have developed a blockchain-based solution for tracking COVID-19 medical equipment
waste to provide a fully secure, transparent, traceable, and trustworthy supply. They conclude that their
application is important in assisting authorities in waste management and application of various other cases
is possible with little changes and adaptations. Kassou et al. (Kassou et al. 2021) also studied medical waste
and proposed that blockchain based medical and water waste management would ensure control of waste
and effective management of disposal. Taeihagh et al. (Taeihagh et al. 2020) explore the opportunities and
challenges of blockchain for clarity of property rights, supporting law and policy goals, and maintaining
anonymity and privacy in sustainable waste management. They also state the challenges the technology
faces, which being a developing technology, being replaceable by existing centralized systems, and data in
blockchain does not prevent waste creation.
Kumar Agrawal et al. (Kumar Agrawal et al. 2021), in their study, focuses on textile and clothing sector
and studied traceability of these chains. They develop a blockchain framework for organic cotton supply
and try to provide traceability through supply network, including all partners. They claim that this
framework would build a technology-based trust among the supply chain partners. (França et al. 2020), in
another study, propose a blockchain framework for solid waste management in a municipality to replace a
paper based system. With this new system, it is expected to ensure integrity, support operational scale gain,
generate reliable and auditable data, and replication of model to other municipalities. Gopalakrishnan, Hall,
and Behdad (Gopalakrishnan, Hall, and Behdad 2020), in their study, analysis the different blockchain
applications specifically developed for waste management like Swachhcoin, Recereum & Plastic. Ongena
et al. (Ongena et al. 2018) also discuss waste management activities and explore the blockchain technology’s
potential in terms of problematic areas and solution recommendations that the technology enables. They
have determined five problematic areas in waste management which are: (1) Fraud and manipulation, (2)
Wrong or loss of information, (3) Manual processes, (4) Lack of knowledge about technology, and (5) Lack
of control. They specifically offer blockchain solution for these topics but also highlight that blockchain is
not the mere problem solver for them.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
167
 RESEARCH METHODOLOGY
The research method of this study is qualitative. In this context, ten textile companies were requested to be
interviewed via e-mail for online interviews, but three companies were accepted. In qualitative studies, the
ideal number of participants could be determined when the data is saturated. This study is still in progress,
and more data will be collected until reaching the purpose. Meetings were held on online platforms like
Zoom and MS Teams at the agreed time and date of interviewees. Before starting the interview, the
interviews were recorded with the permission of the participants. A semi-structured interview technique
was used in the interviews. The main reason for using this technique is to add new questions when necessary
by asking open-ended questions, make the participant experience a conversational mood, and reach more
detailed information through free association. The interviews lasted 70 minutes at an average. Before
starting the interview, the participants were asked to introduce themselves and give information about their
job descriptions briefly. Interviews consist of two parts. In the first part, the company's general sustainability
activities (focusing on environmental practices), waste types, waste management techniques, waste tracking
systems, the situation of the textile industry, and the difficulties encountered were asked. In the second part,
the awareness of blockchain technology was asked. The definition, usage areas, advantages, and
disadvantages of blockchain technology were given, and the tendencies to use such a technique as a waste
tracking system were asked. Fifteen questions were asked during the interview, and detailed explanations
were made where the participant or interviewee did not understand. In this way, the interviews were
concluded in an interactive and discussion atmosphere.
As Bill (2000) suggested, a transcript of the interviews should be made immediately or within a few days
after each interview was terminated. After all the interviews were completed, all the transcript data were
entered into the MAXQDA software program, and the codes were created. After the generated codes were
transferred to excel, they were categorized in parallel with the research questions. Two interviewers
categorized the codes separately then they discussed again in order to support reliability. For the validity,
company’s websites and annual corporate reports are also reviewed. The tables within findings Heading do
not include all of the questions but include all coded answers within categories, and some were supported
by comments. The details of the analysis and the content of the questions are discussed in the findings
section.

FINDINGS
The categorization was framed according to objectives. However, it would be valuable to give detailed
information about the textile sector situation, companies' plans, municipality, and government’s role in
waste management from the view of participant companies.
TABLE 1: THE PARTICIPANT COMPANY PROFILES
Company Size Existence of Distinctive Future Plans
Type Treatment Applications
Plant
X Dye house Middle No Green factory Solar panels,
Micro-level counter system for
water saving
Y Retail Chain Huge No Ecology Zero Waste,
Harmonization Waste tracking system,
Program, Raw Material Maximum reuse,
Control Tool, Global Data validation software,
Recycle Certificate) Automated system for carbon
emission and water traceability,
Circular economy,
Z Textile Large Yes STeP by Oeko-Tex, Advanced treatment plant,
Producer Sustainable Textile Customer-based waste
Production Certificate, management,
Ecologic production Circular economy,

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
168
As indicated in Table 1, all three of the companies are textile companies. However, each firm represents
different rings of the textile supply chain. This situation enables us to understand how waste is managed
both by the manufacturer, the retail company, and the company that does not produce but has a large amount
of waste—this facilitated interpretation from a holistic point of view.
The main reason for choosing Company X; the higher amount of waste and chemicals used in dye house
factories. The fact that the retail company (Y) has no production does not mean that there is no waste and
does not need to manage any waste. It is a large company with a retail chain network means that it works
with too many manufacturers, and those companies can interfere with their waste management processes.
Lastly, the primary motivation for choosing Company Z is their variety of sustainable projects and
certificates.
Sustainability
Table 2 is coded and categorized according to RQ1, RQ2, RQ3, which are listed as:
• RQ1: What is the sustainability perception of textile companies?
• RQ2: What kind of sustainability practices and methods are being implemented?
• RQ3: What are the contributions of textile companies for circular economy?

Table 2 summarizes the sustainable perception, practices, and methods, and lastly, the circular economy
benefits from all three companies' point of views. It should be noted that sustainability practices cover the
circular economy concept, which also serves sustainability. The main difference is circular economy aims
to avoid waste generation from start (design phase) to the end. Its focus is production function, not the social
aspects.
In summary, all the circular economy practices can be accepted as sustainable practices, but not all the
sustainability practices can be accepted as circular economy since circular economy fits in only one bottom
of triple bottom line of sustainability concept. For this reason, separate categorization has been made. The
same context of the codes is shown as bold characters.
Since the participants knew that the main topic of the interview was waste management, they gave answers
focused on waste management while answering perception of sustainability questions. For this reason, the
environmental and economic aspects of sustainability are mentioned chiefly. All three companies are agreed
upon the need for adaption to sustainability; there is still time for maturation. While for Company X,
sustainability is not a priority, they also confirm the importance of the concept like the other two companies.
In general, companies perceive sustainability as a global concept, while big companies consider more and
trying to spread this into their company because their impact on the sustainability is much broader such as
Company Y. This comment from the Material Quality Assurance Manager confirms this finding: ‘It is
crucial to find out which point the work has touched within the scope of sustainability. I strive to share this
logic with our managers and colleagues’.
As sustainable practices, all the companies are trying to avoid the usage of energy, water, and chemicals.
They are trying to follow and adopt Sustainable Development Goals into their procedures. Since Company
Y is a giant retail company who has more efforts such as taking consultancy, innovative and advance
techniques compared to the other participants of the study. Regardless of the company size, companies have
a separate department for sustainable actions. Company X and Z are members of ‘Care for Water’, an act
for global water management strategy generated by giant fashion retailers.
Answers given to the category of Circular Economy questions reveal that all three companies have efforts
for supporting Circular Economy practices such as green production, reuse practices, energy, and water-
saving practices. The important point in Circular Economy is, all the practices should be designed from the
beginning and when a product comes to the end of its life, its materials are retained in economy.
‘When we come to 2023 – 25, we have goals such as using all of our packaging from circular sources,
obtaining all of the raw materials from sustainable sources, creating and implementing a circular
economy-based purchasing strategy, and increasing the share of our product groups in the circular
economy. Our goal is to apply the circular economy at the design stage’ (The Manager of Company Y)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
169
 TABLE 2: CATEGORIZATION OF THE SUSTAINABILITY PART FROM THE VIEW OF
PARTICIPANT COMPANIES
X Y Z
1. Need for adaptation 1. Adaptation 1. Adaptation
2. Not a priority 2. Global problem & impact 2. Important subject
3. Effort to avoid waste 3. Effort to avoid waste 3. Effort to avoid waste
4. Priority is production 4. Chemical content 4. Waste is problem
management 5. Costs
Perception

5. Following the developments

6. Preventing loss
7. Social point of view
8. Brand reputation/image
9. Waste is useless garbage for
others
10. Voluntary actions
11. Circular Economy

1. Setting goals 1. Setting goals 1. Setting goals

2. SDG harmony 2. SDG harmony 2. SDG harmony
3. Environmental Mng. 3. Environmental Mng. 3. -
Sustainability

Team 4. - 4. Care for Water

4. Care for Water act 5. Certification practices member (excellent
5. - 6. Track of inventory for level)
6. Emission filters emissions and water 5. Sustainable Factory
Practices

7. Energy saving practices 7. Chemical testing management Certificate

(natural gas instead of coal) 8. Supplier selection 6. -
8. Counter systems 9. Preparing a Green Collection 7. Project development
9. Filters for stencers Segment
10. Reuse of water by 10. Benchmark analysis
treatment 11. Taking consultancy
12. Research on innovative
solutions for uncontrolled
landfill/disposal
1. Water reuse by treatment 1. Reuse practices 1. Reuse practices
Circular Economy

2. Green production 2. Green purchasing 2. Clean production

3. Energy saving practices 3. Energy saving practices 3. -
Benefit

(natural gas instead of coal) 4. Changing consumer behavior 4. Product life cycle
4. Water saving practices (return) 5. Consortium
5. Maximum reuse 6. Projects
6. Materiality Analysis

Waste Management
In this section RQ4, RQ5, RQ6 will be answered, which are listed as:
• RQ4: What are the waste management methods and practices in textile companies?
• RQ5: What kind of traceability methods are being used for waste?
• RQ6: What are the barriers that textile companies face in waste management (WM)?

When ask dispose of waste, in general, companies are selling, reusing, recycling and giving away to licensed
companies and municipalities. Less the amount of textile loss, more tremendous the benefit both for the
environment and economy. Company Z supports this with this comment:
‘We don't even want waste to be created. In particular, we do not wish fabric and yarn waste to come
out because it affects our productivity. Every 1% loss is precious for us because we work with high

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
170
 endorsement levels. Besides, it is possible to recycle fabric and yarn waste. We sell 90% of our fabric
waste. We do not do any other recycling activities’.
None of the companies uses special software for tracking wastes. All of them uses cloud system and excel.
They are responsible for collecting the wastes inside the company. Then they transferred the wastes to a
licensed company. But they also follow after selling the waste and consider what companies do with their
wastes. Some wastes are given to municipality, which is compulsory. Company X Manager comments that:
‘It is sufficient for us to collect the waste correctly and send it to the licensed company. Particular attention
is paid to this in the audits’.
Company X and Y agree that it is difficult to separate wastes. Company X uses different colored boxing
system to make it easier in sorting the waste when created. This is what the Environment Manager
highlighted during interview. Company Y states that: ‘Waste is seen as rubbish in the sector and society.
However, it is a valuable thing that we have to manage and save it. It is possible for us to reuse or recycle
it. This is our duty to collect and make value out of rubbish’.
Three of the companies say that there is no proper waste tracking system and waste treatment, and it is too
expensive to have it. Besides, the lack of qualified employee in the sector, lack of incentives, cost of
operation and facilities are the other common barriers in waste management. Company Y added more
obstacles such as illegal acts, improper implementations from municipalities, expensive consultancy
services, etc.
‘The other problem is that people don't know how to recycle waste and benefit from it. Knowledge and
skill competence for the sector lacks in this regard. When there is a topic, it is necessary to sit down
and research and find new methods. We try to do this with my team. But when you start doing this, we
run into another problem, printer or online resource or guideline constraints. We are having difficulty
in finding guiding manuals in Turkish sources especially’ (Company Y Manager).
TABLE 3: CATEGORIZATION OF THE WASTE MANAGEMENT PART FROM THE VIEW OF
PARTICIPANT COMPANIES
X Y Z
Methods Methods Methods
1. Recycle 1. Recycle as energy 1. -
2. Landfill/Storage & 2. - 2. Collect-classify-
Temporary landfill 3. Selling to a licensed waste landfill/storage
3. Selling to a licensed disposal firm &Transferring to 3. Selling to a licensed
waste disposal firm & municipalities waste firm &Transferring to
Transferring to 4. Reuse municipalities
municipalities 5. Return 4. -
4. Reuse of refined 6. - 5. -
Methods & Practices
Waste Management

water 7. - 6. -
5. - 7. Treatment facility for
6. Incineration of sludge water and chemicals
7.-
Practices Practices Practices
1. Tracing what firms 1. Tracing procedures applied by 1.Tracing what firms do
do with our waste licensed firms & municipalities with our waste
2. Filter system for 2. Researching on regaining 2.Project development for
stenter value out of waste recycle
3. Refill 3. A new department special for 3.Waste map
waste mng & traceability 4.TPM and quality control
4. Working with innovative firms to avoid waste
5. Energy saving consumption 5.Productivity analysis
6. Green purchasing
7. Sustainable raw material usage
8. LCA calculations

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
171
 1. Online system (Cloud 1. Online System (Cloud system) 1. Online system (Cloud
system) 2. Municipality app (Track of system)
2. Municipality app inventory) 2. Municipality app
3. Excel 3. Excel 3. Excel
4. Manuel data entry 4. Manuel data entry 4. Manuel data entry
Traceability

5. Special employee 5. Special employee 5. Special employee

6. Track of measures 6. Track of measures 6. Track of measures
7. Micro level 7. Spot checking of stakeholder 7. Only tracking fiber and
traceability activity textile waste
8. Waybill system 8. New department for 8. Able to track full record
(filing) traceability of consumption but not
9. Need-based systems 9. Consultancy waste (waste is not the
focus)
1. Lack of traceability 1. Lack of traceability systems 1. Lack of traceability
system 2. - systems
2. Lack of incentives 3. Lack of qualified employee 2. Lack of incentives
3. Lack of qualified 4. Negative perception on reuse 3. -
employee products 4. Negative perception on
4. - 5. Costs (operation) reuse products
5. Costs (facilities) 6. Unproper sorting of waste 5. Costs (textile loss)
6. Unproper sorting of 7. Unconsciousness (Lack of 6. -
waste knowledge on waste) 7. Lack of continuity
7. Lack of recycle 8. Lack of coordination and
system communication
8. Lack of technology 9. Lack of systematized approach
Barriers

9. Hard to reach recycle 10. Not able to intervene

firms municipality procedures
11. Methods used by
municipalities
12. Firms' perception of waste as
garbage
13. Certification processes
14. Illegal acts in sector
15. Access to solution
alternatives
16. Access to education
alternatives (BCs, MCs)
17. Lack of printed resources
18. Expensive consultancy
services
Blockchain Technology
This section answers the last three research questions (RQ7, RQ8, and RQ9):
• RQ7: How familiar the textile companies with blockchain technology (BCT)?
• RQ8: What is the perception of textile companies about integrating BCT in WM systems?
• RQ9: What are the barriers to the blockchain technology in waste management?

In this section, main objective is to detect if the interviewed companies know about blockchain technology
and its potential for waste management. Thereby, the first question is “Do you know the main characteristics
and infrastructure of blockchain technology?” According to their answer and their demand of knowledge, a
brief explanation including technology infrastructure, its history, definition, features and application areas
were presented verbally by the interviewers.
As displayed on Table 4, it is seen that two of the companies have general knowledge on blockchain
technology but not deeper understanding of the infrastructure and how it works for companies. Especially

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
172
 in waste management, they specify that they have not come across with neither discrete not concrete
examples in textile supply chain. Basic knowledge and estimation of application areas are stated by
interviewees. When they were asked to list some possible application areas, one of them (Y) listed supply
chain and social application for wage management. Company X, on the other hand, stated that they don’t
have knowledge on technology, but they have heard about Cleanchain project while they were looking for
green solutions for their processes.
Integration raw in Table 4 are the coded analysis of answers given to the RQ8: What is the perception of
textile companies about integrating BCT in WM systems? Company X and Z state that they see the
integration of this technology as a hardship because of their priorities are different at the existing
circumstances and advantages that the company should exceed its obstacles. They see the existing
technologies as cheap and safe. Therefore, integration of BCT is not expected to be soon. They also
emphasized that this integration should be pioneered by large companies and could be beneficial for life
cycle assessment practices as well. Company Y, on the contrary, has more tendency to integrate this
technology in their processes as they study on this topic as one of their focuses in their waste management
practices. They specially stated that this technology is needed and has opportunities if it is developed further
in the future.
When we ask about the possible barriers to integration of BCT in waste management, they all have common
sense that it will be difficult to apply this technology to the dynamic and scattered structure of textile sector.
In addition to difficulty, they don’t see it as a priority as well. Interviewee from Company Y has explicit
sentences for this finding as:
‘There are so many hygiene factors until we get there. We can definitely say that there is a need, but the
question of whether it is a need today should always be on the table. The primary problem is meeting
the hygiene factors. Textile industry is a very scattered industry, and the unregistered-counterfeit
product problem is at high levels. Such technologies are needed to solve this, but unfortunately, we have
more urgent issues.’
Other barriers stated by interviewees are high investment costs (Company X and Z), lack of qualified
employees to implement this technology into the processes properly (Company Z), and lack of regulative
forces and incentives (Company Y).
TABLE 4: CATEGORIZATION OF THE BLOCKCHAIN PART FROM THE VIEW OF PARTICIPANT
COMPANIES
X Y Z
1. Don't know 1. General knowledge 1. General knowledge
Awareness

2. Cleanchain 2. Tracer 2. No deeper understanding

3. Supply chain application
4. Socially beneficial
1. Hard to integrate 1. Working on it 1. Hard to integrate
2. Advantages are 2. Absolutely needed 2. Advantages are important
Integration
Blockchain Technology

important 3. Promising technology 3. Existing techs are cheaper

3. Not a priority 4. Open to development and safe
4. Unsure of applications 5. Not come across in textile 4. Has room to be developed
5. Large companies as 5. Could be beneficial for LCA
pioneers
1. Not a priority 1. Not a priority 1. Existing systems (cheaper
2. Hard to integrate & 2. Hard to integrate & apply and safe)
apply 3. - 2. Hard to apply
Barriers

3. Investment costs 4. Structure of textile sector 3. Investment costs

4. Not seen as a need 5. Not widespread/global 4. Lack of qualified employee
concern
6. Lack of regulative force &
incentives

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
173
 CONCLUSION
The primary purpose of this study is to understand the potential of blockchain technology in textile waste
management. Since blockchain technology is a new concept, especially in the textile industry, an
exploratory study supported by qualitative techniques was thought be practical to figure out the concepts
and develop a framework. In-depth interview method with companies from textile sector is adopted and
applied as a qualitative approach. For this, ten companies are determined for interviews and then
appointments for online interviews are arranged. However, three companies out of ten have agreed to
contribute to the study. Data were analyzed with the help of MAXQDA software. As briefly discussed in
the finding parts, three companies are aware of the sustainability concept and the importance of waste
management, but it is seen that the more corporate the companies the more impact they have on environment
and the more initiatives they carry out. For the smaller companies, it is seen that larger partner are the force
behind environmental practices. Large companies have more impact on implementing rules and regulations
as have more demand power in the textile sector. This is what Company X confirms during the interview as
well.
In general, companies have various options to manage their waste. These options include reusing, recycling,
or selling their wastes. The findings of our study show that they do not use a specific software program for
waste management. They follow manual data processing procedures or use a cloud system to organize the
whole processes in tracking waste. However, they believe it would be valuable and easier if all suppliers
and company members could access the data regarding wastes on a common platform. All three firms put
forward common barriers for the waste management practices: lack of qualified personnel, technology,
knowledge, and awareness.
As awareness of blockchain technology, the findings show that all three companies has little knowledge on
blockchain and its details. That means the potential of this technology has yet to be explored and
acknowledged by the companies in the textile sector. Literature review also is preliminary evidence for this
finding as very few studies exist in specific to blockchain-based textile waste management systems, and
they lack implemented and tested results. Therefore, we can deduce that companies are in need of seeing
clear picture how this technology can be implemented in waste tracking and if the advantages exceeds
disadvantages when implemented. However, eager and enthusiasm of interviewed companies in learning
new alternative solutions to existing ones is a promising condition for developing this technology and
diversifying blockchain projects in number and quality. Statement of barriers to the application of
blockchain technology is an obstacle and challenge for developments and need to be eliminated. Training
or educating people on this topic would be a proper first step as the companies concern about qualified
employee deficit. As every technology, blockchain is also expensive and complex at the first phases of
development. As it matures and new applications appear, the technology is expected to be more welcomed
and understood by all actors of the textile sector.
There exist a couple of limitations to this study. These would be listed as: (1) Interviewees gave answers
focused on waste management while answering the perception of sustainability since the participants knew
that the main topic of the interview was waste management. For this reason, the social aspect of
sustainability topic was ignored. (2) The ideal number of participants could be determined when the data is
saturated in qualitative studies. This study is still in progress, and more data will be collected until reaching
maturation. (3) The development in research is fast paced, and as this study being conducted, new literature
continues to emerge. That is why it may lack the latest publications. With all these limitations in
consideration, further studies are needed to be conducted and developed. A blockchain framework design
for textile waste management is especially needed and testing of such a framework would contribute a lot
to the theory and practice in terms of providing deep understanding and clear picture.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
174
 REFERENCES
Ahi, Payman, and Cory Searcy. (2013). “A Comparative Literature Analysis of Definitions for Green and Sustainable
Supply Chain Management.” doi: 10.1016/j.jclepro.2013.02.018.
Ahmad, Raja Wasim, Khaled Salah, Raja Jayaraman, Ibrar Yaqoob, Mohammed Omar, and Samer Ellahham. (2021).
“Blockchain-Based Forward Supply Chain and Waste Management for COVID-19 Medical Equipment and Supplies.”
IEEE Access 9. doi: 10.1109/ACCESS.2021.3066503.
European Union. (2019). Briefing: Environmental impact of the textile and clothing industry. Retrieved March 15,
2021, from www.europarl.europa.eu:
https://www.europarl.europa.eu/RegData/etudes/BRIE/2019/633143/EPRS_BRI(2019)633143_EN.pdf
ÇEVRE KANUNU Kanun Numarası: 2872. (1983). Retrieved May 21, 2021, from www.mevzuat.gov.tr:
https://www.mevzuat.gov.tr/MevzuatMetin/1.5.2872.pdf
Çevre ve Şehircilik Bakanlığı. (2015). Resmî Gazete Sayı: 29314 "ATIK YÖNETİMİ YÖNETMELİĞİ BİRİNCİ
BÖLÜM". Retrieved April 08, 2021, from www.resmigazete.gov.tr:
https://www.resmigazete.gov.tr/eskiler/2015/04/20150402-2.htm
Ellen MacArthur Foundation. (2017). THE CIRCULAR ECONOMY IN DETAIL. Retrieved April 08, 2021, from
archive.ellenmacarthurfoundation.org: https://archive.ellenmacarthurfoundation.org/explore/the-circular-economy-
in-detail
Environmental Protection Agency. (2002). RCRA in Focus - Textile Manufacturing. Retrieved May 21, 2021, from
www.epa.gov: https://www.epa.gov/sites/default/files/2015-01/documents/k02028.pdf
EUROPEAN COMMISSION. (2018). Report on Critical Raw Materials and the Circular Economy. Brussels: SWD.
Retrieved March 15, 2021
European Commission. (n.d.). Circular economy action plan. Retrieved August 14, 2021, from ec.europa.eu:
https://ec.europa.eu/environment/strategy/circular-economy-action-plan_en
European Environment Agency. (2016). Municipal waste management across European countries, Briefing no.
03/2016. Retrieved April 08, 2021, from www.eea.europa.eu: https://www.eea.europa.eu/publications/municipal-
waste-management-across-european-countries/copy_of_municipal-waste-management-across-european-
countries#tab-related-interactive-charts
European Environment Agency. (2021). Industrial waste in Europe. Retrieved 2021, from www.eea.europa.eu:
https://www.eea.europa.eu/data-and-maps/indicators/industrial-waste-indicator/assessment-1
European Environment Agency. (2021). Industrial waste in Europe. Retrieved 2021, from www.eea.europa.eu:
https://www.eea.europa.eu/data-and-maps/indicators/industrial-waste-indicator
European Union. (2014). Environmental Improvement Potential of textiles (IMPRO Textiles). Luxembourg:
Publications Office of the European Union. doi:10.2791/52624
França, A. S. L., J. Amato Neto, R. F. Gonçalves, and C. M. V. B. Almeida. (2020). “Proposing the Use of Blockchain
to Improve the Solid Waste Management in Small Municipalities.” Journal of Cleaner Production 244.
Gopalakrishnan, Praveen Kumare, John Hall, and Sara Behdad. (2020). “A Blockchain-Based Traceability System for
Waste Management in Smart Cities.” in Proceedings of the ASME Design Engineering Technical Conference. Vol. 6.
Grand View Research, Inc. (2021). Textile Market Size, Share & Trends Analysis Report By Raw Material (Wool,
Chemical, Silk), By Product (Natural Fibers, Polyester), By Application (Household, Technical), By Region, And
Segment Forecasts, 2021 - 2028. Grand View Research. doi: Report ID: GVR-1-68038-736-0
Hedstrom, G. (2019). Sustainability: What It Is and How to Measure It. Boston: Walter de Gruyter Inc.
doi:10.1515/9781547400423-202
Kassou, M., S. Bourekkadi, S. Khoulji, K. Slimani, H. Chikri, and M. L. Kerkeb. (2021). “Blockchain-Based Medical
and Water Waste Management Conception.” in E3S Web of Conferences. Vol. 234.
Koligkioni, Athina, Keshav Parajuly, Birgitte Liholt Sørensen, and Ciprian Cimpan. (2018). “Environmental
Assessment of End-of-Life Textiles in Denmark.” Pp. 962–67 in Procedia CIRP. Vol. 69. Elsevier B.V.
Kumar Agrawal, Tarun, Vijay Kumar, Rudrajeet Pal, Lichuan Wang, and Yan Chen. (2021). “Blockchain-Based
Framework for Supply Chain Traceability: A Case Example of Textile and Clothing Industry.” doi:
10.1016/j.cie.2021.107130.
Ongena, Guido, Koen Smit, Jarno Boksebeld, Gerben Adams, Yorin Roelofs, and Pascal Ravesteijn. (2018).
“Blockchain-Based Smart Contracts in Waste Management: A Silver Bullet?” Pp. 345–56 in 31st Bled e-Conference:
Digital Transformation: Meeting the Challenges, BLED 2018. University of Maribor Press/Association for
Information Systems Electronic Library, AISeL.
Roxana MOSTEANU, Narcisa, Alessio Faccia, Al Ansari, Mohamed Dawood SHAMOUT, and Fabian Capitanio.
(2020). “Sustainability Integration in Supply Chain Management through Systematic Literature Review.”
21(176):117–23.
Scott, Ian, Miguel de Castro Neto, and Flavio L. Pinheiro. (2021). “Bringing Trust and Transparency to the Opaque
World of Waste Management with Blockchain: A Polkadot Parathread Application.” SSRN Electronic Journal. doi:
10.2139/ssrn.3825072.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
175
Taeihagh, Araz, Arun Khatri-Chhetri, Katrien Steenmans, Phillip Taylor, and Ine Steenmans. (2020). “Blockchain
Technology for Sustainable Waste Management.” Frontiers in Political Science | www.frontiersin.org 2:590923. doi:
10.3389/fpos.2020.590923.
THE EUROPEAN PARLIAMENT AND THE COUNCIL OF THE EUROPEAN UNION. (2008). Directive
2008/98/EC of the European Parliament and of the Council of 19 November 2008 on waste and repealing certain
Directives (Text with EEA relevance). Retrieved April 21, 2021, from https://eur-lex.europa.eu/legal-
content/EN/TXT/PDF/?uri=CELEX:32008L0098&from=EN
Tiseo, I. (2020). Waste Management in the United States - Statistics & Facts. Retrieved March 02, 2021, from
www.statista.com: https://www.statista.com/topics/2630/waste-management-in-the-united-
states/#topicHeader__wrapper
Tiseo, I. (2021). Global waste generation - statistics & facts. Retrieved 2021, from www.statista.com:
https://www.statista.com/topics/4983/waste-generation-worldwide/#topicHeader__wrapper
TÜİK. (2019). Retrieved from http://www.tuik.gov.tr/PreHaberBultenleri.do?id=24877

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
176
 THE USE OF AUTONOMOUS TECHNOLOGIES IN AİRPORT CHECK-
IN PASSENGER FLOW PRE AND POST COVID-19 PROCESS: THE CASE
OF ANKARA ESENBOGA AIRPORT

Erkam Emin AYVAZ1 , Savaş Selahattin ATEŞ2

Abstract ⎯ Today, air transportation is one of the most intensively used transportation areas for high-speed
travel and logistics. Behind this speed, an unceasing technological development, and change lie. Airline,
airport, and air traffic management systems, which are sub-systems of air transportation, are also encountering
these changes day by day. In this study, it is aimed to examine the pre-and post-COVID-19 service times and
flows of the terminal passenger flow in airport systems, one of the subsystems of air transport by adding
autonomous systems, through different scenarios. That scenarios were prepared according to IATA's LoS
standards and hypothetical performances of autonomous systems. Those scenarios were simulated with the
ARENA simulation program, assuming changing autonomous system usage rates before and after COVID-19.
Also, those simulations show the future effect and importance of autonomous systems on airport terminal flow
by comparing the situation before and after the COVID-19, with the assumption that the use of autonomous
systems will increase in the post-COVID-19 process.
Keywords ⎯ Simulation, Check-in, Autonomous Systems, COVID-19, Airport Terminal Passenger Flow

1 Lect. Erkam Emin AYVAZ, Erciyes University, Faculty of Aeronautics and Astronautics, Department of Aviation Management,
Kayseri, Turkey, [email protected]. ORCID: 0000-0001-8213-5117
2 Assoc. Prof. Dr. Savaş Selahattin ATEŞ, Eskişehir Technical University, Faculty of Aeronautics and Astronautics, Department of

Aviation Management, Eskişehir, Turkey, [email protected]. ORCID: 0000-0003-2462-0039

177
 INTRODUCTION
The air transport system has shown tremendous growth over the past 20 years. This growth in air traffic has
brought with it the need for the development of airports (Kov´Acs, et al. 2012). London Heathrow airport, one
of the most crowded airports in Europe, hosts an average of one hundred and fifty thousand passengers per day.
In such cases, facing various problems such as increasing security threats at airports, delays in connecting
flights, and passengers missing their flights becomes an important problem at airports. (IATA, 2004).
In addition to this growth in today's technological development of air transport, it is seen as a great leap
occurred. The concept of robotics and the internet of things are the concepts that we encounter more in our daily
lives. In this context, an increase has been observed in the number of machines communicating with each other.
Emmerson (2010) estimated in his study that by 2020, approximately 50 million machines will be in
communication with each other. In this context, today, intelligent transportation systems (ITS) appear as
advanced applications that aim to provide innovative services related to different transportation modes and flow
management, without embodying intelligence, and that enable various users to be better informed and become
safer, more coordinated and 'smarter' (Chowdhury, 2017).

The COVID-19 pandemic, which entered our lives in the last months of 2019, has also emerged as a factor
affecting the air transport sector and therefore airport subsystems. COVID-19 disease is an epidemic disease
caused by a newly discovered coronavirus (WHO, 2020). If we examine airports as centers used by people for
transportation services, we can say that COVID-19 disease poses greater risks in such busy centers. In this
context, the most serious measures taken in airport operations since the September 11 attacks, within the scope
of the measures taken by the International Civil Aviation Organization with the DOC 10144 publication, have
begun to enter air transport (Rosario & Patricio, 2020). This study aims to plan a Robotic and IoT integrated
passenger flow system to minimize the queues at the airport in the fight against COVID-19 by using robotics
and IoT technologies more effectively in the developing and ever-growing aviation sector and to simulate how
effective might be these systems are in the post-COVID-19 period.

LITERATURE REVIEW

Although IoT and robotic technologies are encountered in many areas of our lives today, it has been
determined by some researches (Alice, Spencer, etc.) that people approach these technologies with a certain
hesitation.
In their study, Joosse and Evers (2017) interviewed 13 of the passengers, as a result of their experiences, in
the trials they conducted with the participation of 16 passengers with the "Spencer" robot at one of the most
crowded airports in the world. According to the results of the interview, 12 of the 13 passengers who participated
in the study stated that they were satisfied with the work that the airport operators did. In the study, examinations
were made under three main headings and these are the speed of the robot, the target user group, and the general
behavior of the robot, respectively. Regarding the speed of the robot, 3 of the participants stated that they could
catch up with the robot in the non-condensed parts of the terminal building and 5 stated that the robot was very
fast. Some of the participants stated that they did not need the guidance of the robot because they had used the
airport before and knew it. Another issue is stated regarding the general driving and usage of the robot. Since
the robot works entirely through sensors, it often tended to stop when approached. Participants stated that the
robot should be equipped with stronger decision-making skills to serve especially passengers under time
pressure in such areas. On the other hand, they concluded that robots can be used effectively in the description
of various regions for passengers who do not know the airport terminal.
Triebel, et al, (2016) also showed in their study that the Spencer robot can approach and interact with a
person, take a target position, and direct the person or group to the target while following the people below.
When it realizes that passengers are no longer following or detects a failure, it stops and waits for reconsultation.
Encouraged by these results, the platform is planned to be first deployed at Schiphol airport shortly.
In addition, the “Airstar” robot was designed by LG electronics at the Seoul Incheon Airport terminal,
providing consultancy services to passengers, trial studies were carried out (FTE, 2018).

178
 The robot named “Rada” used in the Delphi airport terminal 3 building can provide information about
terminal facilities, departure gates, destination weather conditions, real-time flight status, and the airline's
products and services. Greets customers and interacts with them using basic hand gestures and can move around
the hall in predefined paths (FTE, 2018).
In recent years, Rotterdam The Hague Airport has also tested autonomous baggage handling robots. Their
robots worked behind the scenes, manually holding the bags of passengers dropped off at check-in. The airport
says it will closely follow robotic technologies to test whether robots can deliver baggage faster than current
methods and to search for less damaged baggage (Temblador, 2018)
Kansai Airport has experimented with "Kate", a smart check-in kiosk that can be transported autonomously
to busy areas at the airport if needed. A 1-month trial was conducted at Kansai Airport Terminal 1 building for
the passengers of Air France, Air China, Cathay Pacific Airways, Japan Airlines, Korean Air, KLM, All Nippon
Airways, Asiana Airlines, and United Airlines. The aim of the experiments is stated as detecting the areas with
high density at the airports and making check-in procedures to reduce the densities by moving the autonomous
systems to these areas (Wan, 2018).
The use of autonomous robots, which act according to the airport density zone, for check-in purposes
proposed in this study, shows a similarity with the “Kate” project tested at Kansai airport.

Table 1 Check-in Simulation Modeling Studies in the Literature

Author Purpose Result

Bevilacqua & Improved performance Queuing times are shortened.
Ciarapica, 2010
Reducing the check-in queue Reduction in passenger waiting
Marintseva, 2014
times occurred.
Reducing passenger waiting time Passenger waiting times have
Mota, 2015 and increasing service quality decreased and service quality has
increased.
Maximizing service quality and Passenger service quality is
Al-Sultan, 2018
reducing operational costs optimized with 15 different scenarios.
Bolat & Ateş, Determining the impact of COVID- Check-in queue and service time
2020 19 applications on check-in queues increased.

The basic elements of the model proposed and simulated in this study consist of internet technologies,
machine learning, the internet of things, biometric and robotic systems. With the help of these technologies, a
more effective and efficient terminal flow is aimed at the airport terminal.

METHODOLOGY
Airport managements use simulation programs by constructing different scenarios in capacity and flow
calculations (Hafner, 2019; Landau, et al., 2015). In this study, two different scenarios are focused on how
autonomous systems that can be used at Ankara Esenboga Airport affect the check-in flow. The hypothetical
values were simulated with the ARENA simulation program.
In the study, autonomous check-in devices at different rates and the flows and queue formations of
passengers with different characteristics were examined with the ARENA simulation program. In addition, they
are used in the simulation assuming that autonomous check-in devices can reduce queue formations due to their
high processing speed and mobility.
This study aims to reveal the change in airport check-in queues and processing times due to autonomous
check-in devices. In this context, it is assumed that autonomous systems are performed faster than face-to-face
check-in.

179
 Assumptions about check-in desks in the study:
• It is assumed that the waiting times at the check-in desks are distributed according to a
triangular distribution.
• The use of check-in desks is assumed to be 60% in pre-COVID-19 check-in procedures.
• The use of check-in desks after COVID-19 is assumed to be 30%.
• The rate of passengers who do not deliver luggage and do online check-in is assumed
to be 10%.
Assumptions about autonomous systems in the study:
• It is assumed that autonomous systems process faster than normal check-in desks and
self-service check-in systems.
• It is assumed that waiting times in autonomous systems are distributed according to a
triangular distribution.
• It shows a triangular distribution in the form of waiting times (1,2,4) in autonomous
systems.
• The rate of use of autonomous systems before COVID-19 is assumed to be 30%.
• The rate of use of autonomous systems after COVID-19 is assumed to be 60%.
• The rate of passengers who do not deliver luggage and do online check-in is assumed
to be 10%.
Accompanied by these assumptions, the following hypotheses were created and tested in the
study:
H1: In the post-COVID-19 period, autonomous systems reduce the processing time per passenger in airport
check-in.
H2: In the post-COVID-19 period, autonomous systems reduce the total waiting time of passengers in
queues at airport check-in and check-in.

SIMULATION MODEL
Ankara Esenboga Airport check-in operations were simulated with different usage rates in the pre-and post-
COVID period within the scope of autonomous devices and check-in desks. Within the scope of current IATA
data, airport check-in procedures were simulated by the ARENA simulation program (Figure 1 and Figure 2).
Check-in processes were examined in 2 scenarios as pre-COVID and post-COVID. In the first scenario, a
pre-COVID intensive face-to-face check-in scenario is simulated. In the second scenario, it is thought that face-
to-face check-in transactions, which are supposed to decrease in the post-COVID period, will be carried out
more intensively by contactless autonomous systems.
According to IATA LoS standards, it is stated that the transactions at the check-in desks take between 1-10
minutes in pre-COVID check-in processes. On the other hand, it has been stated that autonomous systems
perform check-in processes between 0-5 minutes (IATA, 2014). In the scenarios, it is assumed that the number
of autonomous systems, check-in desks, and personnel used in the pre-and post-COVID period does not change.
The results from the scenarios are detailed in the findings and analysis section.

180
 Figure 1. Pre-COVID-19 Check-in Process Simulation

Figure 2. Post-COVID-19 Check-in Process Simulation

FINDINGS AND ANALYSIS

In the study, the number of passengers was determined as 1000 in both scenarios. Changes in the number
of passengers were ignored and a work schedule was determined by minimizing the departure time of the
personnel from the check-in desks. The data obtained from the ARENA simulation program are given below.
Figure 3 shows the check-in process in the pre-COVID period. The average processing time per passenger
is 0.034 minutes, and the maximum waiting time per passenger is 0.082957 minutes. It is seen that the average
waiting time of the passengers is 0.00023151 minutes and the maximum waiting time is 0.05 minutes. When
we look at the total time, it is seen that the passengers spent an average of 0.1414 minutes in the pre-COVID
check-in process, and the passengers who spent the most time in the check-in process in terms of total time
spent 0.39 minutes.
Figure 4 shows the check-in process in the post-COVID period. The average processing time per passenger
is 0.029 minutes, and the maximum waiting time per passenger is 0.082288 minutes. It is seen that the average
waiting time of the passengers is 0.00001 minutes and the maximum waiting time is 0.014 minutes. When we
look at the total time, it is seen that the passengers spent an average of 0.0923 minutes in the pre-COVID check-
in process, and the passengers who spent the most time in the check-in process in terms of total time spent 0.38
minutes.
When both scenarios are examined, the increase in the use of autonomous systems with the assumption that
face-to-face transactions have decreased in the post-COVID period has brought about changes in the transaction
and waiting times. When we compare all these data, we see the following results:
• When the check-in process is compared, it is seen that the processing time per passenger
decreased by 14% on average in the post-COVID period,
• The waiting time of the passengers waiting the most during the transaction was reduced by 1%,
• The average total waiting time of passengers decreased by 56.7%,
• Maximum waiting times of passengers decreased by 70.8%.

Figure 3. Pre-COVID Check-in Process (Minutes)

181
 Figure 4. Predicted Post-COVID Check-in Process (Minutes)
According to the results obtained from the simulation study, it is seen that the average processing time per
passenger decreased by 14% in the post-COVID-19 period during the check-in process, and the waiting time of
the passengers who waited the most during the process decreased by 1%. With these results, H 1 is accepted.
When the average total waiting time of the passengers and the maximum waiting time in total are examined, the
average total waiting time is 56.7%; In total, it is seen that the maximum waiting times decreased by 70.8%.
According to this result, H2 is also accepted.
CONCLUSION AND DISCUSSION
In the study, 2 scenarios covering Ankara Esenboğa Airport check-in procedures were prepared. The
directing of COVID-19 to technologies that do not require contactless and face-to-face communication has been
effective in the creation of these scenarios. In the analysis part of the study, waiting times in the check-in queue
before and after COVID and performance comparisons of autonomous systems and check-in desks were made.
The change in the usage rates of check-in desks and autonomous systems in the pre-and post-COVID-19
period was assumed during the simulation process. According to these rates, the average and maximum waiting
times of the passengers were calculated and determined.
The study reflects the results of the scenarios created with the foresight that the COVID-19 pandemic will
reduce face-to-face transactions in the future and that the contactless transactions will increase and the contact
will decrease with the idea of guaranteeing human health.
It is clearly seen that thanks to the contactless and faster processing capacities of autonomous systems, it
helps to reduce the waiting time of passengers, and the social distance, which has become important in the
COVID-19 pandemic, prevents the check-in process from prolonging any longer, and check-in processes are
carried out faster and healthier. enables realization. With the increase in the use of autonomous systems, the
average time spent by passengers in check-in procedures is reduced by 56.7%, making it possible to reduce the
number of passengers waiting in a restricted area.
In this study, newly developed autonomous devices with mobility in the airport terminal building are not
included in the study. Considering that autonomous systems that are mobile and perform their operations at
different points where the passenger density of the terminal building is low, will be integrated into airport
terminals in the future, it will be possible to use airport terminal buildings more efficiently with a more effective
distribution of passenger density.
In future studies, it will be useful to obtain data on whether pandemics such as COVID-19 will have a
negative impact on the airport terminal flow by conducting a simulation study in which these autonomous
devices with mobility are integrated with the airport systems and fully address the airport terminal flow.

182
 REFERENCES
Al-Sultan, A. T (2018). Simulation and Optimization for Modeling the Passengers Check-in System at
Airport Terminal. Review of Integrative Business and Economics Research.
Bevilacqua, M., & Ciarapica, F (2010). Analysis of Check-in procedure using simulation: a case study.
IEEE International Conference on Industrial Engineering and Engineering Management.
Bolat, E., & Ateş, S. S (2020). Post COVID-19 Precautions Management in Small-scale Airports:
Evaluation of Check-in Process in Erkilet Airport by Simulation. Journal of Airline and Airport Management,
77-96.
Chowdhury, A. R (2017). IoT and Robotics: A Synergy. PeerJ Preprints.
Emmerson, B (2010). M2M: The internet of 50 billion devices. WinWin Magazine, 19-22.
FTE (2018). Asia leading the charge as airlines and airports embrace customer-facing robots. Future
Travel Experience: Retrieved from https://www.futuretravelexperience.com/2018/10/asia-leading-the-charge-
as-airlines-and-airports-embrace-robots/
Hafner, F. B (2019). Simulation Options for Airport Planning. Washington: National Academy of
Sciences.
IATA (2004). Airport Development Reference Manual 9th Edition. International Air Transport
Association.
IATA (2014). Airport Development Reference Manual. Montreal: IATA.
Joosse, M., & Evers, V (2017). A Guide Robot at the Airport: First Impressions. Vienna.
Kov´acs, G., Harmati, I., Kiss, B., V´amos, G., & Mar´aczy, P (2012). Comparison of Methods for
Passenger Flow Simulation of an Airport Terminal. Proceedings of the 5th WSEAS congress on Applied
Computing conference, and Proceedings of the 1st international conference on Biologically Inspired
Computation, (s. 145-150).
Landau, S., Weisbrod, G., Gosling, G., Williges, C., Pumphrey, M., & Fowler, M (2015). Passenger
Value of Time, Benefit-Cost Analysis and Airport Capital Investment Decisions. Washington: National
Academy of Sciences.
Marintseva, K (2014). Comparative Analysis of Check-in Technologies At The Airport. Vìsnik
Nacìonal nogo ʹAvìacìjnogo Unìversitetu, 97-104.
Mota, M. M (2015). Check-in allocation improvements through the use of a simulation-optimization
approach. Transportation Research Part A, 320-335.
Rosario, M., & Patricio, N (2020). COVID-19 The Potential Impact on Airport Terminal Design. Scott
Brownrigg.
Temblador, A (2018). Robots to Take Baggage to Aircrafts at Rotterdam Airport. Travel Pulse:
Retrieved from https://www.travelpulse.com/news/airlines/robots-to-take-baggage-to-aircrafts-at-rotterdam-
airport.html
Triebel, R., Arras, K., Alami, R., Beyer, L., Breuers, S., Chatila, R., . . . Kucner, T (2016). SPENCER:
A Socially Aware Service Robot for Passenger Guidance and Help in Busy Airports.
Wan, C (2018). The Wrap: Meet Kate, The Intelligent Check-in Kiosk. WebInTravel: Retrieved from
https://www.webintravel.com/the-wrap-meet-kate-the-intelligent-check-in-kiosk/
WHO (2020). Coronavirus. Coronavirus: Retrieved from https://www.who.int/health-
topics/coronavirus#tab=tab_1

183
 THE NEXUS BETWEEN THE LOGISTICS SECTOR AND CLIMATE
CHANGE: AN EMPIRICAL STUDY FOR NEWLY INDUSTRIALIZED
COUNTRIES

Aslı Özpolat 1

Abstract- In this study, the impact of the logistics sector on climate change will be examined. Especially in the
context of international trade, the transportation models used in the logistics sector are very effective on climate
change. The Logistics performance index published by the World Bank contains important clues in this context.
When the relationship between CO2 emission and logistics performance index is investigated, it is seen that
there is a positive relationship between the variables. For this purpose, the factors affecting the CO2 emission
in the logistics sector has been analyzed. As the dependent variable, co2 emissions in the transportation sector
has been used. The independent variables are GDP per capita, industrialization, trade openness, urbanization
and energy use. Panel cointegration analysis between variables has been estimated between 1987-2018.
Parameter estimations has been analyzed with MG analysis. When the results obtained are evaluated, a positive
and significant relationship between energy use, per capita income and industrialization, and a negative and
significant relationship between trade openness and urbanization has been found.

Key Words: causality, logistics, panel data analysis, sectorial CO2 emissions,

1. Introduction

The global climate crisis has begun to be felt more than ever before. While temperatures are well above
normal in many parts of the world, seasonal changes have brought many disasters such as floods and
fires. As a result of the global climate crisis, which expresses the most important result of economic
activities, factors affecting human health emerge and the future of the ecosystem is opened to
discussion. According to the Global Warming 1.5 C0 (IPPC, 2019) report, the global temperature is
expected to increase by approximately 1.50 until 2030. The magnitude and effects of the temperature
increase may vary according to geographical location, level of development, and climatic vulnerability.
However, impacts are being felt in many land and ocean system. According to the Climate Change
2021 (IPCC, 2021) report, the temperature increase between 2001 and 2020 is 1,090 higher than the
increase between 1850-1900. When climate change is evaluated in terms of CO2 emissions, according
to the International Energy Agency's “Global Energy Review” report (2021), CO2 emissions decreased
by 5.8% with the effect of the pandemic. However, this rate is expected to increase by 4.8% in 2021.
When the increase in emissions is evaluated by regions, the increase in emissions in developing
countries is more than two-thirds of the increase in global emissions. In developed countries, the
increase is expected to be 4%. Next-11 countries, on the other hand, support their economic growth
with technologies with high industrial density and low energy efficiency. (Shahbaz et al.. 2016:34).
According to the data of the International Energy Agency (2021), the highest CO2 emission per capita
among the Next-11 countries in 2019 belongs to South Korea. Considering that the CO2 emission per
capita in the world in 2019 is 4.4 metric tons, it is seen that the CO2 emission of South Korea is very
high with a value of 11.3 metric tons. Other countries are as follows: Bangladesh,0.5; Iran, 7.1;
Pakistan, 0.9; Egypt, 2.3; Mexico, 3.6; Philippines, 1.2; Vietnam, 2.4; Indonesia, 2; Nigeria, 0.5 and
Turkey, 4.5. Since the total GDP in the Next-11 countries is expected to be 2-3 times the total GDP in
the G7 countries in 2050 (Shahbaz et al. 2016:35), it is important to determine the factors that
determine the environmental quality of the Next-11 countries. Figure 1 shows the change in CO2
emissions by years.

1
Aslı Özpolat, Gaziantep University, Oğuzeli Vocational School of Higher Education, Management and Organization Department,
Gaziantep, Turkey, [email protected]

184
Figure 1: CO2 emission in global scale (IEA, 2021)

When the change between 1990 and 2021 is examined, it is seen that the pandemic experienced in
2020 decreased the emission, but the emission increased rapidly in other periods. Sectoral emission
values are given in Figure 2.

Figure 2: CO2 Emissions for Sectoral (IEA, 2021)

When the values are examined, electricity and heating are the sectors with the highest emissions, while
the transportation sector is the second sector with the highest emissions. Table 1 includes the logistics
performance indices.
Table 1: Logistics performance Index (World Bank Database, 2021)

Country Name 2007 2010 2012 2014 2016 2018

Bangladesh 2.47 2.74 .. 2.56334 2.663902 2.58
Egypt, Arab Rep. 2.37 2.61 2.98 2.965911 3.184964 2.82
Indonesia 3.01 2.76 2.94 3.081806 2.984537 3.15

185
 Iran, Islamic Rep. 2.51 2.57 2.49 .. 2.601254 2.85
Korea, Rep. 3.52 3.64 3.7 3.666238 3.717126 3.61
Mexico 2.87 3.05 3.06 3.128319 3.114031 3.05
Pakistan 2.62 2.53 2.83 2.825426 2.923219 2.42
Philippines 2.69 3.14 3.02 3.003844 2.856259 2.9
Nigeria 2.4 2.59 2.45 2.808778 2.62809 2.53
Turkey 3.15 3.22 3.51 3.500225 3.423693 3.15
Vietnam 2.89 2.96 3 3.154763 2.976629 3.27

The index is between 1 and 5. Index includes 6 components: customs, infrastructure, international
shipments, logistics competence, tracking & tracing and timeliness. According to the information in
the table, while Korea has the highest performance index, Pakistan has the lowest index. At the world
ranking, Germany is in the first place with 4.19 points, while the Netherlands is in the second place.
Accordingly, the ranking of the next-11 countries among 167 countries is as follows. Bangladesh,100;
Egypt,60; Indonesia, 51; Iran, 81; Korea, 23; Mexico, 53; Pakistan, 95; Philippines, 64; Nigeria, 103;
Turkey, 37; Vietnam, 45. The CO2 emissions of the countries by years are given in Table 2.

Table 2: CO2 emissions per capita ( World Bank Database, 2021)

Country Name 2007 2010 2012 2014 2016 2018

Bangladesh 0.266297 0.34274 0.384025 0.412899 0.466776 0.512837
Egypt, Arab Rep. 2.32935 2.3743 2.457816 2.395917 2.477155 2.502042
Indonesia 1.635207 1.724074 1.959133 1.9208 1.892441 2.178462
Iran, Islamic Rep. 7.341826 7.468427 7.56051 7.883095 7.613494 7.693013
Korea, Rep. 10.3012 11.58854 11.93151 11.54756 12.00325 12.22459
Mexico 4.235668 4.110157 4.099028 3.808063 3.885809 3.741478
Pakistan 0.871591 0.783727 0.767567 0.786821 0.886749 0.98182
Philippines 0.833953 0.889357 0.905849 1.03877 1.205049 1.333691
Nigeria 0.615895 0.572417 0.602049 0.658712 0.583028 0.66711
Turkey 4.152951 4.10801 4.405565 4.410763 4.692472 5.015418
Vietnam 1.233675 1.698806 1.639943 1.83789 2.390741 2.698806

When the data in the table are examined comparatively, it shows that as the logistics performance
increases, the CO2 emission also increases. However, there is not enough data to estimate the
relationship between the logistic performance index and CO2 emissions. Therefore, it is not possible
to reveal a statistically significant relationship.
However, in the light of the specified data and information, the determination of CO2 emissions for
all sectors, as well as sectoral examination, includes elements worth investigating in terms of
preventing environmental degradation. For this purpose, the factors that determine carbon emissions
in the transport sector for the Next-11 countries will be examined. In the first part, there is a general
literature review, while in the second part, analysis and findings are included. In the last section, there
are conclusions and discussions.

186
2. Literature Review

In the literature, studies on sectoral CO2 emissions are quite limited, but studies for the logistics sector
are less. Studies in this area include Alises et al., 2014; Loo & Banister, 2016; Achour & Belloumi,
2016; Engo, 2019; Liu and Feng, 2020; Pham and Nguyen, 2021; Wang et al., 2020; Zhu et al., 2020
can be cited as an example. However, considering the relationship of the logistics sector with other
sectors, its impact on environmental factors includes remarkable findings. Giannakis et. al. (2020), the
increase in final demand of several sectors where the increase in CO2 emissions is high is directly
related to land transport. Therefore, the analysis of this sector is important. Li, et.al. (2021) states that
the logistics industry has three main determinant emission values. These determinants are: fuel
emission factors, vehicle fuel economy, and vehicle miles traveled. In order to control the emission
value, the decrease in the emission of these determinants is not considered sufficient. Therefore, it is
stated that measures such as a carbon tax or the development of an emission trading system should be
taken to reduce the emission value. In addition, it seems that the measures already taken are not
sufficient. For example, Giannakis et. al.. (2020), in their study for Cyprus, concluded that CO2
emissions in land transportation will increase by 24% in 2030 compared to 2016. Parker (2021) found
that there is a strong decoupling from GDP to CO2, while there is a weak decoupling from energy
intensity and urbanization to CO2 emissions for Vietnam between 1995 and 2017. Ulker and etc.
(2021) stated that RFLs (ro-ro and ferry lines) investments should increase when the increase in co2
emissions is taken into account for the Marmara region. They also drew attention to the consideration
of economic cost benefit analysis in environmental investments. Aslan et al. (2018) investigated the
relationship between sectoral CO2 emissions and economic growth for the UK between 1973 and 2015
using the rolling window estimation method. According to the results of the study, it was concluded
that the environmental Kuznets curve (EKC) hypothesis is not valid in the transportation sector.
Congregado et al. (2016) examined the validity of the EKC hypothesis for the transportation sector in
the USA from 1973:1 to 2015:2. According to the results, while the EKC hypothesis was valid in the
transportation sector, the existence of the EKC hypothesis in the industrial sector could not be
determined. Jebli and Hadhri (2018), on the other hand, obtained the result that energy use and
economic growth increase CO2 emissions in their studies for the years 1995-2013.

3. Model and Empirical Results

A model is determine the CO2 emission and the factors determining the emission in the logistics sector between
1987 and 2018 in the next-11 (Bangladesh, Egypt, Iran, Pakistan, Philippines, South Korea, Indonesia, Nigeria,
Turkey, Mexico and Vietnam) countries. While sectoral CO2 emission is the dependent variable in the model,
the independent variables are GDP per capita (real, US$ 2010), trade openness (% share in GDP), industrial
sector added value (% share in GDP), energy use (per capita) and urbanization. (% share of urban population in
total population).
InCOi ,t  0  1InYi ,t  2 InTRi ,t  3 InURBi ,t  4 InINDi ,t  5 InEU   i ,t (1)

In the model, t, i and denote the time period, cross section and error term, respectively. InCOi ,t also expresses
the natural logarithm of CO2 emissions in the Model and InYi ,t expresses the natural logarithm of GDP per
capita. InTRi ,t is trade openness, InURBi ,t is urbanization, InINDi ,t is industrial sector added value and
InENUi ,t is energy use as independent variables. CO2 emissions has been obtained from the IEA and other
variables were obtained from the World Bank Database. In the analysis, first of all, Im, Pesaran, Shin (2013)
unit root analysis of the variables was performed. The results are in Table 3.

187
Table 3. Unit Root test

Variable IPS

Level First Difference

Constant Constant &Trend Constant Constant &Trend

INCO -0.86507 -1.302* -9.13732*** -7.82050***

INEU -0.12905 1.74058 -6.11341 -5.25383***

ININD 3.15404 0.98555 -7.01560 -6.12292***

INTR -0.81027 -1.52907* -5.69311*** -3.68280***

INURB 0.62523 -3.10179** -5.63576*** -4.39936***

INY 5.14518 0.90110 -5.46270*** -5.78119***

Note: probability; *, 10%; **, %5; ***,%1

According to the results in Table 3, all variables are stationary at first difference. That is, all variables are
cointegrated at the I(1) level. After this stage, the long-term relationship between CO2 emissions, GDP per
capita, energy use, trade openness, urbanization and industrialization has been investigated. The existence of a
long-run relationship has been estimated by Pedroni (1999) and Kao (1999) cointegration analyses. The results
are given in Table 4.

Table 4. Results for Panel Cointegration

Test
Value Prob
Pedroni Cointegration -2.801464 0.0025***
Kao Cointegration -2.504356 0.0061***

Note: probability; *, 10%; **, %5; ***,%1

According to the results obtained, both tests support the existence of cointegration. Finally, MG cointegration
estimator was used in the study.
Table 5. Results for cointegration estimator
Variables MG
INEU 0.0427 (0.0334)**
ININD 0.1333 (0.0966)*
INTR -0.121561 (0.000)***
INURB -1.6699 (0.000)***
INY 0.2200 (0.0290)**
Note: Probability; *,10%; **,%5; ***,%1

According to the results obtained, there is a positive and significant relationship between energy use,
industrialization and per capita income and logistics CO2 emissions. On the other hand, the existence of a
negative relationship between urbanization and openness and logistics CO2 emissions was obtained. Hossain
(2011) states that openness and urbanization have a negative effect on CO2 emissions, meaning that
environmental quality is a normal good. The reasons for the effect of reducing CO2 emissions in the
transportation sector may be due to the low density in this sector.

6. Conclusion
The main purpose of the study is to investigate the variables that determine sectoral CO2 emissions and to test
the validity of the sectoral EKC hypothesis. For this purpose, the Pedroni and Kao cointegration tests between

188
CO2 emissions in the logistics sector and GDP per capita, energy use, openness, urbanization and
industrialization between 1987-2018 for the next-11 countries were analyzed. After determining the existence
of a long-term relationship, the long-term parameter coefficients were estimated with the MG cointegration
estimator. While energy use, per capita income and industrialization increase CO2 emissions, openness and
urbanization reduce CO2 emissions. When the results of the study are evaluated, it is seen that it is necessary to
establish sectoral environmental policies. In addition, the determination of environmental taxes by sectors and
the use of green energy technologies are important in terms of reducing and controlling CO2 emissions. The
creation of transportation modes by the logistics sector by making cost-benefit analysis, increasing RFLs
transportation systems, and dissemination of practices such as carbon tax can have an emission-reducing effect.

References
Alises, A., Vassallo, J. M., & Guzman, A. F. (2014). Road freight transport decoupling: A comparative
analysis between the United Kingdom and Spain. Transport Policy, 32, 186–193
Achour, H. and Belloumi, M. (2016). Decomposing the influencing factors of energy consumption in
Tunisian transportation sector using the LMDI method. Transport Policy, 52, 64–71.
Aslan, A. Destek M.A. ve Okumus, I. (2018). Sectoral carbon emissions and economic growth in the US:
Further evidence from rolling window estimation method. Journal of Cleaner Production 200(2018),
ss:402-411
Congregado, E., Feria-Gallardo, J., Golpe, A. A. ve Iglesias, J. (2016). The environmental Kuznets curve ve
CO2 emissions in the USA. Environmental Science and Pollution Research, 23(18), ss:18407-18420.
Engo, J. (2019). Decoupling analysis of CO2 emissions from transport sector in Cameroon. Sustainable Cities
and Society, 51, 101732
Giannakis, E. , Serghides, D., Dimitriou, S. and Zittis, G (2020), Land transport CO2 emissions and climate
change: evidence from Cyprus, International Journal of Sustainable Energy, 39:7, 634-647
IEA, (2021), Global Energy Review. Retrieved from Global Energy Review 2021 – Analysis - IEA.
24.08.2021
IEA (2021). Database. Retrieved from Data overview - IEA. 23.08.2021
Im, K. S., Pesaran, H. And Shin, Y. (2013). Testing for unit roots in heterogeneous panels. Testingfor unit roots
in heterogeneous panels
IPPC, (2019). Global Warming 1.5 Co. Retrieved from Global Warming of 1.5 ºC — (ipcc.ch). 20.07.2021
IPPC, (2021). Climate Change 2021. Retrieved from Sixth Assessment Report (ipcc.ch). 20.07.2021
Jebli, M. B. and Hadhri, W. (2018) The dynamic causal links between CO2 emissions from transport, real
GDP, energy use and international tourism, International Journal of Sustainable Development &
World Ecology, 25:6, 568-577
Hossain, S. Md. (2011). Panel estimation forCO2 emissions, energy consumption, economic growth,
trade openness and urbanization of newly industrialized countries. Energy Policy 39, 6991-6999

Kao, C., (1999). Spurious regression and residual-based tests for cointegration in panel data. J. Economet. 90,
1–44.
Li, W. Wang, L. Li, Y and Liu, B. (2021). A blockchain-based emissions trading system for the
road transport sector: policy design and evaluation. Clımate Policy 21( 3), 337–352
Liu, Y.,and Feng, C. (2020). Decouple transport CO2 emissions from China’s economic expansion:
A temporal-spatial analysis. Transportation Research Part D: Transport and Environment, 79,
102225

189
Loo, B. P. Y. and Banister, D. (2016). Decoupling transport from economic growth: Extending the
debate to include environmental and social externalities. Journal of Transport Geography, 57,
134–144.
Parker S. (2021): A decoupling analysis of transport CO2 emissions from economic growth: Evidence from
Vietnam, International Journal of Sustainable Transportation
Pedroni, P., (1999). Critical values for cointegration tests in heterogeneous panels with multiple regressors.
Oxford Bull. Econ. Stat. 61(Special Issue), 653–670.
Pham, H. T., and Nguyen, T. A. T. (2021). Evaluation of energy intensity of transport service sectors in
Vietnam. Environmental Science and Pollution Research, 28, 11860–11868
Shahbaz, M. Mahalik, M.K, Shah, H.S. ve Sato, J.R. (2016). Time-varying analysis of CO2 emissions, energy
consumption, and economic growth nexus: Statistical experience in next 11 countries. Energy Policy,
98,, ss:3-48
Ülker, D. , Bayırhan, İ., Mersin, K. and Gazioğlu, C. (2021). A comparative CO2 emissions analysis and
mitigation strategies of short-sea shipping and road transport in the Marmara Region, Carbon
Management, 12:1, 1-12
Wang, C., Zhao, Y., Wang, Y., Wood, J., Kim, C. Y., and Li, Y. (2020). Transportation CO2 emission
decoupling: An assessment of the Eurasian logistics corridor. Transportation Research Part D:
Transport and Environment, 86, 102486
World Bank (2021). World Development Report Database. Retrieved from World Development Indicators
| DataBank (worldbank.org). 21.05.2021
Zhu, F., Wu, X.,and Gao, Y. (2020). Decomposition analysis of decoupling freight transport from economic
growth in China. Transportation Research Part D: Transport and Environment, 78, 102201

190
 MAKING THE SHIFT TO A NEW PARADIGM OF LOGISTICS
MANAGEMENT: SPACE LOGISTICS

Enes Demiralay1, Engin Hasan Çopur2, Turan Paksoy3

Abstract ⎯ The recent advances in space technology stimulate the growth rate in global communications,
transportation, media, and technology industries. The sustainability of space operations relies on successful
and effective logistics systems, namely space logistics. Space logistics is defined as the flow management of
material, services, and information to meet the space system lifecycle of ISS, orbital, lunar, or interplanetary
space operations. Experts forecast that satellites will play a significant role in space logistics to manage these
operations in effective and efficient ways. However, there are serious challenges that must be addressed to
ensure the optimal use of the resources. The need to coordinate possible supply lines, develop sustainable
transportation methods, and produce innovative solutions to produce materials capable of resisting the harsh
conditions of space during transportation are some examples of these challenges. This paper aims to explain
why space logistics are so important and how they will be evolved in the future. In addition, this paper does
not introduce the technical and economic issues of satellite logistics only but also some recent technological
advances to deal with these issues. Finally, it also discusses a new and promising opportunity for
international logistics with the help of the means of satellite logistics.

Keywords ⎯ Interplanetary logistics, Orbital logistics, Satellite, Space logistics

1. INTRODUCTION
Since ancient Babylonian times, humans have pursued the dream of exploring space due to the curiosity that is
the essence of humankind. Nearly 65 years ago, the first step into outer space was made by the Soviet Union,
and the first artificial satellite was launched into an elliptical low Earth orbit to be one step ahead of the USA
in the Space Race during the Cold War period. All remarkable achievements and the steady development in
space exploration are due to the long-lasting efforts of the scientists and the researchers to answer difficult
questions of, e.g., how a vehicle can travel out of Earth’s atmosphere or how the communication can be
continuously maintained with the vehicle orbiting in the space. With the developments in space technology,
many challenges have been overcome, and now there is a change in the direction for space research (Cheng et
al., 2009). The unknowns of deep space have attracted the attention of scientists. In addition, living on another
planet has been a never-ending dream of humans. Therefore, it has been one of the favorite stories of many
movies or novels.
Nowadays, the recent achievements of space agencies and private companies convince people that the thought
of living on another planet is not imaginary anymore, and Mars may become the first planet where humans
will colonize in the future. However, Mars will not welcome humans to their new homes since it does not
offer habitats for people to live in. To make Mars inhabitable for people, researchers are currently
investigating new ways to ensure life support on Mars. The general opinion among the researchers is that the
supply lines from Earth to Mars will play an essential role in transporting the goods and information needed to
establish the first colonization successfully. This issue can be addressed by developing a new logistics support
system in the field of space logistics.
The term “space logistics” is not new for the researchers interested in transporting life support from the Earth
to the international space station (ISS), but it has been used more often in the last decade due to the enormous
efforts of both public and private space industries for newly emerging applications in space. In literature,

1
Enes Demiralay, Konya Technical University, Faculty of Engineering and Natural Sciences, Department of Industrial Engineering,
Konya, Turkey, [email protected]
2
Engin Hasan Çopur, Necmettin Erbakan University, Faculty of Aeronautics and Astronautics, Department of Astronautical
Engineering, Konya, Turkey, [email protected]
3
Turan Paksoy, Necmettin Erbakan University, Faculty of Aeronautics and Astronautics, Department of Aviation Management,
Konya, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
191
several different definitions have already been proposed for space logistics. One definition says that space
logistics is “the movement, storage, and tracking of all crew and equipment necessary to carry out a space
exploration campaign” (Gralla et al., 2006; Shull et al., 2006). In another study, an alternative definition is
proposed that space logistics is “the science of planning, creating, and sustaining effective and affordable
infrastructure, supply chain management, and maintenance to support space operations throughout the solar
system” (Herdy & Yañez, 2009). However, the AIAA Space Logistics Technical Committee gives the most
widely used definition in an easily comprehensible manner. According to the committee, space logistics is
“the science of planning and carrying out the movement of humans and material to, from and within space
combined with the ability to maintain human and robotics operations within space” (AIAA SLTC, 2004).
Based on the definitions above, the main aim of space logistics is to develop an efficient, reliable, sustainable,
and safe transportation system for providing the sufficient equipment, information, and human resources
required for ensuring the performance, reliability, and flexibility of the space operations. The success of space
logistics is dependent upon the accuracy of the planning, control, and implementation of the activities
corresponding to the flow of goods and human resources. For example, it is critical to plan when and how the
required goods should be transported from the origin to the destination efficiently and safely. When space
operations are evolved into more complex activities due to deep space exploration or space colonization, the
current logistics system must be completely overhauled. To define the necessary changes in the current space
logistics system due to the complexity of the future space operations, the current framework of the space
logistics system in use must be first reviewed. Then, it must be extended to satisfy future space operations.
Therefore, in the next section, the current and future frameworks of space logistics will be outlined.

2. FRAMEWORK
Since the early years of the space age, logistics has not been considered a major obstacle to the success of
space missions since the artificial satellites are generally designed to be capable of completing missions
without a need for any additional supplies. In addition, for example, during Apollo 11 spaceflight, all test
equipment and other supportive components, including consumables and spares, were carried in the lunar
module. However, this design method does not offer any flexibility for space missions on the International
Space Station (ISS). The reason for this is that the duration of a space mission on ISS could last for years, and
this means that only one single transaction of logistics is not enough to supply all the amounts of the logistics
needs of the crew and the experiments to ISS (Leach & Ewert, 2021).
Therefore, a special strategy was developed for planning and optimizing the logistics operations of the
required items, as well as tracking the items that are being consumed on ISS. This strategy of the logistics
operations from Earth to the low Earth orbit (LEO), in which ISS is operated, is currently being improved by
the advances in logistics technology. Nowadays, there are a few international companies that are currently
developing reusable rockets not only to provide more efficient and affordable logistics operations from Earth
to Earth orbits (E2O) or from Earth orbits to Earth (O2E) but also to make space tourism and space
colonization possible in a time to come.
In the next five years, it is predicted that the current space logistics network will become more complex due to
the space operations needed to explore the Moon and Mars (Chen, 2021; Evans, 2006). According to NASA
(Dunbar, 2018), Moon, as a natural satellite of Earth, may become the home of some scientific facilities which
will be used to explore the mysteries about our planet, the sun, and the whole universe. In long-term plans, a
landing system will be built to enable the astronauts to be delivered to the Moon. In addition, a new space
station will be launched into the low lunar orbit (LLO) to explore new locations on Moon’s surface as well as
to be used as a space logistics center for refueling and maintenance operations. Moreover, this lunar space
station will serve as a scientific facility to research deep space and samples collected from the Moon.
Therefore, both the scientific and service activities planned to be performed on the Moon’s surface and the
space station in the lunar orbit will also require continuous, safe, and efficient logistics operations to ensure
the sustainability of the activities. To achieve these scientific objectives, a novel supply chain network must be
built in which the facilities on the lunar surface are linked to the space station in lunar orbit through supply
chain routes between the Moon and the lunar orbit, as shown in Figure 1. The route from the Moon to the
lunar orbit is represented by M2O, and the transportation in the reverse direction is represented by O2M. In
the network architecture, the facilities in lunar orbits and Earth orbits are also linked to each other by supply
routes which are represented by O2O.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

192
Recently, researchers attempt to build new supply chain models that enable an effective interplanetary
logistics system for the human exploration of Mars and deep space. A schematic framework of a future space
logistics network is given in Figure 1, where ELS, MLS, and MRLS denote the landing spaces on Earth,
Moon, and Mars, respectively. Besides, the transfer shipment point and low Mars orbit are represented by TSP
and LMO, respectively. In addition, the arrow lines show the direction of the flow of the items required for
space missions. The transfer shipment point is usually considered a Lagrange point that lies between two large
bodies and at which the gravitational forces exerted by these bodies are balanced. Due to that, low-cost energy
consumption requires to access the lunar and Earth orbits, and this minimal energy requirement also provides
low-cost transportation (Jagannatha, 2018; Kakoi, 2015; Li & Zheng, 2010). Therefore, the most common
Lagrange point that can serve as a TSP is the Earth-Moon L1 point (Ishimatsu et al., 2016).

Figure 1. A Schematic Framework of The Future Space Logistics Network Architecture.

According to the National Spaceport Network Development Plan prepared for the Federal Aviation
Administration (Global Spaceport Alliance, 2020), U.S. Government should aim for point-to-point
transportation through space since it is believed that long-distance transportation will become faster with the
help of reusable rockets and the countries with this high-level technology could rule the future ear of the
intercontinental logistics system of Earth. To illustrate this complex logistics architecture, the value matrix of
the future supply chains is also given in Table 1 in a more compact and clear form.

Table 1. The value matrix of the future space logistics network

Earth Moon Orbit Mars
E2E
E2O
Transportation of
Transportation of the items
Earth the items between - -
from the sites located on
the sites located on
Earth’s surface to Earth orbits
Earth’s surface
M2M M2O
Transportation of the Transportation of the items
Moon - items between the sites from the sites located on -
located on Moon’s Moon’s surface to the lunar
surface orbits
O2E
O2M O2O O2MR
Transportation of
Transportation of the Transportation of the items Transportation of the
the items from the
Orbit items from the lunar between the celestial orbits, items from Mars orbits to
Earth orbits to the
orbits to the sites located lunar orbits and Lagrange the sites located on
sites located on
on Moon’s surface points Mars’s surface
Earth’s surface
MR2O MR2MR
Transportation of the items Transportation of the
Mars - -
from the sites located on items between the sites
Mars’s surface to Mars orbits located on Mars’s surface

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

193
 3. SAMPLE CASES OF SPACE LOGISTICS
The era of space exploration began in 1957 with Sputnik 1, the first satellite launched into space by Soviet
Russia. People's dreams of space exploration, which started with Yuri Gagarin's launch into space with the
Vostok spacecraft in 1961, came to an end in 1972, with no attempt made after Apollo 17 made the last
manned landing on the Moon. But with the technological advances and achievements in the last decade, the
vision of inhabiting and colonizing space has slowly been reawakened. With the advent of satellite
constellations and habitable bases, space logistics has become an emerging niche in the industry. While the
essence of logistics remains the same, whether in space or on Earth, stricter restrictions and extreme
conditions are forcing the safe transport, storage, and delivery of materials and products beyond and behind
the Earth's atmosphere. In this context, private and public companies are working to transport materials,
products, and even people safely beyond the atmosphere. This section gives information about private
companies currently engaged in space logistics or conducting R&D studies on space logistics and their
logistics activities.
3.1. Space X
SpaceX is a space logistics company established in 2002 to reduce the cost of space transportation to enable
the colonization of Mars. Since 2012, cargo transportation from the world to the ISS and from the ISS to the
world has been carried out by SpaceX. With the spacecraft named Dragon Spacecraft, it has been possible to
transport 6000 kg of cargo from the world to the space station and 3000 kg from the space station to the world.
Dragon was originally developed to carry passengers. It is designed to carry both NASA and commercial
astronauts to targets in low-earth orbit, the Moon and beyond. It has the capacity to carry seven passengers. As
of 2020, it started to transport people to the orbital laboratory within the scope of NASA's Commercial Crew
program. Dragon has visited the orbiting laboratory 20 times so far. In order to reduce space access costs, the
first orbital-class reusable rocket Falcon 9 was developed by Space X. Additionally, to carry more payload,
Space X developed the Falcon Heavy operational rocket, consisting of three Falcon 9s with the world's highest
orbital lift, producing more than 5 million pounds of thrust (SpaceX, 2021).
3.2. Starlink
Starlink is a satellite internet constellation developed by Space X. In the project, which started more than ten
years, the cost of designing, constructing, and placing the constellation exceeded $ 10 billion. It is a company
aiming to provide high-speed broadband internet to places where access is not secure or completely unused,
with traditional ground infrastructure. Satellites are being launched into space to distribute the world's most
advanced broadband internet to the whole world. Starlink satellites are 60 times closer to earth than other
satellites, resulting in lower latency and the ability to support services not possible with traditional satellite
internet. At this point, reusable rockets owned by SpaceX company take the stage and enable 60 satellites to
be carried into space simultaneously (Starlink, 2021).
3.3. D-Orbit
D-Orbit is a next-generation space company that offers active solutions throughout a space mission's entire
lifecycle, including mission analysis and design, engineering, manufacturing, integration, testing, launch, and
end-of-life decommissioning. It has the advantage of competing against different space companies thanks to
its ability to shape launch and deployment services according to customer needs. The ION Satellite Carrier
developed by D-Orbit supplies the correct delivery of micro-satellites to the planned orbits. In addition, the D-
Orbit Decommissioning Device (De3) system has been developed to dispose of end-of-life satellites and avoid
possible satellite debris safely (D-Orbit, 2021).
3.4. Northrop Grumman
Northrop Grumman is a multinational aerospace and defense technology company that advises customers
worldwide to solve challenging problems. The Mission Extension Vehicle, a space product of Northrop
Grumman's SpaceLogistics company, can be docked with existing satellites with the aim to provide the thrust
and status control needed to extend the lifespan of satellites, and to help satellite operators gain cost-saving
advantages (Northrop Grumman, 2021).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

194
In addition, it prevents the further growth of space junk in orbit. By means of these products, Northrop
Grumman intends to ensure environmental and economic sustainability in space. Mission Robotic Vehicle of
SpaceLogistics is designed so that is capable of delivering payloads to satellites in orbit. Mission Extension
Pods is another next-generation life extension system developed by SpaceLogistics. The difference of Mission
Extension Pods from the Mission Extension Vehicle is that it needs the Mission Robotic Vehicle to move, and
it is also smaller and cheaper. The U.S. The Defense Advanced Research Projects Agency has selected
SpaceLogistics as its commercial partner for the Robotic Service of Geosynchronous Satellites program.
3.5. DHL
Established in 1964, DHL is a logistics company that currently operates in more than 220 countries. It has
taken its place as one of the world's leading logistics companies with an annual shipment of 1.6 billion
(Deutsche Post DHL Group, 2020). With the technological developments experienced today, the
establishment of space bases, and the vision of colonization in space, DHL intends to add space logistics to its
route. In this direction, the company actively carries out R&D studies, but no information has been shared
with the public about how space logistics are planned. In addition to R&D studies, the following questions
that need to be answered for space logistics to be successful and sustainable in the future have been
determined by the company, and answers are sought for these questions:
• How will supply lines be planned and managed?
• How will the accumulation of space debris be prevented and reversed?
• As launch rates and emissions increase, are there more sustainable ways to transport objects into space?
• How can products be packaged to survive in zero gravity environments exposed to high radiation and
extreme temperatures?

4. CONCLUSION
Humankind has pursued the curiosity of exploring space since ancient times. In ancient times, space
observations began only by observing the movements of celestial bodies with the naked eye. When the
Renaissance period came, Newton's invention of the telescope allowed human beings to examine the celestial
bodies more easily. In 1957, the first artificial satellite Sputnik 1, has been launched into elliptical low earth
orbit to stay ahead of the USA in the space race during the cold war period. Four years later, Yuri Gagarin and
Soviet Russia have been made the first manned flight into space. During the Cold War period, numerous space
explorations were made by both the USA and Soviet Russia. The first manned flight to the lunar surface with
Apollo 11, one of the most important steps of humanity in space exploration, also has been lived with the cold
war period. Difficulties such as raw material shortages, climatic conditions, and environmental problems in
our world encourage private companies to work on space exploration. In recent years, successes have been
achieved, proving that the dreams of human beings such as space tourism and colonization in space are not
just dreams. While these developments are taking place, new questions have emerged, such as how space
logistics will be provided to realize space tourism or colonization, which criteria affect space logistics, if space
logistics is possible with the problems experienced in our world in terms of sustainability, whether space
logistics can be provided safely or not. Private and public space companies are researching the solution to
these problems. In the study, existing space companies working to make space tourism and space logistics
possible, and the developments achieved by the companies were examined. Considering the technology-
related long-term strategies of companies and the predictions about how the logistics industry will tend to
evolve over time, it is expected that space will be a new marketplace for the logistics industry. There are
several sound reasons for making this forecast. The first reason is that the use of satellites will increase in the
near future to provide more efficient and more affordable internet service to the whole earth since
digitalization will become more important for the daily life of consumers. Therefore, to guarantee the
sustainability of the satellite-based internet service, the required logistics operations must be managed in a
safe, uninterrupted, and efficient manner. The second reason is the desire of human beings to colonize a
planet, and the recent developments prove that it is not impossible, although it will not happen anytime soon.
However, it is also suggested that the colonization will depend on the logistic supports provided by Earth for a
while. Therefore, supply chain management will also become very important for this future plan. According to
the National Spaceport Network Development Plan prepared for the Federal Aviation Administration (Global
Spaceport Alliance, 2020), it is aimed to carry out point-to-point logistics in space logistics when the

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

195
sustainability of supply chains is ensured and the number of reusable rockets is at a sufficient level. In this
context, a framework has been developed for the space logistics network to be established in the future.

5. REFERENCES
AIAA SLTC. (2004). Recommended Government Actions to Address Critical U.S. Space Logistics Needs.
Retrieved from https://www.aiaa.org/docs/default-source/uploadedfiles/issues-and-advocacy/policy-
papers/information-papers/2004-information-papers/sltc-paper.pdf?sfvrsn=3a3c716e_0
Chen, H. (2021). Interdisciplinary Space Logistics Optimization Framework for Large-Scale Space
Exploration (Georgia Institute of Technology). Georgia Institute of Technology. Retrieved from
http://hdl.handle.net/1853/64768
Cheng, X., Guo, J., & Cui, N. (2009). Space logistics development and future trend. 2009 IEEE International
Conference on Mechatronics and Automation, ICMA 2009. https://doi.org/10.1109/ICMA.2009.5246133
D-Orbit. (2021). Launch Deployment. Retrieved from https://www.dorbit.space/launch-deployment
Deutsche Post DHL Group. (2020). About DHL. Retrieved from https://www.dpdhl.com/en/about-us.html
Dunbar, B. (2018). Moon to Mars Overview.
Evans, W. A. (2006). Logistics and supply chain management- A space operations enabler. SpaceOps 2006
Conference. https://doi.org/10.2514/6.2006-5852
Global Spaceport Alliance. (2020). National Spaceport Network Development Plan. Retrieved from
https://www.globalspaceportalliance.com/wp-content/uploads/2020/06/National-Spaceport-Network-
Development-Plan.pdf
Gralla, E. L., Shull, S., & De Weck, O. (2006). A modeling framework for interplanetary supply chain.
Collection of Technical Papers - Space 2006 Conference, 1. https://doi.org/10.2514/6.2006-7229
Herdy, R., & Yañez, D. (2009). VdotTM a revolutionary tool for space logistics campaign planning and
simulation. AIAA Space 2009 Conference and Exposition. https://doi.org/10.2514/6.2009-6551
Ishimatsu, T., De Weck, O. L., Hoffman, J. A., & Ohkami, Y. (2016). Generalized multicommodity network
flow model for the earth-moon-mars logistics system. Journal of Spacecraft and Rockets, 53(1).
https://doi.org/10.2514/1.A33235
Jagannatha, B. B. (2018). Space logistics network optimization with embedded propulsion technology
selection (University of Illinois at Urbana-Champaign). University of Illinois at Urbana-Champaign.
Retrieved from http://hdl.handle.net/2142/102845
Kakoi, M. (2015). Design of transfers from Earth-Moon L 1/L2 libration point orbits to a destination object
(Purdue University). Purdue University. Retrieved from
https://engineering.purdue.edu/people/kathleen.howell.1/Publications/Dissertations/2015_Kakoi.pdf
Leach, H., & Ewert, M. (2021). Analysis of Historical International Space Station Logistical Mass Delivery.
IEEE Aerospace Conference.
Li, M., & Zheng, J. (2010). Indirect transfer to the Earth-Moon L1 libration point. Celestial Mechanics and
Dynamical Astronomy, 108(2). https://doi.org/10.1007/s10569-010-9301-7
Northrop Grumman. (2021). Northrop Grumman: Who we are. Retrieved from
https://www.northropgrumman.com/who-we-are/
Shull, S. A., Gralla, E. L., Armar, N., & De Weck, O. (2006). An integrated modeling tool for sustainable
space exploration. AIAA 57th International Astronautical Congress, IAC 2006, 12.
https://doi.org/10.2514/6.iac-06-d3.3.01
SpaceX. (2021). SpaceX. Retrieved from https://www.spacex.com
Starlink. (2021). Starlink.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

196
 BLOCKCHAIN APPLICATIONS FOR TRACEABILITY AND FOOD
SAFETY IN AGRI-FOOD SUPPLY CHAINS

Şebnem İndap1, Mehmet Tanyaş2

Abstract ⎯ Access to adequate amounts of safe and nutritious food is key to healthy living. With globalization,
the number of suppliers is increasing, food systems are becoming more complex. Increasing geographical
distance from producers to consumers in the agri-food supply chain makes it difficult to maintain food quality
and to quickly recall food when necessary. On the other hand, consumers' concerns about food safety and
quality and demands for food supply chain transparency are increasing and traceability has become critical.
Alternative solutions should be developed to better guarantee the reliability and authenticity of food. In this
context, Blockchain Technology offers great potential, it allows data to be stored in an unalterable way, fast
tracking of all movements in the chain, and has great potential in the traceability of food supply chains. In this
study, Blockchain technology and its applications in the agri-food sector are examined with the method of
literature review. Studies focus on different usage areas of Blockchain technology such as food safety,
traceability, sustainability and supply chain management in the agri-food sector. In current academic studies,
no study has been found on the use of Blockchain in the production and certification of fresh fruit-vegetable
agri-food products in Turkey.

Keywords ⎯ Agri-food, Blockchain, Food Safety, Traceability

1. INTRODUCTION
Access to adequate amounts of safe and nutritious food is key to healthy living. Unsafe foods that contain
harmful bacteria, viruses, parasites or chemicals can cause more than 200 different diseases, from diarrhea to
cancer. Worldwide, approximately 600 million people become ill after eating contaminated food each year,
resulting in around 420 thousand deaths. Foodborne diseases can hinder socioeconomic development by
straining health systems and damaging national economies, tourism and trade. Collaboration between
governments, producers and consumers in international food supply chains (FSC) can help ensure food safety
(WHO, 2019).
The basic principles of the European Union's food safety policy are defined in the General Food Law of
the European Union adopted in 2002. It covers all stages of the food chain, from animal feed and food
production to processing, storage, transportation, import and export of food, as well as retail sales. This
integrated approach means that all food and feed produced and sold in the European Union can be traced
from 'farm to fork' and consumers are informed about the content of their own food (European Commission,
2014, p. 5). So far, this demand has been met by verifying easily manipulated documents such as shipping
documents and invoices. In addition, restructuring of supply chains and associated recalls is very difficult and
time consuming, and almost impossible in complex and global trade structures, especially during epidemics.
More than 3,800 people in Germany contracted the gut bacteria, and the disease was fatal. Weeks after
identifying the cause as Spanish cucumbers, the real cause was identified as Egyptian vegetable sprouts.
During this time, more and more people became ill as the primary source could not be identified. As a result,
alternative solutions should be developed to better guarantee the reliability and authenticity of food. In this
context, Blockchain Technology (BCT) offers great potential (Creydt & Fischer, 2019, pp. 45–46).
With globalization, the number of companies and suppliers is increasing, and food systems are becoming
more complex. The increase in the geographical distance from producers to consumers in the FSC also makes
it difficult to maintain food quality and to quickly recall food when necessary. Not only suspicious customers,
but also food companies suffer economically from product recalls (Zhang et al., 2019, p. 2). According to a
2011 Grocery Manufacturers Association report, most companies going through a level one recall, where the
public health impact has the potential to be the most severe, can expect an average of 10 million USD in
financial impact, excluding indirect losses such as brand reputation and other losses in sales (Accenture, 2018,

1
Maltepe University, Logistics and Supply Chain Management PhD Programme, Istanbul, Turkey, [email protected]
2
Maltepe University, International Trade and Logistics Management Department, Istanbul, Turkey, [email protected]

197
p. 7). In a complex food supply chain, an efficient traceability system can make a significant contribution
to the food recall process. Certain products and ingredients can be quickly isolated from the root of the
problem to prevent further loss (Zhang et al., 2019, p. 2).
Traceability has become a critical element in supply chain management, especially in security-sensitive
sectors such as food, pharmaceuticals, etc. European Union regulations oblige food manufacturers to monitor
all raw materials/ingredients used throughout SC operations. The complexity and fragmentation of modern
FSC networks hinders the development of strong traceability mechanisms. Traceability-oriented FSC
management relies on new technologies such as IoT (Internet of things) that provide real-time information
about products and contamination throughout production and distribution (Casino et al., 2019, p. 2728).
The complexity of global FSC and the economic motivation to provide food products more cheaply
contribute to the prevalence of food fraud. It is estimated that this costs the global food industry around 30
billion Euros each year. Food fraud can also pose public health risks. In 2008, milk mixed with melamine
caused more than 50,000 infants to become ill and resulted in 6 deaths in China. Scandals such as horse meat
in beef products in 2013 and fipronil in eggs in 2017 attracted the attention of the whole world (European
Commission, 2019, p. 4). As a result, authorities lose their credibility, companies lose money and consumers
lose trust in food. Figure 1 shows the product categories with the most recall requests, with fish, meat, oil,
alcohol products, and diet foods at the top.
One of the reasons food fraud is suppliers are not afraid of getting caught because how the food is
produced and where it comes from is anonymous. Having a digital, real-time system for tracking food from a
farm to a retailer would be a strong deterrent to such fraudulent activities, as it would create a digital footprint
(Yiannas, 2018, pp. 49–50).
Companies do not have enough information about the products they buy and sell to meet the challenges
faced by today's global supply chains, such as being safe, sustainable and ethical. While some companies
recognize the value of traceability for efficiency, cost savings, and product priority in the market, they must
first overcome the distrust associated with product identity and traceability. Although Blockchain (BC) alone
does not provide traceability, it can change the way business is done. When implemented effectively, it can
connect the links of the supply chain and provide efficiency, transparency and accountability. Better and
reliable data can help optimize business decisions and achieve higher standards for production, efficiency and
sustainability (Accenture, 2018, p. 3).

Figure 1. Top 10 Categories with the Most Recall Requests in 2018 (European Commission, 2019, p. 8).

198
 Food safety regulations are becoming more stringent around the world. For example, the U.S. Food Safety
Modernization Act has established additional record keeping requirements, including a section on tracking
food. This will inevitably raise the bar on minimum expectations for traceability of food in the coming years
(Yiannas, 2018, p. 50).
From a sustainability perspective, increased traceability and transparency will allow food system
participants to optimize their supply chains and reduce food losses. The current estimate is that about a third
of the food produced globally is wasted. Food loss in the USA is approximately 161.1 billion USD each year
(Yiannas, 2018, p. 50). By pairing Blockchain digital identifiers with food products, traceability in supply
chains can be achieved for growth conditions, batch numbers and expiration dates. In this way, consumers can
learn about the ecological footprints of their food, food surplus distribution can be directed and food waste can
be prevented. The inability to change records can enable the identification of the source of foodborne illness
and prevent fraud (Antonucci et al., 2019, p. 6129). Blockchain could improve the ability to link foodborne
outbreaks with food vehicles, helping fewer and smaller outbreaks and fewer people being harmed. It can
enable more efficient analysis and future preventive measures to identify the root cause of epidemics
(Yiannas, 2018, pp. 48–49).
Food safety after Covid-19 will be one of the prominent issues in the coming years. Food safety will
require the development of new areas of expertise and investment in both how food is produced and how it is
supplied. The components of the food we buy and how it is delivered to our tables will gain more importance
(Kalın, 2020). Blockchain technology, which allows data to be stored in an unalterable way and fast tracking
of all movements in the chain, also has great potential in the traceability of FSC.
Blockchain could improve the ability to link foodborne outbreaks with food vehicles, helping fewer and
smaller outbreaks and fewer people being harmed. It can enable more efficient analysis and future preventive
measures to identify the root cause of epidemics (Yiannas, 2018, pp. 48–49).
In the second part of the study, the Literature Review was summarized. In the third part, global and
Turkish agri-food sector, food safety, traceability and transparency in the agri-food industry were examined.
As a result, various suggestions have been made for solving problems in agri-food supply chains with
Blockchain.

2. METHODOLOGY - LITERATURE REVIEW

In the first phase of the literature review, the keywords were determined as “blockchain” or “blockchain
technology” and “food industry” or “agri-food” and “food safety” or “traceability” or “transparency” or
“sustainability” or “IoT” or “big data”.
Academic resources were searched with keywords in Web of Science, Scopus, Science Direct, Google
Scholar and YÖK National Thesis Center databases. Table 1 was created by utilizing the “inclusion and
exclusion criteria” of Zhang et al. (2019).

Table 1. Inclusion and Exclusion Criteria

Inclusion Criteria Exclusion Criteria
Studies on Food or Agri-Food Sector Studies on sectors other than Food or Agri-Food
English or Turkish studies Studies in languages other than English or Turkish
Articles, conference proceedings, theses, reports News, reports other than those published by relevant
published by relevant professional organizations professional organizations, technical studies
2010 and later studies Studies before 2010

A total of 136 studies were accessed, first non-English or non-Turkish sources and then technical-oriented
sources were excluded, and the summaries of 90 sources which were identified were reviewed in terms of
their relevance to the scope of the study.
Among the 90 studies selected at the first stage, due to the innovation of technology, the oldest study was
1 in 2016, 4 in 2017, 27 in 2018, 48 in 2019, and as of March 15, 2020, 10 studies were in 2020. The data in
Figure 2 show that research on blockchain in the food industry tends to gain increasing research interest from
2018.

199
 Figure 2. Types of Research by Years

Articles are mostly published in food industry related journals such as “Food Control”, "Global Food
Security", "Foods", "British Food Journal" and "Journal of the Science of Food and Agriculture" and
technology-related journals such as "International Journal of Information Management", "International Journal
of Information Technology" and "International Journal of Innovative Technology and Exploring
Engineering”.
The 90 studies were categorized according to research topics and methods used, as summarized in Table
2. Studies focus on different usage areas of Blockchain technology in the agri-food sector such as safety,
traceability, sustainability and supply chain management. Also, although it examines the agri-food sector, 14
studies focusing on Blockchain Technology, IoT (Internet of Things) and Big Data technologies were
examined.

Table 2. Research Topics and Methods Used

Methods Used
Reviewed Conceptual Modeling/
Research Topics Case Study/
Sources Literature Questionnaire/ Study/ System Design/
Use Case
Review Interview Conceptual Optimization/
Analysis
Modeling Simulation
BCT, IoT, Big Data 14 5 1 4 5
Food Safety 16 4 4 3 7
FSC Management 25 6 8 4 4 7
Traceability 29 3 11 3 6 10
Sustainability 6 3 2 1 1 2
Total 90 21 25 9 18 31

Of the 90 studies, 29 focused on traceability and 16 on food safety. In terms of the methods used, in 31
studies modeling/ system design/ optimization/ simulation, in 21 studies literature review, in 25 studies case
study/ use case analysis, in 18 studies conceptual modelling and in 9 studies questionnaire/ interviews were
carried out.

200
 In the second phase of the literature review, the abstracts of the selected articles were reviewed for
relevance to the scope of the study. 27 studies were selected to be examined in detail to ensure diversity in
research topics and methods used. Table 3 contains a summary of these 27 studies.

Table 3. Literature Review

No Author, Year Research Type Method Sector

Antonucci et al., 2019 Journal Article Literature Review Agri-Food

Problem A review on blockchain applications in the agri-food sector
Summary The aim of the study is to show scientific studies on the use of Blockchain in the
1 agri-food sector. The reviewed studies focused on software topics such as
architecture and smart contracts. It has been found that Blockchain Technology
is promising and has application potential in various industries, but is still
immature and difficult to implement.
Arena et al., 2019 Conference Paper Simulation Food – Extra Virgin Olive Oil
Problem An IoT Blockchain-based framework for certifying extra virgin olive oil supply
chain
Summary In the study, a Blockchain-based Bruschetta system is presented for product
2
certification by monitoring the entire supply chain of Extra Virgin Olive Oil.
Bruschetta Extra Virgin Olive Oil uses IoT technology to connect quality control
sensors and work on the Blockchain. A performance evaluation based on
simulations was carried out to test the proposed solution.
Bumblauskas et al., Journal Article Use Case Analysis Food- Eggs
2019
Problem A blockchain use case in food distribution
3 Summary The aim of the study is to track products from farm to fork using Blockchain and
IoT enabled technologies at a company in the USA. One of the research
questions is how Blockchain can be used and applied to move goods across
global supply chains more accurately and transparently.
Buttafoco, 2019 Master’s Thesis Case Study Food – Extra Virgin Olive Oil
Problem Food traceability solutions in the EU: Use of Blockchain technology in tracking
extra virgin olive oil
Summary The research question “How and to what extent Blockchain technology can
4 create a sustainable solution to improve the traceability of extra virgin olive oil
within the EU” was addressed. By evaluating the feasibility of Blockchain
technology, a possible alternative to existing applications has been examined. By
focusing on Extra Virgin Olive Oil, gaps and opportunities in existing
traceability systems were identified.
Casino et al., 2019 Journal Article Modeling Food Supply Chain
Problem Modeling food supply chain traceability based on blockchain technology
Summary The study aimed to develop a distributed functional model to provide
decentralized and automated food supply chain traceability based on Blockchain
5 technology and smart contracts. To evaluate the feasibility of the proposed
modeling approach, a food traceability use-scenario is presented. The
applicability of the model is explained by the development of a fully functional
smart contract and a local custom blockchain. The overall benefits of the
proposed model were evaluated based on predefined key performance indicators.
Creydt & Fischer, 2019 Journal Article Literature Review Food
Problem Algorithm-driven food traceability with blockchain
Summary By conducting a literature review, Blockchain application areas and limitations
6 were discussed and alternative strategies were presented. The implementation of
Blockchain technology together with IoT approaches could revolutionize the
food industry. Due to the complexity, it will take some time for the Blockchain
technology to be fully established.

201
No Author, Year Research Type Method Sector

Croxson et al., 2019 Conference Paper Questionnaire, Food Supply Chain

Q-Method
Problem Making Sense of Blockchain in Food Supply-Chains
Summary Q-methodology was used to examine whether Blockchain technology could be
the solution for accurate tracking and tracing of products in New Zealand
7 agricultural industry. Empirical research has revealed 4 distinct groups in the
industry with different Blockchain perspectives and potential. The results
showed that although industry experts believed the implementation of
Blockchain was inevitable and would solve current problems, factors such as
high installation costs and complexity of technology could be barriers. Further
research has been proposed on how to solve food safety problems with new
technologies such as BC, IoT (internet of things) and AI (artificial intelligence).
Galvez et al., 2018 Journal Article Literature Review, Food
Case Study
Problem Future challenges on the use of blockchain for food traceability analysis
8 Summary One way to solve traceability issues and ensure transparency is to use BCT to
store data in chemical analysis in chronological order so that it is impossible to
manipulate later. The potential of BCT to provide traceability and authenticity in
the food supply chain has been examined. Although achieving food traceability
with Blockchain technology seems promising, the limits to consider are outlined.
Gerdan, 2019 Master’s Thesis Modeling, Food – Eggs
Prototype & Interface Design
Problem Food Safety with Blockchain Technology and an Example Application for the
Egg Industry
9 Summary The innovations that BCT will bring to the food supply chain traceability and the
advantages it offers have been researched and the benefits it will bring to food
safety have been examined. An organic egg producer under the name of
Ecological Egg Ltd. was designed to be examined. A digital platform using BCT
called Blockchain Lab. was assumed and a food safety solution prototype was
created for the egg producer and sample user interfaces were designed.
Green, 2018 Master’s Thesis Conceptual Study Agri-Food
Problem Application of blockchain technology to agri-food markets
Summary Potential Blockchain applications in the agri-food market are explored, focusing
10 on food safety and environmental certifications.
It has been identified that the technology has high potential, but there are
obstacles to overcome before reaching this potential, such as changes to
precision farming equipment and other farm machinery, or greater integration of
real-world connections to the Blockchain such as RFID and NFC chips.
Hancock, 2019 Report Questionnaire Food - Beef

Problem The Integration of Blockchain Technology to the Beef Industry-A Comparative

Analysis
Summary It is aimed to investigate Blockchain applications in the beef industry in the USA
and Kenya. Previous studies have not focused on differences between countries,
11 this study aimed to address this gap. The primary research method was applied
as the main method, by interviewing various stakeholders in the industry about
supply chain issues and opportunities. It has been found that the integration of
Blockchain into the beef industry is broadly advisable, but varies greatly
depending on the developmental stage and regulatory environment of industries
at different countries.

202
No Author, Year Research Type Method Sector

Kamble et al., 2019 Journal Article ISM Agriculture

DEMATEL
Problem Modeling the blockchain enabled traceability in agriculture supply chain
12 Summary In this study, the relationships between the factors enabling the adoption of
Blockchain Technology in the agriculture supply chain were determined.
The findings show that among the factors identified, traceability is the most
important reason for the implementation of Blockchain Technology in the
agriculture supply chain, followed by auditability and immutability.
Kamilaris et al., 2018 Report Multiple Case Study Agriculture- Food

Problem The Rise of the Blockchain Technology in Agriculture and Food Supply Chain
Summary This article examines the impact of Blockchain Technology in the agriculture
13 and food supply chain, presents current ongoing projects and initiatives, and
discusses the general implications, challenges and the potential.
The findings show that blockchain is a promising technology for a transparent
food supply chain, but many barriers and challenges still exist that hinder its
wider popularity among farmers and systems.
Kayıkçı & Subramanian, Conference Paper Literature Review, Food-Milk
2018 Modeling,
Questionnaire
Problem Feasibility of Food Loss Reduction with Blockchain in the Emerging Economy
Context
14 Summary This article analyzes the feasibility of reducing food loss and the adaptation
challenges of Blockchain technology.
To understand the barriers, a survey was conducted for the Turkish dairy supply
chain. The results of the research show that barriers such as trust between
businesses, trust in technology, standards, consumer perception, legislation and
regulation are the key to reducing food loss in the supply chain.
Kim & Laskowski, 2017 Report Use Case Analysis Agri-Food
Problem Agriculture on the Blockchain: Sustainable Solutions for Food, Farmers, and
15 Financing
Summary This research includes cases showing how Blockchain, soil sensors, satellite
tracking and drones can increase crop yields, improve crop and soil quality, and
reduce waste along the food chain.
Kim et al., 2018 Conference Paper System Design, Food
Modeling
Problem Integrating Blockchain, Smart Contract-Tokens, and IoT to Design a Food
Traceability Solution
16 Summary The study introduces Harvest Network, a theoretical food traceability application
“from farm to fork” integrating Blockchain and IoT devices. The aim is to create
a distributed ledger accessible to all stakeholders in the supply chain. A basic
framework has been established to build a prototype or simulation using existing
technologies and protocols.
Lezoche vd., 2020 Journal Article Literature Review Agri-Food
Problem A survey of the supply chains and technologies for the future agriculture
Summary More than 100 studies on new technology and supply chain methods have been
17 analyzed and compared to understand the future of agri-food. With the
integration of 4 new technologies focused on - IoT, Blockchain, Big Data and
Artificial Intelligence, it has been seen that smart agriculture targets such as
water saving, soil protection, limiting carbon emissions and increasing
productivity can be achieved.

203
No Author, Year Research Type Method Sector

Musah, 2019 Master’s Thesis Questionnaire Food - Cocoa Bean

Problem Farmers’ awareness of emerging technologies
Summary The aim of the study is to determine the effect of Blockchain Technology in the
18 cocoa bean supply chain. A combination of questionnaire and interview methods
was used. It has been found that Blockchain Technology is highly effective in
cocoa bean food supply chains and contributes positively to reducing widespread
unethical activities through transparency and traceability.
Ndraha et al., 2018 Journal Article Literature Review Food – Cold Chain
Problem Time-temperature abuse in the food cold chain: Review of issues, challenges,
and recommendations
Summary Temperature violations in food cold chains operating in different countries, cold
chain solutions focusing on food quality and safety were reviewed. Food waste
19 can be reduced by better temperature management in food cold chains. In order
to ensure the integrity of the food cold chain, it has been recommended to create
guidelines by the relevant regulatory authorities considering energy use,
economic risk, social impact, environmental issues and sustainability, and it has
been suggested to apply Big Data and Blockchain Technology in the regulation
process.
Sander et al., 2018 Journal Article Questionnaire Food - Meat
Problem The acceptance of blockchain technology in meat traceability and transparency
Summary The aim of the study is to investigate meat traceability by identifying different
20 perspectives and opinions of meat supply chain stakeholders. The acceptance
potential of BCT as a viable transparency and traceability system was evaluated.
Consumer survey and interviews with meat industry supply chain stakeholders
were conducted. The perspectives of a 4-stage supply chain and government and
3pTSPs as an inclusive interest group are discussed.
Tian, 2018 PhD Thesis Modeling, Food - Cold Chain
Simulation,
AHP, FCEM
Problem An information system for food safety monitoring in supply chains based on
HACCP, Blockchain and Internet of Things
Summary By creating a decentralized information system for food supply chains, it is
aimed to guarantee food quality and safety.
21
1) An attempt has been made to establish a new decentralized information
system based on IoT and Blockchain Technology.
2) This information system is integrated with supply chain risk management
methods to create a safety control and traceability system for food supply chains.
3) How the new system improves the performance of food supply chain
management is discussed and the developed Fuzzy-AHP assessment method is
used to conduct a case study in a Chinese agri-food cold chain logistics business.
Unurjargal, 2019 Master’s Thesis Case Study, Food - Banana, Fish,
Conceptual Modeling Mongolian Meat
Problem Blockchain-supported food supply chain reference architecture
Summary The aim of the study is to create a Blockchain supported food supply chain
reference architecture model, with additional guides that can be applied in real-
22 life supply chain cases. A case-oriented bottom-up approach was used to build
the reference architecture with the help of the BOAT framework as the main tool
to organize the case details. 3 food supply chain cases were used in the
development of the reference architecture and the Mongolian meat trade supply
chain case study was examined with the proposed solution and finally evaluated
by local experts.

204
No Author, Year Research Type Method Sector

Yiannas, 2018 Journal Article Case Study, Conceptual Food - Mango & Pork
Evidence Study
Problem A New Era of Food Transparency Powered by Blockchain
23 Summary Walmart and IBM presented 2 PoCs to demonstrate that Blockchain Technology
offers a convenient way to track and verify food from farm to store quickly and
precisely. PoCs are focused on the traceability and authenticity element of the
Blockchain solution. The time taken to view mangoes with Blockchain
Technology decreased from 7 days to 2.2 seconds.
Yıldızbaşı & Üstünyer, Journal Article Conceptual Modeling Agriculture, Food -
2019 Vegetables Fruits
Problem Blockchain design in agri-food supply chain
Summary The aim of the study is to facilitate the inspection and registration system by
integrating vegetables and fruits into the SCM process with BCT without the
24 need for an intermediary, within the scope of the market law implementation in
Turkey's agenda. From the producer to the user, product traceability and
transparency of content information are ensured throughout the process. In
addition, the verification process that BCT provides increases the reliability of
agri-foods. The proposed study is an example that can be developed, as it is
adaptable to different supply chains.
Yılmaz, 2019 Master Thesis Modeling Food Supply Chain
Problem Designing a Blockchain system that will track the delivery of products from the
supplier to the consumer
25 Summary To solve the difficulties experienced in food supply chain management, a
proposal has been made regarding the use of Blockchain Technology.
Information about Ethereum technology and smart contracts and an example of a
working Blockchain is presented.
Zhang et al., 2019 Conference Paper Literature Review Food
Problem A content based literature review on the application of blockchain in food supply
chain management
Summary The aim of the study is to explore how Blockchain is used in the field of food
supply chain and how it can help address food safety issues.
26 In order to answer the following research questions, an in-depth analysis of the
literature was conducted to fill in the gaps:
1) What are the researches on the application of Blockchain in food supply chain
management?
2) What benefits can Blockchain bring to the food supply chain?
3) What are the challenges of implementing Blockchain in the food supply
chain?
Zhao et al., 2019 Article Systematic Literature Network Agriculture-Food
Analysis
Title Blockchain technology in agri-food value chain management: A synthesis of
applications, challenges and future research directions
Summary SLNA was used to review Blockchain Technology, including key applications
and challenges in the agri-food value chain.
27
Blockchain Technology along with IoT, has been shown to improve agri-food
value chain management in 4 main aspects: traceability, information security,
production and sustainable water management.
6 challenges were identified, including high cost, storage capacity and
scalability, privacy leakage, regulation issue, throughput & latency issue and
lack of skills.

205
 3. FINDINGS
3.1. Global Agri-Food Sector
Agri-food trade has more than doubled in real terms since 1995, reaching USD 1.5 trillion in 2018.
Developing countries have become active participants in global markets and account for about one-third of
global trade. Technological advances have made possible the transformation of production and trade
processes, resulting in the emergence of global value chains in food and agriculture. With advances in digital
technology, as farmers and firms find it easier to communicate, they can better coordinate their operations
across borders and become part of global value chains (FAO, 2020, pp. v–x).
It is estimated that the world economic growth will be 2.7% from 2020 to 2029 and approximately 703
million people will be added to the world population that was 7.5 billion in 2019 in the next ten years.
Increasing per capita income and population growth are among the main factors that increase global trade in
agricultural products (USDA, 2020, p. 54).
Agriculture in the world and in Turkey is faced with the following challenges (Accenture, 2017, p. 1;
Aşarkaya, 2015, p. 5; FAO, 2018, p. 6; TARMAKBİR, 2020, pp. 12–14; TTGV, 2021, p. 12):
• Climate change threatens water resources, production areas and food security
• Meeting the food demands of the growing population: The world population is expected to reach 9.6
billion by 2050. Food production will need to double to feed the world's population.
• Limited and decreasing arable land
• Reducing food waste
• Ensuring food safety
• Changes in consumer behavior and diets
• Population aging, migration to cities and decrease in agricultural population
• Reducing supply chain costs
• Maintaining high quality phytosanitary standards
• Maintaining profitable agricultural activities
• Increasing the income of small-scale food producers
Robust, resilient and sustainable agri-food supply chains are difficult to achieve because they involve
more uncertainty and risk than other supply chains (Lezoche et al., 2020, p. 4). Esteso et al. (2018, pp. 4425–
4426) identified four types of product-based uncertainty:
• Product: shelf life, rate of deterioration, lack of homogeneity, food quality and safety
• Process: harvest yield, lead time, resource needs, production
• Market: demand, market prices and
• Environment: weather, pests and diseases, and regulations.
Mismanagement of these sources of uncertainty can have a very negative impact on the safety, quality,
quantity and waste of products, as well as on human, technological and natural resources. New technologies
have a great impact on the reduction of uncertainty as they enable precise data to be obtained in real time.
These technologies, with their capacity to make smart decisions throughout the entire supply chain, from
farmers to customers; will help increase efficiency, sustainability, flexibility, agility and endurance (Lezoche
et al., 2020, p. 4).
In the current agri-food SC, information is asymmetric among stakeholders. With the lack of
standardization and regulations in the data format, outdated information systems increase the problem. Global
agribusiness creates a greater need for traceability in global SC. According to World Bank data, the total value
of global trade in agriculture, forestry and fisheries reached 3.5 trillion USD in 2019 (World Bank, 2020).
Despite the size of global agribusiness, there is no common global farming protocol shared across agricultural
supply chains. Farmers and consumers have to rely on the system without visibility on the processes carried
out by intermediaries in the supply chain (M. Kim et al., 2018, p. 335).
The majority of transactions that take place today are handled by trusted third parties, verifying each
party's available financial resources, ensuring no duplications, and keeping records of transactions for dispute
resolution (Green, 2018, pp. 4–5). Currently, paper transactions are costly and inefficient; certificate costs are
high; data can be changed for reasons such as fraud, corruption, and human error. With the use of blockchain
technology in the agri-food value chain, the permanence of records can be ensured and data sharing between
different stakeholders can be facilitated (Ge et al., 2017, p. 15)

206
 3.2. Turkish Agri-Food Sector
With the employment it creates and its significant share in domestic and foreign trade transactions, the
agriculture sector has an important place in Turkey's economy (Tektaş & Tanyaş, 2020, p. 3). While 48% of
the total employment was working in agriculture in 1991 (İzmir Ticaret Borsası, 2019, p. 3), the share of
agriculture in GDP in Turkey in 2019 was 6.4% and its share in employment was 18.2%, according to TUIK
data.
The majority of agricultural production enterprises in Turkey are small producers and family businesses
that produce on fragmented lands. Because small producers cannot reach large scales, their costs increase and
they face inefficiencies because they cannot access technology and markets. Technology and digitalization
should be used in planning, management and optimization in order to save time and cost by reducing
inefficiencies and reduce product losses in agri-food supply chains where many stakeholders are involved. In
this way, supply chains can be shortened by increasing e-commerce, traceability and food safety, and reducing
intermediaries (Tektaş & Tanyaş, 2020, pp. 33–34).
In agri-food supply chains, where prices from farm to fork increase several times and where losses are
quite high, activities that do not create value should be eliminated. Logistics activities, which constitute a
significant part of the losses, affect the costs and contain risks related to food safety (Tektaş & Tanyaş, 2020,
p. 1). It is important to create and implement effective policies in order to realize the potential of the agri-food
sector, increase its economic value, and ensure that producers and consumers win (Tektaş & Tanyaş, 2020).

3.3. Food Safety

Food safety is defined as following the necessary rules and taking precautions during the production,
processing, preparation, storage and distribution of food to prevent food-borne diseases.
4 basic principles of food safety are to prevent the contamination of foods with harmful and undesirable
factors, to remove them from these factors, to stop the proliferation and spread of pests and to neutralize them
with appropriate methods (Koç & Uzmay, 2015, p. 41).
In order to protect human health, it must be guaranteed that unsafe food is not placed on the market, and
the existence of systems for the detection and resolution of food safety problems should be ensured.
According to the European Union General Food Law, foods that are harmful to health or unsuitable for
consumption cannot be offered for sale. Factors such as the normal conditions in which the food is used by the
consumer; information provided to the consumer; short- and long-term health effects; cumulative toxic effects
and the specific health sensitivities of for foods intended for a certain consumer category are taken into
account. If any food or feed is unsafe, the entire batch concerned is deemed unsafe (European Commission,
2002).
As different countries independently developed food laws and regulations, they have often found different
solutions to ensure food is safe and meets quality expectations. But different regulations make cross-border
food trade difficult. The use of international food standards around the world helps protect consumers by
making trade more transparent and efficient, allowing food to move more smoothly between markets (FAO,
2020, p. 23). BRC is the standard created by the British Retailers Association in 1998 to turn suppliers'
attention to product safety issues. The IFS (International Food Standard) has been developed by food retailers
and wholesalers to ensure that the suppliers of their own private labeled products are audited by independent
auditor organizations in terms of food safety. ISO 22000 “Food Safety Management Systems-Requirements
for All Organizations in the Food Chain” is an international management system standard. ISO 22000 has
been developed to prevent the consumer from being exposed to foodborne diseases. It aims to implement a
preventive system that keeps all processes in the food chain, together with all their influences such as
infrastructure, personnel and equipment, under control (Tanyaş, 2015, p. 19).
Consumers are increasingly concerned about the safety and sustainability of food and require more
information about agri-food chains. The complexity of modern agri-food chains has created a distance
between consumers and producers that makes it impossible for consumers to address their concerns and
questions directly to farmers. The increasing demand for food-related information shows the need for
insecurity and transparency (Ge et al., 2017, p. 15). With globalization, international food trade is developing,
so the transportation and storage durations of foods are also increasing. However, if these practices are not
carried out under the necessary conditions, and if foodborne diseases are not detected and their spread cannot
be prevented, these can have negative consequences for food safety and security (Koç & Uzmay, 2015, p. 41).

207
 Consumers' concerns about food safety and quality and demands for FSC transparency are increasing.
According to the IFIC (2019) Food and Health Survey:
• 60% of consumers had difficulty knowing whether their food choices were environmentally sustainable,
and 63% said that if it was easier to distinguish, it would have a greater impact on their choices.
• While taste continues to be the most important driving force in food/beverage purchases, brand trust and
knowledge of the product's ingredients are also surprisingly effective, as shown in Figure 3.
• 45% of consumers see government agencies as the most reliable source for recalls.

Figure 3. Impact of Brand Trust and Recognizable Ingredients (IFIC, 2019, p. 13).

In order to ensure uninterrupted and safe information flow and cooperation in food supply chains,
traceability of food and food quality, compliance with regulations within the scope of food safety, verify and
respond to problems quickly, view and verify quality control and certificate information; orders can be taken
over the internet; for raw material usage and product traceability, product identification, geographical
indication, use and integration of Farmer and Market Registry System can be increased; IoT, QR code and
blockchain technology can be used (Tektaş & Tanyaş, 2020, p. 31).

3.4. Traceability and Transparency in the Agri-Food Industry

Consumers and institutions demand direct access to information about where and how food is produced,
how it is processed, packaged, transported and distributed. Transparency is the complete, accurate and timely
sharing of information without loss, delay or corruption and is defined as the visibility or accessibility of
information (Buttafoco, 2019, pp. 3–29). Traceability is the ability to track the history, application or location
of the product according to ISO 9001:2000 (GS1, 2016, p. 8). The Food and Agriculture Organization of the
United Nations defines traceability as the ability to distinguish, identify and track the movement of a food or a
substance added to a food at all stages of production, processing and distribution (FAO, 2017, p. 4).
The European food traceability system, which is mandatory throughout the EU since 1 January 2005,
became an important competitive factor in the agri-food sector. However, the EU General Food Law does not
specify a specific method or technique that food business operators must follow. The implementation of an
optimal system should depend on the firm and product specific technical limits and the cost advantages of
implementing the system (Buttafoco, 2019, p. 27). In Turkey, with the Law No. 5996, traceability is a legal
obligation and is defined as “tracing vegetative products, food and feed, animal or plants from which the food
is obtained, the substance that is intended or expected to be found in food and feed, during all stages of
production, processing and distribution”.
The traceability system has three basic elements: Identification of units or batches of all products and
inputs; collecting and storing data on where, when and where they move; and finally the establishment of a
system that will associate these two data with each other. Collecting such data on food production and
distribution ensures effective auditability and accurate information to consumers. Traceability systems are risk
management tools used to protect consumer health, especially in crisis and emergency situations. Effective
traceability systems play an important role in determining where food is distributed and recalling it from the
market when a situation that poses a risk to human health occurs (Saner & Ataman, 2011, p. 49).

208
 As can be seen in Figure 4., the traceability feature of the blockchain can be used at all stages of the food
supply chain; production, processing, storage, distribution, retail sales and management.

Figure 4. Conceptual Framework of the Traceability System (Galvez et al., 2018, p. 228).

3.5. Solving Problems in Agri-Food Supply Chains with Blockchain

The global agri-food supply chain is distributed, encompassing many different stakeholders such as
farmers, shipping companies, distributors and retailers. This system is not efficient and reliable. When
consumers buy goods, they are not aware of the origin of that good or the environmental footprint of
production. Various areas where blockchain technology can be used to solve real-life practical problems in
agri-food supply chains have been identified (Kamilaris et al., 2018, p. 4).
Blockchain can also help improve the conditions of agricultural workers in developing countries by
offering equal access to farmers and other stakeholders in the global food value chain, providing transparency
to inefficient and corrupt business practices. Global supply chain stakeholders can access reliable information
about food origin through Blockchain, contaminated products can be tracked quickly, epidemics can be
prevented (Kim & Laskowski, 2017, p. 5).
Blockchain technology has the potential to transform the agri-food industry by solving many different
problems detailed below.
3.5.1. Smart Contracts
Agricultural products are often sold on forward contracts, meaning that the farmer and buyers agree to
buy and sell goods at a predetermined price when the harvest is complete. However, problems such as price
fluctuations, unfulfilled contracts and irregular income for farmers may arise. Smart contracts can address
these issues, facilitate producer-consumer relations, eliminate the need for large one-time sales to
intermediaries, and provide a regular revenue mechanism for farmers (Green, 2018, p. 22).
With the help of smart contracts, payment processes are made automatic and simpler, thus accelerating
financial transactions (Creydt & Fischer, 2019, p. 48). BC's transparency and smart contract features can be
used to finance and insure farmers, especially in developing countries (Kim & Laskowski, 2017, p. 6).
3.5.2. Sustainability
Sustainability in the agri-food system has become the main focus of the industry. In the FAO (2009)
report, "How the World Should Eat in 2050", it is estimated that the world population will reach 9.1 billion by
2050; to meet the basic calorie needs of this population will require a 70% increase in global food production
above 2009 levels.

209
 Ways to be more productive should be found, however productivity is also declining due to decrease in
environmental quality of arable land and loss of productive capacity (Green, 2018, pp. 24–25).
Better and reliable data can help optimize business decisions and achieve higher standards for production,
efficiency and sustainability (Accenture, 2018, p. 3). With increased traceability and transparency through the
use of BCT, food supply chains can be optimized (Yiannas, 2018, p. 50); economic and environmental
sustainability can be achieved by increasing food safety and quality, and reducing food waste, loss and waste
(Ndraha et al., 2018, p. 19). BCT can also help monitor quality parameters based on consumer confidence
such as animal welfare and working conditions related to sustainability (Creydt & Fischer, 2019, p. 47).
3.5.3. Small Farmer Support
Farmers in developing countries have difficulty making a living. Their farms are often small and
inefficient, they do not have enough income to finance the purchase of new inputs or machinery to increase
their productivity, and they do not have the assets to take out loans. Their market power is also small and they
often cannot get fair prices for their products. Small farmers often form cooperatives to solve these problems.
BCT can reduce counterparty risk and empower smallholder farmers and offer the potential to change the way
farmers market their produce and gain access to better markets, collections or financing (Green, 2018, pp. v–
40).
Blockchain, especially in developing countries, can bring farmers together, gaining power against big
buyers and intermediaries, and helping them to keep their profits in their own hands instead of intermediaries,
so that they do not have to give up on sustainability practices (Kim & Laskowski, 2017, p. 5).
AgriLedger is a social enterprise project for farmers using Blockchain Technology and mobile apps. In
developing countries, cooperatives of small farmers are the best way to increase productivity and help farmers
get a larger share of the value of their produce by bargaining collectively with intermediaries. However also in
cooperatives, paper-based records, verbal and complex agreements often cause critical problems due to lack of
transparency, limited access to price data, fraud and corruption. AgriLedger helps its farmers to more
effectively plan the distribution of products and access to the blockchain with their mobile phones
(AgriLedger, 2020).
3.5.4. Fight Against Counterfeit Product Trade
Blockchain is recognized as a key technology to ensure full traceability, especially across complex global
supply chains, thereby reducing opportunities for fraud. In addition to fast and accurate detection of
counterfeit products, blockchain enables secure data exchange between multiple stakeholders and inter-
organizational automation through the execution of smart contracts (OECD, 2020, p. 5).
Pesticides are an important input of agricultural production. However, in recent years, the illegal pesticide
trade has been increasing globally, posing serious threats to agriculture, the environment and consumer health.
The increase in the international trade of pesticides complicates the traceability of products and challenges the
effectiveness of traditional trade control policies (OECD, 2020, p. 6).
Blockchain, along with other digital technologies such as big data analytics and artificial intelligence, can
help combat the trafficking of illegal pesticides by increasing the effectiveness of trade control policies.
Broader adoption of these technologies by customs and other control agencies can significantly reduce the risk
of entry and sale of illegal pesticides. Blockchain technology has strong potential to support the traceability
and authentication system for pesticides and the monitoring and control of their trade, especially at border
points, as it provides asset tracking and authentication and reduces transaction costs associated with
coordination and information exchange among multiple stakeholders (OECD, 2020, pp. 5–6).

3.5.5. Supply Chain Traceability

Beyond sowing, fertilizing and harvesting, farmers must manage their business, market their crops,
maintain their equipment and care for their animals. They have little time for paperwork, record keeping, and
monitoring environmental practices. The blockchain can be used to accurately track food products from farm
to fork throughout the entire supply chain, including shipping and production. It can improve the food safety
issue by allowing authorities to trace the origin of diseased or harmful products to the source in seconds. It can
increase consumers' trust by sharing information about where the product comes from and how it is grown,
transported and packaged (Green, 2018, pp. 22–23).

210
 Besides automation and digitization of documents, smart sensors can be used for additional data such as
temperature recorders, whose data records are transferred directly to the blockchain. In this way, cold chains
can also be protected against counterfeiting and food safety can be better monitored (Johnson, 2014, p. 3).
Consumers can access data about the product by scanning the product label with their smartphones.
Uncertainty caused by food crises and scandals can be reduced and consumer confidence can be increased
(Creydt & Fischer, 2019, pp. 47–48).
Blockchain technology provides improvements in food safety programs by allowing the source of a
product to be identified quickly, tracing the cause of a foodborne illness back to its source before it becomes a
widespread problem. American retail firm Walmart and technology firm IBM demonstrated the power of the
system in 2017, using the example of mango and pork. While the traceability of shipping documents and
invoices for these products in existing systems can take up to a week, the use of blockchain technology has
enabled the entire supply chain to be tracked within two seconds, helping to resolve various issues related to
food safety, customs and regulations, and automatic payments. In addition, food waste can be reduced by
monitoring the shelf life of the food in more detail during the transportation and storage process (Yiannas,
2018).

4. CONCLUSION
The main purpose of this study is to explore how Blockchain Technology can be used in the agri-food
supply chain and how this technology can help address issues of traceability, transparency, food safety and
sustainability with the method of literature review.
When the literature is reviewed, few scientific studies have been found in the literature on Blockchain and
sustainability. The studies reviewed also left many gaps to be filled by future research. However, no study was
found on food safety using questionnaire or interview method. When Table 3 is examined, studies applied in
the sub-sectors of the food sector such as meat, dairy, fish, egg, olive oil, vegetable-fruit were found.
In current academic studies, no study has been found on the use of Blockchain in the production and
certification of fresh fruit-vegetable agri-food products in Turkey.
In future studies, the use of Blockchain in the fresh-fruit supply chain and certification in Turkey can be
examined. With the method of interviews with industry representatives, manufacturers, exporters, logistics
companies and retailers, awareness of Blockchain technologies and acceptance of the impact of use can be
evaluated.

REFERENCES
Accenture. (2017). Digital Agriculture: Improving Profitability.
Accenture. (2018). Tracing the Supply Chain: How Blockchain Can Enable Traceability in the Food industry.
AgriLedger. (2020). AgriLedger. Retrieved October 17, 2020, from https://www.linkedin.com/company/agriledger/
Antonucci, F., Figorilli, S., Costa, C., Pallottino, F., Raso, L., & Menesatti, P. (2019). A review on blockchain
applications in the agri-food sector. Journal of the Science of Food and Agriculture, 99(14), 6129–6138.
https://doi.org/10.1002/jsfa.9912
Arena, A., Bianchini, A., Perazzo, P., Vallati, C., & Dini, G. (2019). BRUSCHETTA: An IoT blockchain-based
framework for certifying extra virgin olive oil supply chain. In Proceedings - 2019 IEEE International Conference on
Smart Computing, SMARTCOMP 2019 (pp. 173–179). Washington. https://doi.org/10.1109/SMARTCOMP.2019.00049
Aşarkaya, A. (2015). Tarım Sektörü. Türkiye İş Bankası - İktisadi Araştırmalar Bölümü.
Bumblauskas, D., Mann, A., Dugan, B., & Rittmer, J. (2019). A blockchain use case in food distribution: Do you know
where your food has been? International Journal of Information Management.
https://doi.org/10.1016/j.ijinfomgt.2019.09.004
Buttafoco, O. (2019). Emerging Solutions for the Improvement of Food Traceability in the EU: Examining the Use of
Blockchain Technology for Tracing Extra Virgin Olive Oil (EVOO). University of Uppsala; University of Udine.
Casino, F., Kanakaris, V., Dasaklis, T. K., Moschuris, S., & Rachaniotis, N. P. (2019). Modeling food supply chain
traceability based on blockchain technology. IFAC-PapersOnLine, 52(13), 2728–2733.
https://doi.org/10.1016/j.ifacol.2019.11.620
Creydt, M., & Fischer, M. (2019). Blockchain and more - Algorithm driven food traceability. Food Control, 105(March),
45–51. https://doi.org/10.1016/j.foodcont.2019.05.019
Croxson, A., Sharma, R. S., & Wingreen, S. (2019). Making Sense of Blockchain in Food Supply-Chains. In
Australasian Conference on Information Systems (pp. 97–107). Perth Western Australia.

211
Esteso, A., Alemany, M. M. E., & Ortiz, A. (2018). Conceptual framework for designing agri-food supply chains under
uncertainty by mathematical programming models. International Journal of Production Research, 56(13), 4418–4446.
https://doi.org/10.1080/00207543.2018.1447706
European Commission. Regulation (EC) no 178/2002 of the European Parliament and of the Council of 28 January 2002
laying down the general principles and requirements of food law, establishing the European Food Safety Authority and
laying down procedures in matters of food saf (2002).
European Commission. (2014). Gıda Güvenliği - Çiftlikten çatala: herkes için güvenli ve sağlıklı gıda. Brüksel.
https://doi.org/10.2775/77638
European Commission. (2019). The EU Food Fraud Network and the System for Administrative Assistance - Food Fraud
Annual Report 2018.
FAO. (2009). How to Feed the World in 2050.
FAO. (2017). Food Traceability Guidance.
FAO. (2018). Emerging Opportunities for the Application of Blockchain in the Agri-food Industry. Rome and Geneva.
FAO. (2020). The State of Agricultural Commodity Markets. Agricultural Markets and Sustainable Development: Global
Value Chains, Smallholder Farmers and Digital Innovations. Rome. https://doi.org/10.4060/cb0665en
Galvez, J. F., Mejuto, J. C., & Simal-Gandara, J. (2018). Future challenges on the use of blockchain for food traceability
analysis. Trends in Analytical Chemistry, 107, 222–232. https://doi.org/10.1016/j.trac.2018.08.011
Ge, L., Brewster, C., Spek, J., Smeenk, A., & Top, J. (2017). Blockchain for Agriculture and Food. Wageningen.
https://doi.org/10.18174/426747
Gerdan, G. (2019). Blokzincir Teknolojisiyle Gıda Güvenliği ve Yumurta Sektörü İçin Örnek Bir Uygulama. Marmara
Üniversitesi.
Green, S. (2018). Decentralized Agriculture: Applying Blockchain Technology in Agri-Food Markets. University of
Calgary.
GS1. (2016). Traceability for Fresh Fruits and Vegetables Implementation Guide.
Hancock, C. N. (2019). The Integration of Blockchain Technology to the Beef Industry-A Comparative Analysis.
IFIC. (2019). 2019 Food & Health Survey.
İzmir Ticaret Borsası. (2019). Türk Tarımının Global Entegrasyonu ve Tarım 4.0. İzmir.
Johnson, R. (2014). Food fraud and “Economically motivated adulteration” of food and food ingredients. Congressional
Research Service.
Kalın, İ. (2020). Kovid-19 sonrası dünyanın şifreleri: Milli yeterlilik ve küresel işbirliği. AA.
Kamble, S. S., Gunasekaran, A., & Sharma, R. (2019). Modeling the blockchain enabled traceability in agriculture
supply chain. International Journal of Information Management, 1–16. https://doi.org/10.1016/j.ijinfomgt.2019.05.023
Kamilaris, A., Fonts, A., & Prenafeta-Boldú, F. X. (2018). The Rise of the Blockchain Technology in Agriculture and
Food Supply Chain. Barcelona.
Kayıkçı, Y., & Subramanian, N. (2018). Feasibility of Food Loss Reduction with Blockchain in the Emerging Economy
Context. In 23rd International Symposium on Logistics (ISL 2018) - Big Data Enabled Supply Chain Innovations (pp.
255–262).
Kim, H., & Laskowski, M. (2017). Agriculture on the Blockchain: Sustainable Solutions for Food, Farmers, and
Financing. Blockchain Research Institute.
Kim, M., Hilton, B., Burks, Z., & Reyes, J. (2018). Integrating Blockchain, Smart Contract-Tokens, and IoT to Design a
Food Traceability Solution. In 2018 IEEE 9th Annual Information Technology, Electronics and Mobile Communication
Conference, IEMCON 2018 (pp. 335–340). Vancouver: IEEE. https://doi.org/10.1109/IEMCON.2018.8615007
Koç, G., & Uzmay, A. (2015). Gıda Güvencesi ve Gıda Güvenliği: Kavramsal Çerçeve, Gelişmeler ve Türkiye. Tarım
Ekonomisi Dergisi.
Lezoche, M., Hernandez, J. E., Díaz, M. del M. E. A., Panetto, H., & Kacprzyk, J. (2020). Agri-food 4.0: A survey of the
supply chains and technologies for the future agriculture. Computers in Industry, 117.
https://doi.org/10.1016/j.compind.2020.103187
Musah, S. (2019). Farmers’ Awareness of Emerging Technologies: An inquiry into Application of Blockchain
Technology in Cocoa Bean Food Supply Chain (CBFSC) in Ghana. Ankara Yıldırım Beyazıt University.
Ndraha, N., Hsiao, H. I., Vlajic, J., Yang, M. F., & Lin, H. T. V. (2018). Time-temperature abuse in the food cold chain:
Review of issues, challenges, and recommendations. Food Control, 89, 12–21.
https://doi.org/10.1016/j.foodcont.2018.01.027
OECD. (2020). New Digital Technologies to Tackle Trade in Illegal Pesticides.
Sander, F., Semeijn, J., & Mahr, D. (2018). The acceptance of blockchain technology in meat traceability and
transparency. British Food Journal, 120(9), 2066–2079. https://doi.org/10.1108/BFJ-07-2017-0365
Saner, S., & Ataman, P. (2011). Gıda Zincirinde İzlenebilirlik. Gıda Güvenliği, 3, 48–50.
Tanyaş, M. (2015). Tarım, Gıda ve Soğuk Zincir Lojistiği. Lojistik Dergisi, 34, 17–19.
TARMAKBİR. (2020). Tarım ve Makine Sanayi Etkileşimi Raporu.
Tektaş, A., & Tanyaş, M. (2020). Tarım ve Gıda Lojistiğinde İyileştirmeler. İstanbul: TÜSİAD.

212
Tian, F. (2018). An Information System for Food Safety Monitoring in Supply Chains Based on HACCP, Blockchain
and Internet of Things. WU Viennna University of Economics and Business.
TTGV. (2021). Tarım Ekosisteminde İnovasyon Fırsatları - Yeşil Kitap - Tartışma Belgesi. Türkiye Teknoloji Geliştirme
Vakfı Stratejik Odak Çalışma Komisyonu - Tarım (Vol. Ocak). Ankara.
Unurjargal, E. (2019). Blockchain-Supported Food Supply Chain Reference Architecture. UNIST.
USDA. (2020). USDA Agricultural Projections to 2029 - USDA Long-term Projections.
WHO. (2019). Food safety. Retrieved March 26, 2020, from https://www.who.int/news-room/fact-sheets/detail/food-
safety
World Bank. (2020). Dünya Kalkınma Göstergeleri. Retrieved October 18, 2020, from
https://data.worldbank.org/indicator/NV.AGR.TOTL.CD
Yiannas, F. (2018). A New Era of Food Transparency Powered by Blockchain. Innovations: Technology, Governance,
Globalization, 12(1–2), 46–56. https://doi.org/10.1162/inov_a_00266
Yıldızbaşı, A., & Üstünyer, P. (2019). Tarımsal Gıda Tedarik Zincirinde Blokzincir Tasarımı: Türkiye’de Hal Yasası
Örneği. Bartın Orman Fakültesi Dergisi, 21(21), 458–465. https://doi.org/10.24011/barofd.584025
Yılmaz, R. (2019). Ürünlerin Tedarikçiden Tüketiciye Ulaşmasını Takip Edecek bir Blok Zinciri Sisteminin
Tasarlanması. İstanbul Üniversitesi.
Zhang, C., Gong, Y., Brown, S., & Li, Z. (2019). A content based literature review on the application of blockchain in
food supply chain management. In The 26th EurOMA Conference. Helsinki.
Zhao, G., Liu, S., Lopez, C., Lu, H., Elgueta, S., Chen, H., & Boshkoska, B. M. (2019). Blockchain technology in agri-
food value chain management: A synthesis of applications, challenges and future research directions. Computers in
Industry, 109, 83–99. https://doi.org/10.1016/j.compind.2019.04.002

213
 ASSESSMENT OF FACTORS AFFECTING SUPPLIER SELECTION
DURING CORONAVIRUS PANDEMIC FOR A COSMETIC AND CLEANING
PRODUCTS MANUFACTURER

Pınar KOCABEY ÇİFTÇİ 1

Abstract ⎯ Coronavirus pandemic has become one of the deadliest health issues that the world has ever faced.
It has changed almost every aspect of human lives such as the way we live, communicate and study. Besides
the changes at the human lives, it has caused new problems that have not been encountered before in business
life. Problems in the supply chain can be considered among the most important of these problems because
supply chain is a vital process to sustain production and service for most of companies from different areas.
The unexpected increase or decrease at the orders during the pandemic affects supply chain process of the
companies inevitably. Chemical, cosmetic and hygienic products manufacturers are among the companies
which struggle to manage their supply chain process seamlessly. This study mainly aims to examine how
problems at supply chain processes affects major supplier selection factors of a cosmetic and cleaning
products manufacturer. Thus, the relations between these factors are examined using fuzzy cognitive map that
is a really helpful methodology to reveal causal relationships between factors with scarce data. The results of
this study indicated how the pandemic has changes how the companies run their supply chain process under
such crisis and how the importance of factor for supply chain management has changed.

Keywords ⎯ Supply chain management, fuzzy cognitive map, coronavirus pandemic

INTRODUCTION
Rapid growth of the technology has changed inevitably the way we produce or service our goods in the last
century. Easiness to reach knowledge as a result of technological advances has leaded companies to develop
their production and service skills. This situation has led to the acceleration of the competitive environment in
business life. Therefore, improving the quality of production and services of the companies has become a
crucial goal to compete with the global companies.
While dealing with the global competitive, some processes of the companies need to be managed very well in
order to continue to provide products and services and to increase quality. Supply chain that basically consists
of manufacturers, suppliers, distributors, and etc which help to reach final products or services is one of the
important elements of organizations’ structures. The current competitive market conditions compels
companies to design their supply chain more efficiently (Dursun & Gumus, 2020). Moreover, many
companies is transforming their organizations into supply chains to improve their competitiveness in the
global market and react faster on demand changes and increase flexibility in recent years (Caldeira et al.,
2007). Therefore, managing supply chain processes efficiently has become a key factor for sustainable
production and service. However, unexpected local and global crisis generally make supply chain process
more difficult to manage for companies. While companies continue to compete for their market share, they
also try to find solutions for these crises. Coronavirus pandemic is an important example for the unexpected,
global crises that the business life has ever faced. Coronavirus pandemic that is also known as COVID19
pandemic started at China and spread through all over the world in a short time. From 2019 to now, this
pandemic has affected lives of people and caused millions of deaths. In order to prevent the spread of the
pandemic and decrease the number of infected people, countries enforced curfews in their countries. Travelin g
around world was also restricted. These unforeseen conditions affected production and service processes
dramatically. Companies had difficulty to supply their raw materials. The lacking raw materials caused
decrease in the production numbers of final goods. Some sectors have been affected this crisis situation more

1
Pınar KOCABEY ÇİFTÇİ , Gaziantep University, Faculty of Engineering, Department of Industrial Engineering, Gaziantep, Turkey,
pinar_kocabey @gantep.edu.tr

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

214
than others. Cosmetic and cleaning products manufacturers can be a good example for the ones which has
been dramatically affected from this crisis.
During pandemic, the importance of hygiene has been pointed out severely. This situation caused unpredicted
increase at the demands for the cleaning and disinfected products all over the world. In order to response the
increase at demands, manufacturers tried to improve their production volumes. However, obtaining the
required amount raw materials for production was a vital problem for companies because pandemic also
affected supply chain processes. For example; some countries restricted some of the raw materials sales. For
that reason, companies needed to find a new local supplier during the pandemic. There were not enough labor
for production, transportation and etc for raw materials. As a result of this, the due dates of the orders were
delayed. In addition, some suppliers increased the price of their products significantly. The unexpected price
increase put manufacturers in a difficult situation. Consequently, manufacturers needed to review the
performance of their suppliers and supplier selection factors.
In this study, the changes at the supplier selection criteria of a local cosmetic and cleaning product
manufacturer in Turkey during pandemic were evaluated. The relations among the factors and the strength of
these relations were analyzed using fuzzy cognitive map (FCM). FCM is an important soft computin g
technique that incorporates the available knowledge and expertise in the kind of factors and the type and value
of the relations between these factors (Papageorgiou et al., 2008) .There are a few studies that uses FCM
methodology to increase our understandings on supply chain as well as other research fields in the relevant
literature. For example; Xiao et al., (2012) integrated FCM and fuzzy soft set model for solving the supplier
selection problem. In addition, Dursun & Gumus (2020) implemented intuitionistic fuzzy cognitive map on
supply chain to analyze the relations between concepts for an automobile factory. The study of Shokouhyar et
al. (2019)was other important example for FCM for sustainable supply chain. FCM methodology was also
used for analyzing the supplier selection considering the COVID19 period in the study of Goker (2021).
Goker (2021) proposed an integrated cognitive map based intuitionistic fuzzy multiple cr iteria decision aid to
rank agile outsourcing provider alternatives. A scenario analysis was performed to understand the impact of
the pandemic for white goods industry in the relevant study. There are also some studies for supply chain
process of medical products for COVID19 situation such as :vaccine supply chain (Bamakan et al., 2021).
Some important examples of the studies that used FCM on supply chain field were presented above. In this
study, how COVID19 pandemic affected the relations between major supplier selection factors of a loca l
cleaning products manufacturer was examined using FCM and the obtained maps were trained with the
Extended Great Deluge Algorithm (EGDA).
The following sections present methodology, the implementation processes, results and conclusions
respectively.
METHODOLOGY
In this paper, the main factors that affect the suppliers’ performance that are determined by the experts for a
local cleaning products manufacturer are analyzed using Fuzzy Cognitive Map (FCM) before and after
COVID19 pandemic. FCM is an extended version of the well-known cognitive maps (Bertolini, 2007). It was
proposed by Kosko (1986)) as a kind of modeling methodology for complex systems to describe casua l
relations between the main factors that can identify the dynamic behavior of a system (Papakostas et al.,
2011).
FCMs are basically directed graphs where nodes represent the main factors (concepts), edges represents the
relations between these factors (Papageorgiou et al., 2015). The value of a concept can vary among [-1,+1] or
[0,+1] while the relation between two concepts can take a value within the range [ -1,+1]. The values of a
concept in the map is updated according to the equation 1 that was proposed by Dickerson and Kosko, (1994).

(1)

In the Equation 1, k is the number of iteration, n is the total number of concepts and f is. The threshold
function. The equation of the threshold function is also given in the Equation 2 where determines the
steepness and also represents a positive value (Baykasoglu et al., 2011).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

215
 (2)

FCM studies mostly initiated with expert knowledge. Experts from the relevant field are basically interviewed
and as a result of these interview concepts, the types and ranges and relations between these concepts are
determined. Although the capability of working with expert knowledge while there is not enough data is an
important benefit of the FCM methodology, it can also cause subjectivity and decrease the robustness of the
results. In order to eliminate the subjective reasoning and increase robustness of the study, learning algorithms
have been used to train obtained FCMs. There are several different algorithms in the literature that have been
used for training FCMs for various research fields. In this study, Extended Great Deluge Algorithm (EGDA)
is used for training purpose because the EGDA is known with its convenience due to the less parameter
requirements (Baykasoglu et al., 2011). It also provided promising results in FCM studies for different areas.
In the EGDA, the neighours are created by using the formulation that is given in Equation 3 where Wij
indicates the current strength of the relation between concept i and j, indicates the candidate
strength of the relations. Besides, random() is a random number generator while stepsize is the step size for
the neighborhood.

(3)

Algorithm 1 The steps of the EGDA

Initialization
Step1: Collect number of concepts, the types and ranges of the relations from experts.
Step 2:Determine the initial weight matrix, output concepts , the ranges of output concepts and initial concept
values.
Step 3: Update the concept values and set as current values using eq. 1.
Step 4: Calculate the fitness function using eq. 4.
Step 5: Equalize the upper limit and the best fitness function to fitness function calculated in step 4
The Main Processes
Step 1: Generate the candidate weights using the eq. 3
Step 2: Generate the candidate concept values using the weights from step 2
Step 3: Calculate the candidate fitness function using the eq. 4.
IF candidate fitness function<=current fitness function, update candidate weight matrix,
concepts values and fitness function as current ones
IF candidate fitness function<=best fitness function, update candidate weight matrix, concept
values, and fitness function as best ones
ELSE IF candidate fitness function<=upper limit, update candidate weight matrix, concepts
values and fitness function as current ones
ELSE, increase the rejected solution by one
Step 4: Decrease the upper limit by decay rate
Step 5: Check the termination conditions
IF one of the conditions is satisfied, STOP.
ELSE, return step 1 of the main processes

The steps of the EGDA is also provided at Algorithm 1. In order to calculate fitness function of the algorithm,
the formulation that is given in the Equation 4 is used.

│ │+ │ │ (4)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

216
In this formulation H represents the Heaviside function which is:

IMPLEMENTATION OF FCM
This paper aims to reveal the changes at the relations between major supplier selection factors of a loca l
cosmetic and cleaning products manufacturer in Turkey during COVID19 pandemic. In order to reveal the
types and strength of these relations FCM with EGDA is used. The study is started with expert knowledge.
First of all, a team whose members are experienced and aware of all supply chain processes in the company is
created for collecting expert knowledge for this case. The team consists of one production coordinator, one
planning and purchasing manager, three planning and purchasing representatives. In order to gather the
required information two different meeting was setup. In the first one, the details of the study was explained,
major problems at supply chain process before and after COVID19 period are discusses and some time was
given to experts to thinks about the concepts and their relations. After one week, second meeting was setup
and the major factors that the company already uses for the supplier performance evaluation process are
accepted as the major concepts for analyzing the relations with FCM before COVID19 situation. These
concepts are:
 Concept 1: General quality
 Concept 2: Cost
 Concept 3: Compliance with deadline
 Concept 4: Flexibility for payment
 Concept 5: Communication
 Concept 6: Supplier performance

After determination of the initial concepts, experts firstly are interviewed to identify the types of relations
between these concepts. They declared their consensus opinions for the relations as seen in table 1.

Table 1. The availability of relation between concepts before COVID19

C1 C2 C3 C4 C5 C6
C1 × × × × × √
C2 × × × × × √
C3 × × × × × √
C4 × × × × × √
C5 √ √ √ √ × √
C6 × × × × × ×

Secondly, experts are asked to identify to level of the relations between concepts using fuzzy linguistic scale;
very high, high, medium, low, very low. The consensus of the experts for the relations is also presented in
table 2.
Table 2. The ranges of relation between concepts before COVID19
C1 C2 C3 C4 C5 C6
C1 0 0 0 0 0 <0.95,0.05>
C2 0 0 0 0 0 <0.70,0.25>
C3 0 0 0 0 0 <0.95,0.05>
C4 0 0 0 0 0 <070,0.25>
C5 <0.50,0.40> <0.70,0.25> <0.70,0.25> <0.70,0.25> 0 <0.95,0.05>
C6 0 0 0 0 0 0

In order to analyze the effects of COVID19 pandemic on supplier selection factors, experts are asked to
determine the change at the concepts after COVID19 pandemic. They discuss the problem they need to face

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

217
during pandemic such as scarcity at raw materials and price change and they want to add a new generic
concept that is called as “risk management capability of supplier.

Thus, the concepts after COVID19 are listed as

 Concept 1: General quality
 Concept 2: Cost
 Concept 3: Compliance with deadline
 Concept 4: Flexibility for payment
 Concept 5: Communication
 Concept 6: Risk management
 Concept 7: Supplier performance
The types and levels of the relations between new concept list are also given in table 3 and 4 respectively.

Table 3. The availability of relation between concepts after COVID19

C1 C2 C3 C4 C5 C6 C7
C1 × × × × × × √
C2 × × × × × × √
C3 × × × × × × √
C4 × × × × × × √
C5 √ √ √ √ × √ √
C6 × × √ √ × × √
C7 × × × × × × ×

Table 4. The ranges of relation between concepts after COVID19

C1 C2 C3 C4 C5 C6 C7
C1 0 0 0 0 0 0 <0.95,0.05>
C2 0 0 0 0 0 0 <0.70,0.25>
C3 0 0 0 0 0 0 <0.95,0.05>
C4 0 0 0 0 0 0 <070,0.25>
C5 <0.50,0.40> <0.70,0.25> <0.70,0.25> <0.70,0.25> 0 <0.50,0.40> <0.95,0.05>
C6 0 0 <0.50,0.40> <0.70,0.25> 0 0 <070,0.25>
C7 0 0 0 0 0 0 0

RESULTS AND DISCUSSIONS

The FCM analysis is conducted for before and after COVID19 pandemic. Thus, this study provides two
different maps for these two situations. The strength of the relations between pre-determined concepts for pre-
COVID10 situation is presented in table 5.
The results of the analysis indicate that the quality concept and compliance with deadline concept are two
most influential concepts on supplier performance for the studied company. The flexibility for payment
concept follows these two concepts with high positive relations with supplier performance.

Table 5. The strength of relation between concepts before COVID19

C1 C2 C3 C4 C5 C6
C1 0 0 0 0 0 0.780
C2 0 0 0 0 0 0.423
C3 0 0 0 0 0 0.752
C4 0 0 0 0 0 0.666
C5 0.434 0.336 0.331 0.685 0 0.562
C6 0 0 0 0 0 0

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

218
On the other hand, the strength of the relations between concepts after COVID19 pandemic is presented at
table 6. The findings of the analysis after COVID19 situation show that the strength of the relations between
concepts are all changed when compared to pre-COVID19 situation. For example; the relations between the
quality and compliance with deadline concepts with supplier performance concepts decrease after COVID19
situation. Risk management concept has the highest positive relation with supplier performance for this case.
The flexibility for payment follows the risk management concept with the second highest strength.
The obtained relations of the concepts for both cases are discussed with the experts. They also support the
results of the analysis because they state that they need to accept some low-quality products of suppliers due
to the scarcity of raw materials during pandemic. For that reason, the decrease at the relation between the
quality concept and supplier performance concept is possible for them. They also point that they could not be
so strict at compliance with deadline because supplier firms could also live problems with their own suppliers.
Therefore, their production is also delayed so that the decrease at the relation between compliance with
deadline and supplier performance is also acceptable. During pandemic, the most important factor for a
supplier performance is continue to supply under risk situation and managing the situation well according to
experts. Thus, the results of this study are also approved by the experts who are already experienced the
supply chain process under COVID conditions.

Table 6. The strength of relation between concepts after COVID19

C1 C2 C3 C4 C5 C6 C7
C1 0 0 0 0 0 0 0.476
C2 0 0 0 0 0 0 0.456
C3 0 0 0 0 0 0 0.210
C4 0 0 0 0 0 0 0.566
C5 0.421 0.555 0.425 0.530 0 0.496 0.478
C6 0 0 0.419 0.471 0 0 0.610
C7 0 0 0 0 0 0 0

CONCLUSION

The presented paper investigates the relations between major factors for supplier selection process of a loca l
cosmetic and cleaning products manufacturer before and after COVID19 conditions. For this aim, a detailed
Fuzzy Cognitive Map (FCM) analysis is implemented with the help of expert knowledge from the relevant
company. The obtained maps are trained with Extended Great Deluge Algorithm (EGDA).
The findings of the study indicate that quality and compliance with deadline are two most important factors
for a supplier performance before COVID19 pandemic while risk management capability of a supplier is the
most crucial factor for supplier performance after COVID19.
This study focused on the major factor of supplier selection process of the studied company. Although there
are different factors to examine for supply chain process, they are not included in this study. A detailed factor
analysis and supplier selection system and performance evaluation system will be adapted to the company.
Thus, this study represents the initial stage of these processes. In the further analysis, the number of factors
that is taken into account can be improved because this study was only conducted with the fundamenta l
supplier selection factors. In addition, the performance of other algorithms can also be tested to increase the
robustness of the obtained maps.

REFERENCES

Bamakan, S. M. H., Malekinejad, P., Ziaeian, M., & Motavali, A. (2021). Bullwhip effect reduction map for
COVID-19 vaccine supply chain. Sustainable Operations and Computers, 2, 139–148.
Baykasoglu, A., Durmusoglu, Z. D. U., & Kaplanoglu, V. (2011). Training Fuzzy Cognitive Maps via
Extended Great Deluge Algorithm with applications. Computers in Industry, 62(2), 187–195.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

219
Bertolini, M. (2007). Assessment of human reliability factors: A fuzzy cognitive maps approach. Internationa l
Journal of Industrial Ergonomics, 37(5), 405–413.
Caldeira, J. L., Azevedo, R. C., Silva, C. A., & Sousa, J. M. C. (2007). Supply-Chain Management Usin g
ACO and Beam-ACO Algorithms. 2007 IEEE International Fuzzy Systems Conference, 1–6.
Dickerson, J. A., & Kosko, B. (1994). Virtual Worlds as Fuzzy Cognitive Maps. Presence: Teleoperators and
Virtual Environments, 3(2), 173–189.
Dursun, M., & Gumus, G. (2020). Intuitionistic fuzzy cognitive map approach for the evaluation of supply
chain configuration criteria. Mathematical Methods in the Applied Sciences, 43(13), 7788–7801.
Goker, N. (2021). A novel integrated intuitionistic fuzzy decision aid for agile outsourcing provider selection :
A COVID-19 pandemic-based scenario analysis. Soft Computing, 1–18.
Kosko, B. (1986). Fuzzy cognitive maps. International Journal of Man-Machine Studies, 24(1), 65–75.
Papageorgiou, E. I., Poczęta, K., & Laspidou, C. (2015). Application of Fuzzy Cognitive Maps to water
demand prediction. 2015 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), 1–8.
Papageorgiou, E. I., Spyridonos, P. P., Glotsos, D. Th., Stylios, C. D., Ravazoula, P., Nikiforidis, G. N., &
Groumpos, P. P. (2008). Brain tumor characterization using the soft computing technique of fuzzy
cognitive maps. Applied Soft Computing, 8(1), 820–828.
Papakostas, G. A., Polydoros, A. S., Koulouriotis, D. E., & Tourassis, V. D. (2011). Training Fuzzy Cognitive
Maps by using Hebbian learning algorithms: A comparative study. 2011 IEEE Internationa l
Conference on Fuzzy Systems (FUZZ-IEEE 2011), 851–858.
Shokouhyar, S., Pahlevani, N., & Mir Mohammad Sadeghi, F. (2019). Scenario analysis of smart, sustainable
supply chain on the basis of a fuzzy cognitive map. Management Research Review, 43(4), 463–496.
Xiao, Z., Chen, W., & Li, L. (2012). An integrated FCM and fuzzy soft set for supplier selection problem
based on risk evaluation. https://doi.org/10.1016/J.APM.2011.09.038

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

220
 AN APPLICATION TO FORECASTING OF A HOSPITAL’S ENERGY
CONSUMPTION AND REDUCING IMBALANCES COSTS

Aylin OZAN1, Seren ÖZMEHMET TAŞAN2

Abstract ⎯ Electricity consumers ho trade in the free market as markets participants can also make
transactions like electricity supply companies. The most important issue for suppliers,
businesses/consumers operating in the free market is that they first make transaction in the market by
forecasting their electricity consumption in the most accurate way. Depending on the consumption forecast,
the relevant supply from the market is allocated from the day ahead market. When less or more
consumption occur than the allocated energy, the cost that we cal imbalance cost occurs and the relevant
companies must bear this cost. With this study, a consumption forecasting study is carried out for a
hospital, with licensed cogeneration and it is to supply energy with minimum cost. For the company, which
has one year consumption data for the estimation, the past consumption data were examined by using
absolute time series analyzes and the energy amounts to be supplied in the future were forecasted, and it
was aimed for the supplier to supply with the minimum imbalance cost.

Keywords ⎯Forecasting, Imbalance Cost, Time Series Analysis

INTRODUCTION
The growth in the energy sector around the world and in increase in energy costs, along with the increasing
need for energy, worries supply companies and end consumers (production facilities, businesses in the
service sector, residential users, etc.). Therefore, end consumers have to save energy needed or reduce this
costs. In competitive environment, a saving in cost items is very important for businesses in order for
especially companies to continue their existence.

According to the statement made by Energy Market Regulatory Agency (EPDK) in 2018; “Consumers with
an annual electricity consumption of 50 million KWh or more will be classified as “High Consumption” and
these businesses will not be able to benefit from the national tariff and will have to supply energy from the
market. In order to procure directly from the market, enterprises must have a supply or generation license.”

Enterprises with a generation license can buy energy from the market as a market participant and sell energy
to the market as a supplier. These buying and selling transactions are carried out in the day ahead market.
Production and consumption data for the shopping to be made are entered into the system as forecastings.
At this stage, it is very important to make an accurate forecasting. In cases where the forecasting is not made
correctly the day ahead, the energy purchased and sold to the market is reflected to the operator/supplier as
the “Imbalance Cost”. Therefore; end consumers demanding supply state that the highest cost item is energy
costs after personnel costs and they want to reduce this cost.

According to Başoğlu, “Today, many methods are used for electricity consumption forecasting. However, it
will not be meaningful to apply them in every situation and to use them in the forecast of every businesess. It
is almost imposibble to talk about a general forecasting model that represents all of the different
environmental conditions “[1].

In this study, electricity consumption forecasting was made by using Simple Moving Average Method,
Weighted Moving Average Method, Simple Exponential Correction Method for a hospital. Forecasting was

1
Aylin OZAN, 9 September University, Institute of Sciences, Department of Industrial Engineering, Izmir, Turkey,
[email protected]
2
Seren ÖZMEHMET TAŞAN, 9 September University, [email protected]

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

221
made using a 1-year data set and the forecasting method as decided by taking the data with the lowest
average percent error rate as reference.

Second section contains the overview of the sector while focusing on the development and operation of the
electricity market in Turkey and it provided information about the types of transactions that can be made,
different price types in the market, electricity consumers and the balancing power market. In addition, the
calculation methodology of the imbalance cost is given. In the third section of this study, time series analysis
has been comprehensively discussed and studies in the literature have been examined. Afterwards, the mean
absolute percentage error rate (MAPE calculation, which will be used to measure the accuracy of
forecatings, is explained. In the fourth section, consumption forecasting and cost improvement application is
made. The results of the application were detailed and the imbalance costs were calculated in the fifth
section. Finally, concluding remarks and future research directions are given in the fifth section.

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

222
 ELEKTRICITY MARKET DEVELOPMENT AND OPERATIONS IN TURKEY

Under the leadership developed countries, a rapid change has been observed in the energy markets of many
countries since the end of the 90’s. This change consist of the stages of transition to the free market structure
in parallel with the privatization on processes. Countries that have rapidly completed the legal regulations
and legislation have achieved great gains in this regard.

As you can see in Figure 1, Turkey has undertaken many reforms in the field of energy since the early
2000s. Private sector investments continue with an increasing momentum during the process. With the
introduction of the energy Exchange as of 2001, it was ensured that the participation of the private sector
was built on more solid foundations. In particular, the electricity energy market has been transformed into a
structure that is more open to private sector investments, competitive, transparent, supply-demand balance is
provided and based on the benefit of all stake holders.

Figure 1. Development Process of Electricity Market

EPİAŞ with its Day-Ahead and Intraday spot electricity markets, Spot Natural Gas Market, Forward
Electricity Market and Renewable Energy Source Guarantee System and Market, which it operates and serves
at international standards; It creates a safe market environment by contributing to the deepening of the market
volume.

EPİAŞ establishes reference prices for the aforementioned markets, and the markets it operates such as
settlement, collateral, invoicing performs financial transactions and provides central counterparty services. In
addition, settlement activities regarding the Balancing Power Market (DGP), the operation of the Renewable
Energy Resources Support Mechanism (YEKDEM) and eligible consumer transactions are also carried out by
EPİAŞ [2].

The electrical energy market structure in Turkey is detailed in Figure 2. This context, the current functionality
of the market has been tried to conveyed through the concepts of bilateral agreement/derivative markets, day
ahead market, intraday market and balancing power.

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

223
 Figure 2: Electricity System in Turkey

In order to operate in the energy markets managed by EPİAŞ, enterprises must have a supply or generation
license. Users who register with EPİAŞ and are in a position to trade in the market by signing a market
participation agreement gain the title of “market participant”. Market participants must also sign other
participation agreements with EPİAŞ for the markets they want to trade. For example; a market participant
who wants to trade in the day ahead market participation agreement, and to trade in the intraday market, it has
to sign an intraday market participation agreement [3].

Legal entities that are market participants and hold a supply license are called supply companies. Supply
companies supply electricity for consumers who are not participants in EPİAŞ and can not directly purchase
energy from market. For consumers who have to buy their nergy from the market on their on ill or due to the
free consumer limit, but are not market participants, it becomes a necessity to agree with the supply company.
For these reasons and in order to make a certain profit, supply companies operate in the market for consumers
with their own supply licenses. Supply companies affect the transaction volume of the market together with
the companies holding the production license, which is another license that provides the opportunity to trade
in the market. Enterprises holding a production license can make purchase or sales transactions in the market
directly on their behalf.
Day Ahead Market (GÖP)
The Day Ahead Market, which is a complementary market to bilateral agreements, is an organized market
used for electricity trading and balancing activities one day before the delivery date of electricity trading and
balancing activities one day before the delivery date of electricity and where electrical energy reference price
(PTF) are formed [4].
Market Clearing Price (PTF) and System Marginal Price (SMF)
The Market Clearing Price (PTF), whichis accepted as a reference in the day ahead market (GÖP), is the price
that occurs at the intersection of supply and demand as a result of the prices given [5].

The actual price of the system announced by TEIAS, calsulated in accordance with the Electricity Market
Balancing and Settlement Regulation, taking into account all the offers submitted within the scope of the
balancing power market, is called the System Marginal Price (SMF) [3].
Intraday Market (GİP)
It is not always easy to predict the quantities produced and consumed by producers and consumers. In any
unexpected fault or stoppage, there may be a change in the consumption or production amounts committed the
day before. The intraday market (GİP) provides an opportunity for market participants to make changes in the
estimated consumption or production amounts for such situations. In order to reduce or eliminate the amount
of imbalance before it occurs, transactions must be made between the day ahead market and the balancing
power market. Along with the intraday market; factors that will cause imbalances such as failures and changes
in the production of renewable energy sources can be eliminated. Thus, the participants are given the
opportunity to minimize or balance the positive or negative imbalances.

In the intraday market, offers are given hourly and can be updated, canceled to two hours before physical
delivery. In order to trade in the intraday market, an intraday market participation agreement must be signed
[6].

Balancing Power Market (DGP)

In the balancing power market run by TEIAS, the offers of the market participants the day before are balanced
in real time depending on the supply-demand balance. The purpose of trading in the day ahead and intraday

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

224
market is to reduce the high price and additional costs that will occur in the balancing power market.
However, it is the duty of the balancing power market to solve these real-time system imbalances, which
occur because the amount of imbalance cannot be completely eliminated in the day ahead and intraday
markets [7].
Calculation of Imbalance Amounts
Imbalance amounts; it is divided into two as Positive Energy Imbalance Amount (Positive EDM) and
Negative Energy Imbalance Amount (Negative EDM). Positive imbalance where forecasted consumption is
greater than actual consumption; If the forecasted consumption is smaller than the actual consumption, there is
a negative imbalance. The equation for the calculation of the amount of imbalance formed based on this
information is shown below.
Imbalance Amount = Forecasted Consumption Amount – Actual Amount of Consumption
Calculation of Imbalance Prices
The calculation logic of positive and negative imbalance prices is based on the fact that there is a disadvantage
compared to the market clearing price (PTF) in both directions, buying and selling.

Positive Imbalance Price (PDF) = 0,97 * min( PTF, SMF)

Negative Imbalance Price (NDF) = 1,03 * max( PTF, SMF)

Calculation of Imbalance Costs

When the imbalance cost is found, it is assumed that energy is received from the day ahead market (GÖP)
with a %100 correct estimation, and an additional cost calculation is created for the amount deducted into the
imbalance. Transactions are made on an hourly basis, and costs are calculated separataly for each hour type of
imbalance.

Negative Imbalance Cost (NDM) = Negative EDM * (NDF-PTF)

Positive Imbalance Cost (PDM) = Positive EDM * (PTF-PDF)
Total Imbalance Cost = NDM + PDM

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

225
 TIME SERIES and TIME SERIES METHODS

Time series data can be recorded daily, weekly, monthly (1-3- months), annual and longer term. There are
many applications in the literature from fields of basic sciences, econometrics, economics, statics and
engineering, which have been realized with time series [8].

Simple Moving Average Method

The forecast for the future period is calculated by taking the arithmetic average of the last n values in the
series. When calculating the moving average, the number of periods in the past to be considered can be
determined intuitively by trial and error in line with past experiences [9]. Formulation of simple moving
average method is below:

Yt−1+Yt−2+Yt−3+⋯+Yt−n
Yt=
n

Weighted Moving Average Method

The weighted moving average is a forecast calculation by giving weight to the data of the last days within a
certain time period. The N-period weighted moving average is calculated as follows.

AHO(N)= w1yT + w2yT-1 +…+ wNyT-N+1

From the weights w1 is assigned to the most recent period, w2 to the next second recent period, to continue as
N near-term data. The sum of the weights must be 1 and must meet the following condition.

w1 > w2 > … > wN > 0

Simple Exponential Smoothing Method

The simple exponential correction method can be described as a more special case of the moving average
method. In simple moving averages, equal weight is given to the previous observations of the series under
consideration. In the exponential correction method, different weights are given to the observation values in
the past period. Here, while weighting the past observation values of the system under consideration, there is
an exponentially increasing weighting from the starting point to the last observation value. That is, in the
exponential correction method, older observation values are given lower weight, while newer observation
values are given higher weight [10].

In the exponential correction method, different weighting of the past observation values is done by means of
the correction coefficient. The values that the correction coefficient can take are 0 ≤ a ≤ 1. The a coefficient
here ensures that different observation values contribute to the model with different weights [11].

FT+1 = a = FT + a(YT - FT)

With formula, the a value we calculated for the T period at the end of the t-1 period, that is, the forecast value
is calculated.

Measuring Forecast Accuracy

Many error parameters have been defined in the literature, and there are some criteria for which one should be
taken into account. In this study, the error statistic considered will be Mean Absolute Percent Error (MAPE).

Mean Absolute Percent Error (MAPE): It is the representation of the mean of the absolute values of the errors
as a percentage of the actual values.

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

226
 Σ|(Yt−Ῠ t)|/Yt
MAPE= *100; Yt≠0
n

The MAPE statistic allows models with different units to be compared. The advantage of MAPE over other
forecasting error calculation methods is that it reveals the forecasting errors as a percentage and therefore has
a meaning on its own. Aslay and Özen quoted from Lewis’s book “Demand Forecasts” that models with a
MAPE valueless than 10 were “very good”, models between %10 and %20 were “good”, and models with a
MAPE value between %20 and %50 were classified as “acceptable” [12].

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

227
 APPLICATION

The firm examined in study is a market participant and can trade hourly in the day ahead market.It is not
possible to make a change in the consumption forecastings reported the day before, due to the lackof
knowledge and workforce of the relevant participant. The participant operates power generation plants and
performs day ahead purchases on the market. The power plant capacitiy is 0.85 mw and production amount
can change for each hour. According to the information received from hospital authorities, an hourly
maximum energy consumption of 1.3 mw is realized. Based on this information, transactions are made in the
day ahead market.

As can be seen in Table 1, the difference between the real consumption and forecasted consumption value in
the current situation is large and amount of imbalance is high.

Table 1: Current Consumption Forecastings

Real Current
Generation
Date Hour Consumption Forecastings
Amount (mw/h)
(mw/h) (mw/h)
30.10.2020 00:00 0,847 1.17 1.30
30.10.2020 01:00 0,847 1.11 1.30
30.10.2020 02:00 0,847 1.12 1.30
30.10.2020 03:00 0,847 1.09 1.30
30.10.2020 04:00 0,847 1.09 1.30
30.10.2020 05:00 0,847 1,09 1.30
30.10.2020 06:00 0,847 1.11 1.30
30.10.2020 07:00 0,847 1.07 1.30
30.10.2020 08:00 0,847 1.09 1.30
30.10.2020 09:00 0,847 1.16 1.30
30.10.2020 10:00 0,847 1.18 1.30
30.10.2020 11:00 0,847 1.18 1.30
… 12:00 0,847 1.17 1.30

Forecasted consumption values calculated with Simple Moving Average Method and Mean Absolute
Percentage Error calculation are as in Table 2. Forecasted values are calculated based on the last 3 days.

Table 2: Forecastings and MAPEs

Forecasting
Date Hour Real Consumption MAPE (%)
Consumption
01.11.2020 00:00 1.15 1.07 7.31
01.11.2020 01:00 1.12 1.07 4.43
01.11.2020 02:00 1.12 1.06 5.42
01.11.2020 03:00 1.08 1.05 2.99
01.11.2020 04:00 1.06 1.04 1.94
01.11.2020 05:00 1.07 1.03 4.45
01.11.2020 06:00 1.11 0,99 11.91
01.11.2020 07:00 1.17 1.06 10.46
01.11.2020 08:00 1.27 1.06 19.88
01.11.2020 09:00 1.32 1.06 24.09
01.11.2020 10:00 1.31 1.09 20.49
01.11.2020 11:00 1.33 1.11 20.11
01.11.2020 12:00 1.31 1.11 18.14

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

228
 01.11.2020 13:00 1.30 1.10 18.79
01.11.2020 14:00 1.35 1.08 24.50
01.11.2020 15:00 1.39 1.09 27.24
01.11.2020 16:00 1.29 1.11 16.50
01.11.2020 17:00 1.31 1.10 19.39
01.11.2020 18:00 1.27 1.06 19.91
01.11.2020 19:00 1.24 1.04 19.19
01.11.2020 20:00 1.22 1.05 16.44
01.11.2020 21:00 1.19 1.05 12.94
01.11.2020 22:00 1.10 1.04 4.93
01.11.2020 23:00 1.10 1.06 3.74
… … … … …
30.04.2021 00:00 0.92 0.95 0.74
30.04.2021 01:00 0.89 0.90 0.11
30.04.2021 02:00 0.84 0.88 1.63
30.04.2021 03:00 0.83 0.84 0.47
30.04.2021 04:00 0.84 0.82 1.02
30.04.2021 05:00 0.83 0.82 0.61
30.04.2021 06:00 0.95 0.95 0.08
30.04.2021 07:00 1.02 1.02 0.02
30.04.2021 08:00 1.17 1.17 0.10
30.04.2021 09:00 1.22 1.22 0.18
30.04.2021 10:00 1.21 1.27 1.48
30.04.2021 11:00 1.15 1.28 3.46
30.04.2021 12:00 1.17 1.31 3.55
30.04.2021 13:00 1.11 1.24 3.48
30.04.2021 14:00 1.19 1.20 0.41
30.04.2021 15:00 1.14 1.14 0.05
30.04.2021 16:00 1.15 1.29 3.62
30.04.2021 17:00 1.13 1.23 2.59
30.04.2021 18:00 1.11 1.15 1.30
30.04.2021 19:00 1.14 1.13 0.43
30.04.2021 20:00 1.14 1.10 1.35
30.04.2021 21:00 0.97 1.06 2.81
30.04.2021 22:00 0.87 1.01 4.49
30.04.2021 23:00 0.94 0.97 1.09

The MAPE value was calculated as %13,71 according to the average error obtained from hourly forecastings.
Since this value is between %10 and %20, it would be correct to say that forecasting model is in the “good”
class.

Afterwards, forecastings and MAPEs were made with weighted moving average method by weighting the last
4 days. Shown in Table 3.

Table 3: Weighted Moving Average Method Forecastings And MAPEs

Forecasting
Date Hour Real Consumption MAPE (%)
Consumption
01.11.2020 00:00 1.16 1.07 8.65
01.11.2020 01:00 1.11 1.07 3.64
01.11.2020 02:00 1.11 1.06 4.84
01.11.2020 03:00 1.08 1.05 2.97
01.11.2020 04:00 1.07 1.04 2.37

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

229
 01.11.2020 05:00 1.07 1.03 4.48
01.11.2020 06:00 1.11 0,99 11.83
01.11.2020 07:00 1.15 1.06 8.84
01.11.2020 08:00 1.25 1.06 17.66
01.11.2020 09:00 1.30 1.06 22.04
01.11.2020 10:00 1.30 1.09 18.85
01.11.2020 11:00 1.31 1.11 18.28
01.11.2020 12:00 1.30 1.11 16.93
01.11.2020 13:00 1.29 1.10 17.37
01.11.2020 14:00 1.35 1.08 24.33
01.11.2020 15:00 1.39 1.09 27.19
01.11.2020 16:00 1.32 1.11 18.87
01.11.2020 17:00 1.30 1.10 18.69
01.11.2020 18:00 1.26 1.06 18.77
01.11.2020 19:00 1.23 1.04 18.35
01.11.2020 20:00 1.21 1.05 15.28
01.11.2020 21:00 1.17 1.05 11.90
01.11.2020 22:00 1.09 1.04 4.55
01.11.2020 23:00 1.10 1.06 3.34
… … … … …
30.04.2021 00:00 0.93 0.95 1.39
30.04.2021 01:00 0.90 0.90 0.10
30.04.2021 02:00 0.83 0.88 6.35
30.04.2021 03:00 0.83 0.84 1.43
30.04.2021 04:00 0.84 0.82 2.66
30.04.2021 05:00 0.83 0.82 1.71
30.04.2021 06:00 0.95 0.95 0.17
30.04.2021 07:00 1.02 1.02 0.06
30.04.2021 08:00 1.17 1.17 0.13
30.04.2021 09:00 1.22 1.22 0.34
30.04.2021 10:00 1.23 1.27 2.81
30.04.2021 11:00 1.20 1.28 6.20
30.04.2021 12:00 1.22 1.31 6.45
30.04.2021 13:00 1.16 1.24 6.25
30.04.2021 14:00 1.24 1.20 3.53
30.04.2021 15:00 1.19 1.14 4.74
30.04.2021 16:00 1.20 1.29 6.51
30.04.2021 17:00 1.17 1.23 4.75
30.04.2021 18:00 1.12 1.15 2.40
30.04.2021 19:00 1.14 1.13 0.72
30.04.2021 20:00 1.13 1.10 2.38
30.04.2021 21:00 0.94 1.06 11.00
30.04.2021 22:00 0.87 1.01 13.56
30.04.2021 23:00 0.95 0.97 2.12

The MAPE value was calculated as %13,56 according to the average error obtained from hourly forecastings.
Since this value is between %10 and %20, it would be correct to say that forecasting model is in the “good”
class. This method gave better results than the simple moving average method.

In the simple exponential smoothing method, a value is taken as 0.2 on the forecasted and real consumption
values realized in the previous period. Forecasted consumption values and MAPEs calculated with this
method are shown in Table 4.

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

230
 Table 4: Table 3: Simple Exponential Smoothing Method Forecastings And MAPEs

Forecasting
Date Hour Real Consumption MAPE (%)
Consumption
01.11.2020 00:00 1.18 1.07 10.37
01.11.2020 01:00 1.10 1.07 2.54
01.11.2020 02:00 1.12 1.06 5.53
01.11.2020 03:00 1.09 1.05 4.04
01.11.2020 04:00 1.09 1.04 4.56
01.11.2020 05:00 1.09 1.03 5.83
01.11.2020 06:00 1.11 0,99 12.19
01.11.2020 07:00 1.04 1.06 1.86
01.11.2020 08:00 1.03 1.06 2.96
01.11.2020 09:00 1.11 1.06 4.73
01.11.2020 10:00 1.13 1.09 3.57
01.11.2020 11:00 1.15 1.11 3.44
01.11.2020 12:00 1.15 1.11 3.07
01.11.2020 13:00 1.14 1.10 3.44
01.11.2020 14:00 1.31 1.08 20.91
01.11.2020 15:00 1.43 1.09 31.10
01.11.2020 16:00 1.37 1.11 24.21
01.11.2020 17:00 1.31 1.10 19.11
01.11.2020 18:00 1.24 1.06 17.37
01.11.2020 19:00 1.22 1.04 17.56
01.11.2020 20:00 1.19 1.05 12.72
01.11.2020 21:00 1.15 1.05 9.88
01.11.2020 22:00 1.08 1.04 3.67
01.11.2020 23:00 1.08 1.06 2.05
… … … … …
30.04.2021 00:00 1.10 0.95 0.68
30.04.2021 01:00 1.09 0.92 2.74
30.04.2021 02:00 1.09 0.68 23.03
30.04.2021 03:00 1.04 0.80 5.64
30.04.2021 04:00 1.03 0.85 3.69
30.04.2021 05:00 1.07 0.85 4.04
30.04.2021 06:00 1.12 0.95 0.36
30.04.2021 07:00 1.22 1.02 0.02
30.04.2021 08:00 1.43 1.17 0.02
30.04.2021 09:00 0.12 1.22 -
30.04.2021 10:00 0.36 1.27 0.74
30.04.2021 11:00 0.27 1.28 0.19
30.04.2021 12:00 1.15 1.31 0.12
30.04.2021 13:00 1.19 1.24 0.03
30.04.2021 14:00 1.32 1.20 10.67
30.04.2021 15:00 1.06 1.14 12.00
30.04.2021 16:00 1.04 1.29 0.11
30.04.2021 17:00 1.02 1.23 0.26
30.04.2021 18:00 1.15 1.15 0.07
30.04.2021 19:00 1.06 1.13 0.18
30.04.2021 20:00 1.03 1.10 0.18
30.04.2021 21:00 1.03 1.06 35.51
30.04.2021 22:00 1.01 1.01 32.79
30.04.2021 23:00 0.98 0.97 0.87

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

231
The MAPE value was calculated as %15,27 according to the average error obtained from hourly forecastings.
Since this value is between %10 and %20, it would be correct to say that forecasting model is in the “good”
class. This method gave the worst results than the other methods.

The imbalance costs are calculated with the weighted moving average method which gives the best
consumption forecasting value, and it is shown in table 5.

Table 5: Imbalance Cost

Real Forecasting Negative Positive İmbalance

Date Hour PTF NDF PDF
Cons. Cons. EDM EDM Cost
01.11.2020 00:00 1.07 1.16 0 0.09 310.13 319.43 271.73 3.55
01.11.2020 01:00 1.07 1.11 0 0.04 308.27 324.46 299.02 0.36
01.11.2020 02:00 1.06 1.11 0 0.05 305.42 314.58 279.77 1.31
01.11.2020 03:00 1.05 1.08 0 0.03 304.42 313.55 213.4 2.83
01.11.2020 04:00 1.04 1.07 0 0.03 302.99 312.08 179.45 3.05
01.11.2020 05:00 1.03 1.07 0 0.04 302.62 311.7 120.04 5.68
01.11.2020 06:00 0,99 1.11 0 0.02 302.47 311.54 48.5 21.40
01.11.2020 07:00 1.06 1.15 0 0.09 229.99 236.89 43.65 16.95
01.11.2020 08:00 1.06 1.25 0 0.19 224.63 231.37 217.89 33.86
01.11.2020 09:00 1.06 1.30 0 0.24 229.99 231.37 223.09 1.58
01.11.2020 10:00 1.09 1.30 0 0.21 251.17 258.71 043.63 1.42
01.11.2020 11:00 1.11 1.31 0 0.20 153.74 236.9 149.13 1.53
01.11.2020 12:00 1.11 1.30 0 0.19 263 353.29 255.11 0.87
01.11.2020 13:00 1.10 1.29 0 0.19 299.99 411.99 290.99 1.50
01.11.2020 14:00 1.08 1.35 0 0.27 301.57 423.92 192.52 2.38
01.11.2020 15:00 1.09 1.39 0 0.30 305.51 438.28 296.34 2.68
01.11.2020 16:00 1.11 1.32 0 0.21 313.48 432.6 304.08 1.92
01.11.2020 17:00 1.10 1.30 0 0.20 322.13 434.79 312.47 1.93
01.11.2020 18:00 1.06 1.26 0 0.20 321.62 434.27 311.97 1.92
01.11.2020 19:00 1.04 1.23 0 0.19 320.51 418.18 310.89 1.83
01.11.2020 20:00 1.05 1.21 0 0.16 314.14 414.06 307.63 1.55
01.11.2020 21:00 1.05 1.17 0 0.12 313.27 412 303.87 1.19
01.11.2020 22:00 1.04 1.09 0 0.05 307.8 412 298.57 0.45
01.11.2020 23:00 1.06 1.10 0 0.04 317.14 369.77 290.98 0.33
… … … … … … … … … …
30.04.2021 00:00 0.95 0.93 0.02 0 356.72 367.42 329.91 0.14
30.04.2021 01:00 0.92 0.90 0.02 0 356.72 367.42 341.46 0.01
30.04.2021 02:00 0.68 0.83 0 0.15 356.34 367.03 341.09 0.60
30.04.2021 03:00 0.80 0.83 0 0.03 356.11 366.79 330.77 0.13
30.04.2021 04:00 0.85 0.84 0.01 0 355.12 365.77 320.22 0.76
30.04.2021 05:00 0.85 0.83 0.02 0 353.51 364.12 300.70 0.74
30.04.2021 06:00 0.95 0.95 0 0 351.61 362.16 198.85 0.24
30.04.2021 07:00 1.02 1.02 0 0 352.03 362.59 201.76 0.01
30.04.2021 08:00 1.17 1.17 0 0 355.95 366.63 226.93 0.19
30.04.2021 09:00 1.22 1.22 0 0 355.76 366.43 234.74 0.51
30.04.2021 10:00 1.27 1.23 0.04 0 356.20 366.89 234.74 0.38
30.04.2021 11:00 1.28 1.20 0.08 0 357.21 367.93 291 0.85
30.04.2021 12:00 1.31 1.22 0.09 0 353.03 363.62 303.64 0.89
30.04.2021 13:00 1.24 1.16 0.08 0 352.43 363.00 269.11 0.82
30.04.2021 14:00 1.20 1.24 0 0.04 358.23 368.98 330.99 1.15
30.04.2021 15:00 1.14 1.19 0 0.05 358.76 381.09 348 0.58
30.04.2021 16:00 1.29 1.20 0.09 0 355.01 365.66 344.36 0.89

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

232
30.04.2021 17:00 1.23 1.17 0.06 0 356.01 366.69 296.82 0.62
30.04.2021 18:00 1.15 1.12 0.03 0 359.62 370.41 310.03 0.30
30.04.2021 19:00 1.13 1.14 0 0.01 376.71 388.01 365.41 0.09
30.04.2021 20:00 1.10 1.13 0 0.03 372.17 383.34 346.29 0.68
30.04.2021 21:00 1.06 0.94 0.12 0 369.02 381.10 357.95 1.41
30.04.2021 22:00 1.01 0.87 0.14 0 358.42 410.97 347.67 7.20
30.04.2021 23:00 0.97 0.95 0.02 0 355.04 386.29 344.39 0.64

Table 6: Comparison of Key Data

Simple
Simple Moving Weighted Moving Exponential Hospital Current
Data
Average Method Average Method Smoothing Status
Method
Total Real
Consumption 5.140,60 5.140,60 5.140,60 5.140,60
(mwh)
Total Forecasting
Consumption 4.778,98 4.779,23 4.774,35 5.647,2
(mwh)
Total Negative
182,69 172,67 220,78 38,25
EDM (mwh)
Total Positive
176,24 179,38 222,61 912,82
EDM (mwh)
Total Imbalance
10.803,88 10.519,55 13.254,84 29.456,47
Cost (tl)

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

233
 RESULTS

Businesses (like a hospital) have to put the concept of cost improvement on their agenda in order to survive or
grow in the sector. Energy costs are one of the most basic expenses of enterprises and at the same time they
have the greatest monetary value.

Forecastings and imbalances of the first and last days of the month are shown in Table 6. In the first stage, the
total forecasted consumption amount was found to be 4.778,98 mwh, according to hourly consumption
forecastings obtained by simple moving average method based on 6 month data. The actual consumption
amount is 5.140,60 mwh. The imbalance cost was realized as 10.803.88 TL with the first method. With the
weighted moving average method, the total consumption forecasted amount is 4.779,23 mwh and the
imbalance cost is 10.519,55 TL. With the simple exponential smoothing method, the total consumption
forecasted amount is 4.774,35 mwh and the imbalance cost is 13.254,84 TL. This is the most appropriate
forecasting method for the hospital, as the lowest imbalance and MAPE ratio is provided by weigted moving
average method.

When these comparisons and forecasting errors are examined, it is seen that the forecasting made according to
the relevant hours of the day is closer to the truth. In all three analysis methods, an improvement was achieved
in the cost according to the current situation of the company. It was observed that this improvement was
higher in the forecasting made using the data of the relevant hour. In the next process, it will be correct to
make the consumption forecasting of the company with the relevant hour data that provides the highest cost
improvement.

As a result, it is aimed to save the energy cost of the enterprise with the cost improvement made in this study.
Cost improvement is a very important requirement for businesses to survive in market conditions. Forecasting
the consumptions closest to the real value by using correct forecasting methods forms the basis of cost
improvement. In the light of this study, many companies will be able to achieve a significant improvement in
energy supply costs.

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

234
 REFERENCES

[1] Başoğlu B., Bulut M., Kısa Dönem Elektrik Talep Tahminleri İçin Yapay Sinir Ağları Ve Uzman
Sistemler Tabanlı Hibrit Sistem Geliştirilmesi, Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi,
2017, 32(2), 575-583.

[2] EPİAŞ Web Sitesi. https://www.epias.com.tr/epias-kurumsal/hakkimizda

[3] Resmi Gazete, Elektrik Piyasası Dengeleme Ve Uzlaştırma Yönetmeliğinde Değişiklik Yapılmasına Dair
Yönetmelik, https://www.resmigazete.gov.tr/eskiler/2019/08/20190809-4.htm

[4] Gün Öncesi Piyasası.

https://www.google.com/search?q=g%C3%BCn+%C3%B6ncesi+piyasas%C4%B1&source=lmns&bih=657&
biw=1366&hl=tr&sa=X&ved=2ahUKEwiYwe6OzJzyAhWM-qQKHU4IBW0Q_AUoAHoECAEQAA

[5] Ceyhan G., Türkiye’de Elektrik Piyasa Takas Fiyatı Ve Sistem Marjinal Fiyatı Farkı Üzerine İstatistiksel
Bir Çalışma, EPİAŞ, https://blog.metu.edu.tr/e162742/files/2016/08/PTF_vs_SMF_original.pdf

[6] GİP Kullanım Kılavuzu, https://www.epias.com.tr/wp-content/uploads/2019/04/G%C4%B0P-

Kullan%C4%B1c%C4%B1-Klavuzu_05.04.2019.pdf

[7] Biçen Y., Elektrik Enerji Piyasasında Arz-Talep Dengesi ve Fiyat Etkileşimi, 4. Uluslararası Mühendislik
ve Bilimde Yenilikçi Teknolojiler Sempozyumu, Antalya, Türkiye, 3-5 Kasım 2016.

[8] Akdi Y., (2003). Zaman Serileri Analizi

[9] Demez F., Hastanelerde Kapasite Ölçümü Talep Tahmin Yöntemleri ve Bir Araştırma, Yüksek Lisans
Tezi, Beykent Üniversitesi, Sosyal Bilimler Enstitüsü, İstanbul, 2016, 450024.

[10] Çağlar, T. (2007). Talep Tahmininde Kullanılan Yöntemler ve Fens Teli Üretimi Yapan Bir İşletmede
Uygulanması, (Yüksek Lisans Tezi), Kırıkkale: Kırıkkale Üniversitesi , Fen Bilimleri Enstitüsü.

[11] Üreten, S. (2005). Üretim/İşlemler Yönetimi, Stratejik Kararlar ve Karar Modelleri, Ankara: Gazi
Kitabevi.

[12] Aslay F., Özen Ü., Meteorolojik Paramatreler Kullanılarak Yapay Sinir Ağları İle Toprak Sıcaklığının
Tahmini, Politeknik Dergisi, 2013, 16(4), 139-145.

© 19th International Logistics and Supply Chain Congress

October 20-21, 2021, Gaziantep, Turkey

235
 A MIXED INTEGER PROGRAMMING MODEL FOR SUPPLY CHAIN
NETWORK DESIGN BASED ON NET PRESENT VALUE CONCEPT
Rahmi Baki1

Abstract - Today, businesses gain a competitive advantage against their competitors thanks to the
efficiency of the supply chain networks they have designed. In the current research, a dynamic and
multi-product supply chain network including supplier, factory, warehouse and customer components
is designed. Through the proposed mixed integer programming model (MIPM), it is aimed to make
strategic decisions such as raw materials to be purchased, products to be produced and stored, and
distribution planning. Businesses avoid investment decisions that require high investment costs,
especially in periods when inflation and interest rates are high. Therefore, while making investment
decisions, the total profit should be analysed by considering the concept of net present value.
However, the concept of net present value is not generally explored in the supply chain network design
literature. In the model proposed in the research, production and distribution decisions are taken in a
way that maximizes total profit in terms of net present value, and thus it is aimed to design the supply
chain network in a way that will provide competitive advantage to the companies. In the study, a
sample application was solved through the GAMS package program. It is expected that the results
obtained in the study will benefit companies that aim to activate their supplier chain strategies,
especially in the high interest and inflation environment.
Keywords: Mixed integer programming model, Supply chain, supply chain network design.

INTRODUCTION
Today, there are many factors that affect the performance of companies. The most important of these
factors can be listed as customer satisfaction, product and service quality and distribution costs
(Bashiri, Badri & Talebi, 2012). Businesses that want to keep production and distribution costs at a
low level and not to lag behind their competitors in the market have to manage their supply chains
effectively. The supply chain is a process that establishes the link between the supplier and the
customer, from the purchase of raw materials to the consumption of final products (Lummus &
Vokurka, 1999). Today, supply chains have become more complex than in the past with their wide
geographical scope (Tolooie, Maity & Sinha, 2020). Managing this complex process can be achieved
with an effective supply chain network (SCN) design.

The most important decisions made in SCN design are supplier, production and shipping preferences.
A SCN includes many choices, such as from which suppliers to buy raw materials, which
transportation method to use, or which customer group to serve. These decisions must be made
effectively so that products or services can reach suppliers from suppliers to factories and customers at
a low cost and smoothly (Kheirabadi, Naderi, Arshadikhamseh & Roshanaei, 2019). Therefore, many
mathematical models have been proposed for generation and distribution network design in the
literature.

In the study, a mixed integer programming model (MIPM) was developed to design a multi-product
and dynamic SCN. In the study, the production and distribution network consists of four components:
supplier, factory, warehouse and customer. With the developed approach, it is aimed to make
decisions such as supplier selection, raw material quantities to be purchased from suppliers, product
quantities to be produced in factories and product quantities to be transferred. The study aims to make
production and distribution decisions for the SCN in future scenarios, taking into account net present
value (NPV) in a way that maximizes profit.

1
Aksaray University, Department of Management Information Systems, Aksaray, Turkey, [email protected].

236
NPV is the discounted value of a future cash inflow or outflow with a certain interest rate (Gürsoy,
2007). The concept of today refers to the date when the main transaction took place (Dabbagoğlu,
2008). NPV is a method that is frequently used to evaluate the profitability of current and potential
investments and enables effective financial decisions to be made. The NPV method analyses the flows
of the project in question over a period of time and converts them into equivalent cash flows using
economic factors (Remer & Nieto, 1995). While choosing between alternatives, the alternative with
the highest total NPV will maximize the total value (Crundwell, 2008).

In the objective function (OF) of the study, the total profit is maximized by considering the NPV.
Especially in countries with high inflation rate, the SCN needs to be designed taking NPV into
account. Evaluating the total profit to be obtained in different investment scenarios from the NPV
perspective and taking decisions accordingly will provide companies with a competitive advantage
and increase their profitability. The proposed model allows the SCN to be planned in a way that
maximizes the NPV of profit, taking into account the variation of profitability over time and the
interest rate in the relevant country. Especially in scenarios where investment costs and inflation rate
are high, it becomes more important to design the SCN by evaluating it in terms of NPV. Under such
conditions, companies' strategic investment decisions may change directly and businesses may delay
important investment decisions such as establishing a new production facility, purchasing physical
equipment or warehouse center. For this reason, especially large-cost investment decisions should be
planned considering NPV. However, in the literature on the subject, the concept of NPV has not
generally been taken into account. In the current study, it is aimed to fill this gap in the literature and
to contribute to companies that want to develop their supply chain strategies, especially in a high
inflation environment.

The remainder of the research is organized as given below. In the second chapter, researches on SCN
design in the literature are presented. In the third chapter, the proposed model is introduced and
explained. In the fourth chapter, a case study is solved and the findings are evaluated. In the fifth
chapter, the results obtained from the study, suggestions for future research and limitations of the
study are presented.

LITERATURE REVIEW
The success of supply chain management in a company can only be achieved by making effective
production and distribution decisions. For this purpose, many researchers have recently proposed
various mathematical models for use in SCN design. Some of the studies on the subject are included in
this part of the research.

Özceylan, Paksoy, and Bektaş (2014) defined a MIPM that optimizes decisions to be made in closed-
loop supply chain problems. In the study, it is aimed to minimize transportation, purchasing and
renewal costs. The OF is defined in a way that minimizes transportation costs between all levels,
purchasing costs of subgroups and replacement costs of products for operating workstations. Dotoli,
Epicoco & Falagario (2015) proposed an integrated approach of Data Envelopment Analysis and fuzzy
set theory to manage SCN design problems by considering imprecise input and output data. After the
actors of each stage are listed, it is targeted to maximize the efficiency of the SCN by considering the
current capacity and customer demands. In the OF, it is aimed to maximize the defined ratio by
considering the weighted output sum and weighted input sum. Mousazadeh, Torabi & Zahiri (2015)
developed a bi-objective MIPM for SCN design in the pharmaceutical industry. The proposed
approach will contribute to strategic decisions such as the opening of production centers and tactical
decisions such as flow between distribution centers. The first OF aims to minimize installation costs,
production costs, transportation costs and inventory costs, while the second OF aims to minimize
unmet demand. Jeihoonian, Zanjani & Gendreau (2016) designed a closed-loop SCN for long-life
products with different quality levels. To achieve this aim, a MIPM is proposed based on the general
understanding that the number of each sub-assembly depends on the quality condition.

237
Mirmajlesi & Shafaei (2016) presented an integrated approach to solve a multi-period, multi-product
and multi-stage supply chain problem for short-lived products. In the study, the case problem was
solved by using Tabu Search Algorithm, Differential Evolution Algorithm and MIPM. In the OF, it is
aimed to minimize the supplier costs, recovery and collection centers. Nemati, Madhoshi &
Ghadikolaei (2017) investigated the impact of sales and operations planning on the overall
performance of the supply chain. The proposed model has been tested on a dairy manufacturing
company with several raw material suppliers, distribution centers, production facilities and a wide
variety of product families. In the study, the OF aims to optimize the total profit obtained by
subtracting the production, installation, inventory, return, shipping, purchasing and ordering costs
from the total profit from sales. Shi, Liu, Tang, and Xiong (2017) describe a model in which overall
costs, carbon emission and network sensitivity are optimized using Genetic Algorithm and MIPM.
Two OFs are presented in the proposed model. The first OF is to minimize the overall cost of the
system, which consists of facility opening, operating and transportation costs of facilities. In the
second OF, it is aimed to minimize the overall carbon emissions that occur in the production, storage
and transportation of products. In the study by Varsei & Polyakovskiy (2017), a general model was
developed for SCN design, covering economic, environmental and social objectives. Model was tested
on a large-scale wine company in Australia. Three OFs were used in the study. In the first OF, which
specifies the economic targets of the firm, transportation, production, purchasing and storage costs are
minimized. In the second OF, which specifies environmental targets, emissions produced in all
transport activities along the supply chain are taken into account. In the third OF, which shows the
social objectives, the total social impact of the relevant suppliers and facilities in the SCN is
calculated.

Dehghan, Nikabadi, Amiri & Jabbarzadeh (2018) proposed a hybrid model for closed-loop SCN
design under uncertain conditions. The proposed approach has been tested on a real case study in an
olive oil supply chain. In the OF, an OF that minimizes contracting, opening, raw material,
transportation and processing costs with suppliers is used. Patil, Vishnu & Sridharan (2018) developed
a supply chain approach that considers different dimensions of reliability. Accordingly, a MIPM in
which a four-stage SCN is designed has been proposed. The OF of the SCN model minimizes total
production, transportation and fixed costs. Kheirabadi, Naderi, Arshadikhamseh & Roshanaei (2019)
developed a Lagrange-based MIPM considering quantity discount and shipping mode selection. In the
OF of the model, the total cost consisting of operating costs, transportation costs and raw material
costs is minimized. Sinaki, Sadeghi, Suer & Çelikbilek (2019) presented a model to integrate
production planning and sustainable SCN. In the study, the SCN consists of two stages, including a
factory adopting the cellular production system and customers. In the proposed model, the OF
optimizes the sum of the satisfaction levels consisting of profit, early access and carbon emission
targets.

In the study developed by Sangaiah, Tirkolaee, Goli & Dehnavi-Arani (2020), a MIPM using Cuckoo
Optimization Algorithm is proposed in order to provide liquefied natural gas sales planning for a
certain time period. In the proposed model, it is aimed to minimize sales costs. Tolooie, Maity &
Sinha (2020) proposed a two-stage MIPM for designing reliable SCNs under unexpected disruptions
and demand uncertainty. Thus, it is aimed to strengthen the selected facilities in the face of capacity-
related problems, facility interruptions and uncertain demand. In the OF, the total cost, consisting of
the cost of establishing facilities and distribution centers, the transfer cost, and the penalty cost of not
meeting customer demand, is minimized. Fakhrzad & Goodarzian (2021) developed a multi-objective
mathematical model for citrus SCN design using Ant Colony Optimization and Simulated Annealing
approaches. In the OF of the study, the total cost consisting of transportation cost, operating cost,
inventory holding cost, waste cost and storage cost is minimized. Hajimirzajan et al. (2021), a strategic
planning framework was presented for the realization of crop planning and agri-food supply chain
management in terms of sustainable climate. In the proposed model, the OF minimizes the sum of
planting, storage, transportation, cold storage rental and penalty costs.

In the studies in the literature, it has been seen that different OFs are suggested for an effective SCN
design. In the OFs of the proposed models, targets such as reducing general costs, increasing total

238
income, reducing carbon emissions and meeting social objectives are used. However, the concept of
NPV, which has a critical importance especially when there is a high inflation rate and big investment
decisions are evaluated, has often been ignored. In the literature review, only two studies were found
that included the concept of NPV in the model they proposed. Hombach, Cambero, Sowlati, and
Walther (2016) presented a model for planning second-generation biofuel supply chains, taking into
account the policy tools of European biofuel regulations. Bashiri, Badri & Talebi (2012), developed a
system that will contribute to the strategic and tactical decisions of companies. In the study, it is aimed
to take final decisions to maximize profit in the NPV perspective in order to plan an effective SCN,
and as a result, to gain competitive advantage to businesses.

PROPOSED MODEL
In this part of the study, the proposed MIPM for planning a dynamic generation-distribution network
is introduced. The production and distribution network in the proposed model consists of supplier,
factory, warehouse and customer components. In the developed approach, potential suppliers for raw
materials, different factory alternatives for production, different warehouses for product storage and
customer sets at different distances to potential warehouses were considered.

In the proposed model, decisions such as the selection of suppliers and the amount of raw materials to
be supplied from the suppliers, the amount of products to be produced in the factories, the amount of
products to be transported from each factory to each warehouse and the amount of products to be
transported from each warehouse to each customer will be reached. The assumptions of the model are
as follows; each supplier, factory and warehouse has a limited capacity, stocks can only be kept in
warehouses, there is no stock in any warehouse at the beginning of the period, product transfer
between warehouses is not allowed and products cannot be delivered to customers from factories. In
addition, all customer demands must be met. The indices, parameters and decision variables used for
mathematical modelling of the problem are as follows;

N (n∈N): Set of strategic periods

T(t∈T): Set of suppliers

F(f∈F): Set of factories

D(d∈D): Set of warehouses

M(m∈M): Set of customers

U(u∈U): Set of products or raw materials

: Capacity of supplier t at period n

: Capacity of factory f at period n

: Capacity of warehouse d at period n

: Demand of customer m for product u at period n

: Total profit at period n

FO: Interest rate

: Selling price of the product u

: Price of product u at supplier t

239
 : Opening cost of factory f

: Fixed cost of factory f

: Variable cost of product u at factory f

: Storage cost of product u at warehouse d

: Transportation cost of product u from supplier t to factory f

: Transportation cost of product u from factory f to warehouse d

: Transportation cost of product u from warehouse d to customer m

: 1 if the supplier t is selected at period n; 0 otherwise

: 1 if the factory f is active at period n; 0 otherwise

: 1 if the warehouse d is active at period n; 0 otherwise

: Quantity of product u transferred from supplier t to factory f at period n

: Quantity of product u transferred from factory f to warehouse d at period n

: Quantity of product u transferred from warehouse d to customer m at period n

: Quantity of product u produced in factory f at period n

: Quantity of product u held in warehouse d at period n

The mathematical formulation of the proposed model is presented as shown below;

Objective Function (OF);

Maximize Z = ∑ ∈ . (1)

In the OF, it is aimed to maximize the total profit. Equation (1) indicates the OF in which the profits in
NPV for each period are maximized.

Constraints;

∑ ∈ = , ∀m ∈ M, ∀u ∈ U, ∀n ∈ N. (2)

Equation (2) states that the products transferred to the customer in all periods must be equal to the
customer demand.

+∑ ∈ ∑ ∈ =∑ ∈ ∑ ∈ + , ∀d ∈ D, ∀n ∈ N. (3)

Equation (3), which shows the flow balance in the warehouses, shows that the sum of the amount of
product kept in the warehouse in the previous period and the amount of product coming from the
factory to the warehouse in that period should be equal to the sum of the product amount going from
the warehouse to the customer and the amount of the product remaining in the warehouse in the
current period.

∑ ∈ ∑ ∈ =∑ ∈ , ∀f ∈ F, ∀n ∈ N. (4)

Equation (4) states that the required raw material must be ordered from the supplier in order to
produce the desired amount of product in a factory.

240
 =∑ ∈ , ∀f ∈ F, ∀u ∈ U, ∀n ∈ N. (5)

Equation (5) shows that the amount of product produced in a factory in a period should be equal to the
amount of product delivered from that factory to the warehouses. That is, products should not be
stored in factories.

∑ ∈ ≤ . , ∀f ∈ F, ∀n ∈ N. (6)

Equation (6), states that a factory should not produce more than the capacity of that factory.

∑ ∈ +∑ ∈ ∑ ∈ ≤ . , ∀d ∈ D, ∀n ∈ N. (7)

Equation (7), emphasizes that more products should not be stored in a warehouse than the capacity of
that warehouse.

≤ . , ∀t ∈ T, ∀u ∈ U , ∀n ∈ N. (8)

Equation (8), ensures that the quantity ordered from a supplier does not exceed that supplier's capacity.

=∑ ∈ ∑ ∈ ∑ ∈ . (9)

∑ ∈ . (10)

∑ ∈ . (11)

∑ ∈ ∑ ∈ (12)

∑ ∈ ∑ ∈ (13)

∑ ∈ ∑∈ ∑ ∈ . +∑ ∈ ∑ ∈ ∑ ∈ . +
∑ ∈ ∑ ∈ ∑ ∈ . (14)

∑ ∈ ∑ ∈ ∑ ∈ , ∀n ∈ N. (15)

Equation (9) total revenue from products delivered to customers, Equation (10) total cost of opening
factories, Equation (11) fixed cost of factory operation, Equation (12) variable cost related to the
amount of product produced in the factory, Equation (13) ) shows the cost due to the products kept in
the warehouses, Equation (14) shows the transportation cost as a result of transportation between the
supplier-factory-warehouse-customer, and Equation (15) shows the total raw material cost. Thus, net
profit is obtained as a result of subtracting costs from total income.

≤ , ∀f ∈ F. (16)

Equation (16), prevents a factory from closing in future periods.

∈ (0,1). (17)

∈ (0,1). (18)

∈ (0,1). (19)

≥ 0. (20)

≥ 0. (21)

≥ 0. (22)

241
Equations (17), (18) and (19) indicate that variables must be binary. Equation (20), (21) and (22)
prevent the variables from taking negative values.

APPLICATION
In order to observe the application of the proposed model, the model is tested on a numerical example.
In this example; 3 suppliers (T1, T2, T3), 2 factories (F1, F2), 4 warehouses centers (D1, D2, D3, D4),
4 customers (M1, M2, M3, M4), and 2 product types (U1, U2) is assumed. Detailed information about
the sample (capacity, demand, selling price and cost) is given in the Appendix A.

In practice, the production-distribution plan for the 3-month planning period was determined and the
monthly interest rate was accepted as 0.014%. Production and distribution decisions are made with the
aim of maximizing profits, taking into account NPV. The application of the model is solved in the
GAMS/CPLEX package program.

The quantities of raw materials to be supplied from suppliers and sent to factories are presented in
Table 1. For example, in the first period; 10000 raw materials will be taken from T1 for U1, 5000 of
which will be moved to F1 and 5000 to F2. 10000 raw materials taken from T2 for U1 will be moved
to F2, and 20000 raw materials from T3 for U2 will be transferred to F2.

Table 1. Amount of raw materials to be sent from suppliers to factories

Period 1 Period 2 Period 3
Supplier Product
F1 F2 F1 F2 F1 F2
U1 5000 5000 6000 4000 5000 5000
T1
U2 - - - - - -
U1 - 10000 - 5000 - 10000
T2
U2 - - - - - -
U1 - - 5000 - - -
T3
U2 20000 - 4000 6000 15000 -

The amount of products produced in the factories are given in Table 2. For example, in the first period,
5000 units of U1, 15000 units of U1 and 20000 units of U2 will be produced in F2. In addition, the
amount of products to be sent from the factories to the warehouses is given in Table 3. In the first
period, 5000 U1 and 10000 U2 will be moved from F1 to D2. 10000 units of U2 will be sent from F1
to D3. From F2, 13000 U1 will be moved to D1, and 2000 to D3. However, 10000 U2 will be retained
at the end of the first term in D3. At the end of the second period, the amount of product in the
warehouse is 5000.

Table 2. Amount of products to be produced in factories

Factory Product Period 1 Period 2 Period 3
U1 5000 11000 5000
F1
U2 - 4000 15000
U1 15000 9000 15000
F2
U2 20000 6000 -

242
 Table 3. Amount of products to be sent from factories to warehouses
Period 1 Period 2 Period 3
Factory Product
D1 D2 D3 D1 D2 D3 D1 D2 D3
U1 5000 5000 - 6000 4000 - 5000 5000 -
F1
U2 - - - - - - - - -
U1 - 10000 - - 5000 - - 10000 -
F2
U2 - - - - - - - - -
The products delivered from the warehouses to the customers are given in Table 4. When the first
period is examined, it is understood that 6000 units of U1 will be sent from D1 to M1, 4000 units to
M3 and 3000 units to M4. 5000 units of U1 will be delivered from D2 to M2. In addition, 2000 units
of U2 will be delivered from D2 to M1, 3000 units to M2, 4000 units to M3 and 1000 units to M4.
2000 U1 will be transmitted from D2 to M2. As a result, the profit at the end of the three periods,
taking into account NPV, was calculated as TL 43568444.

Table 4. Quantities of products to be sent from warehouses to customers

Warehous Produ Period 1 Period 2 Period 3
e ct M1 M2 M3 M4 M1 M2 M3 M4 M1 M2 M3 M4
600 400 300 600 300 500 500 300 200
U1 - - -
0 0 0 0 0 0 0 0 0
D1
100 500 600
U2 - - - - - - - - -
0 0 0
500 700 700 400 500
U1 - - - - - - -
0 0 0 0 0
D2
200 300 400 100 400 500 100 300
U2 - - - -
0 0 0 0 0 0 0 0
200
U1 - - - - - - - - - -
0
D3
300 200 500
U2 - - - - - - - - -
0 0 0

CONCLUSIONS AND DISCUSSION

Effective management of the supply chain can only be achieved by taking the entire system into
account and making decisions in this perspective. These decisions are very important for companies to
survive in today's competitive business world. For this reason, companies should make their supplier,
production and distribution decisions through an efficient SCN they designed. In this study, a MIPM
has been developed, which aims to design a multi-product, dynamic and efficient production and
distribution network with supplier, factory, warehouse and customer components.

In the proposed model, decisions are made about how much raw material will be purchased from
which supplier, the type and amount of product to be produced in the factories, how much product will
be transported from which factory to which warehouse, how much product will be stored in the
warehouses and how much product will be delivered to the customers from which warehouses. The
application of the model is analyzed on a numerical example. In the numerical example, there are 2
product types, 3 suppliers, 2 factories, 4 warehouse centers and 4 customers. However, the monthly
interest rate was accepted as 0.014% and production and distribution planning was carried out for a 3-
month planning period. The sample application is solved through the GAMS package program and
order, production, storage and distribution decisions are presented for each period.

In the OF of the study, the total profit is defined by considering the NPV. Thus, the proposed model
ensures that the SCN is planned in a way that maximizes the present value of profit, taking into
account the time-dependent variation of profitability and the interest rate in the relevant country.

243
Especially in cases where investment decisions are discussed and there is a high inflation environment,
it becomes more important to design the SCN by evaluating it in terms of NPV. Under such
conditions, investment decisions of firms may directly change. In the current research, it is aimed that
enterprises take their production and distribution decisions in a way that maximizes their profits by
considering the NPV understanding and design their SCNs effectively.

The model proposed in this study can be improved by adding different dimensions of the supply chain.
For example, in future studies, different investment scenarios in different sectors can be analyzed in a
way that maximizes total profit in terms of NPV. In addition, considering some elements such as legal
environmental regulations or the concept of reverse logistics will contribute to the literature. However,
adding the constraints considering demand uncertainty and quantity discount can be considered as the
next step of the research. In addition to these, some heuristic methods can be used to shorten the
solution time of large-scale problems.

244
Appendix A. Application Information

, ,
n=1 n=2 n=3 n=1 n=2 n=3 500 250 350 10,000
30,000 20,000 25,000 6,000 7,000 10,000 750 150 250 15,000
10,000 5,000 10,000 2,000 4,000 5,000 200 200
20,000 15,000 15,000 7,000 6,000 6,000
25,000 25,000 25,000 3,000 4,000 5,000
15,000 15,000 15,000 4,000 3,000 3,000 2,000 30 5 6
25,000 25,000 25,000 4,000 5,000 6,000 3,000 20 4 5
15,000 15,000 15,000 3,000 4,000 2,000 35 3 4
10,000 10,000 10,000 1,000 2,000 3,000 25

(u = 1) (u = 2) (u = 1) (u = 2)
f=1 f=2 f=1 f=2 d=1 d=2 d=3 d=1 d=2 d=3
t=1 14 20 t=1 16 22 f=1 4 2 3 f=1 4 2 3
t=2 22 16 t=2 24 18 f=2 3 5 4 f=2 3 5 4
t=3 18 26 t=3 30 28

(u = 1) (u = 1) FO
m=1 m=1 m=1 m=4 m=1 m=1 m=1 m=4 FO 0,014
d=1 8 6 5 7 d=1 8 6 5 7
d=2 6 8 9 7 d=2 6 8 9 7
d=3 10 8 9 7 d=3 10 8 9 7

245
References

Bashiri, M., Badri, H., & Talebi, J. (2012). A new approach to tactical and strategic planning in production–
distribution networks. Applied Mathematical Modelling, 36(4), 1703-1717.
Crundwell, F. (2008). Finance for engineers: Evaluation and funding of capital projects. Springer Science &
Business Media.
Dabbağoğlu, K. (2008). Finansal kiralama işlemlerinde net bugünkü değer hesaplamaları. Mali Çözüm
Dergisi/Financial Analysis, 87, 51-66.
Dehghan, E., Nikabadi, M. S., Amiri, M., & Jabbarzadeh, A. (2018). Hybrid robust, stochastic and possibilistic
programming for closed-loop supply chain network design. Computers & Industrial Engineering, 123, 220-231.
Dotoli, M., Epicoco, N., & Falagario, M. (2015). A technique for supply chain network design under uncertainty
using cross-efficiency fuzzy data envelopment analysis. IFAC-PapersOnLine, 48(3), 634-639.
Gürsoy, C. T. (2007). Finansal yönetim ilkeleri. Doğuş Üniversitesi Yayınları.
Fakhrzad, M. B., & Goodarzian, F. (2021). A new multi-objective mathematical model for a Citrus supply chain
network design: Metaheuristic algorithms. Journal of Optimization in Industrial Engineering, 14(2), 127-144.
Hajimirzajan, A., Vahdat, M., Sadegheih, A., Shadkam, E., & El Bilali, H. (2021). An integrated strategic
framework for large-scale crop planning: sustainable climate-smart crop planning and agri-food supply chain
management. Sustainable Production and Consumption, 26, 709-732.
Hombach, L. E., Cambero, C., Sowlati, T., & Walther, G. (2016). Optimal design of supply chains for second
generation biofuels incorporating European biofuel regulations. Journal of cleaner production, 133, 565-575.
Jeihoonian, M., Zanjani, M. K., & Gendreau, M. (2016). Accelerating Benders decomposition for closed-loop
supply chain network design: Case of used durable products with different quality levels. European Journal of
Operational Research, 251(3), 830-845.
Kheirabadi, M., Naderi, B., Arshadikhamseh, A., & Roshanaei, V. (2019). A mixed-integer program and a
Lagrangian-based decomposition algorithm for the supply chain network design with quantity discount and
transportation modes. Expert Systems with Applications, 137, 504-516.
Lummus, R. R., & Vokurka, R. J. (1999). Defining supply chain management: A historical perspective and
practical guidelines. Industrial Management & Data Systems, 99(1), 11-17.
Mirmajlesi, S. R., & Shafaei, R. (2016). An integrated approach to solve a robust forward/reverse supply chain
for short lifetime products. Computers & Industrial Engineering, 97, 222-239.
Mousazadeh, M., Torabi, S. A., & Zahiri, B. (2015). A robust possibilistic programming approach for
pharmaceutical supply chain network design. Computers & Chemical Engineering, 82, 115-128.
Nemati, Y., Madhoshi, M., & Ghadikolaei, A. S. (2017). The effect of Sales and Operations Planning (S&OP) on
supply chain’s total performance: A case study in an Iranian dairy company. Computers & Chemical
Engineering, 104, 323-338.
Özceylan, E., Paksoy, T., & Bektaş, T. (2014). Modeling and optimizing the integrated problem of closed-loop
supply chain network design and disassembly line balancing. Transportation Research Part E: Logistics and
Transportation Review, 61, 142-164.
Patil, P., Vishnu, C. R., & Sridharan, R. (2018). Reliable supply chain network design model: Development and
analysis. Materials Today: Proceedings, 5(11), 23549-23558.
Remer, D. S., & Nieto, A. P. (1995). A compendium and comparison of 25 project evaluation techniques. Part 1:
Net present value and rate of return methods. International Journal of Production Economics, 42(1), 79-96.
Sangaiah, A. K., Tirkolaee, E. B., Goli, A., & Dehnavi-Arani, S. (2020). Robust optimization and mixed-integer
linear programming model for LNG supply chain planning problem. Soft Computing, 24(11), 7885-7905.
Shi, J., Liu, Z., Tang, L., & Xiong, J. (2017). Multi-objective optimization for a closed-loop network design
problem using an improved genetic algorithm. Applied Mathematical Modelling, 45, 14-30.
Sinaki, R. Y., Sadeghi, A., Suer, G., & Çelikbilek, C. (2019). A weighted multi-objective mathematical model
for cell scheduling and environmentally sustainable supply chain network. Procedia Manufacturing, 39, 1559-
1566.
Tolooie, A., Maity, M., & Sinha, A. K. (2020). A two-stage stochastic mixed-integer program for reliable supply
chain network design under uncertain disruptions and demand. Computers & Industrial Engineering, 148,
106722.
Varsei, M., & Polyakovskiy, S. (2017). Sustainable supply chain network design: A case of the wine industry in
Australia. Omega, 66, 236-247.

246
 FACTORS AFFECTING GREEN LOGISTICS POLICIES AND THEIR
EFFECTS ON LOGISTICS PERFORMANCE INDEX: COMPARISON OF
THE EUROPEAN UNION AND TURKEY

1
Abdullah n , Abdurrahim Furkan Mercimek2

Abstract ⎯ Globalization and increasing competitiveness have caused the logistics industry to play an
important role in international trade. Developments in logistics and transportation infrastructure are of vital
importance in terms of increasing trade volume and international competitiveness. The increase in logistics
activities, on the other hand, creates a negative impact on the environment with the increase in greenhouse
gas and carbon dioxide emissions. It is aimed to establish a sustainable balance between green logistics and
environmental, social and economic targets, which are created to reduce the negative impact of logistics
activities. In this study, as an introduction to the subject, the factors that can affect the decisions and policies
taken in the field of green logistics are emphasized and the effects of these factors on the supply chain are
discussed. Subsequently, the policies implemented in the context of the transportation sector are examined in
terms of the European Union and Turkey, and the factors affecting the ranking of the countries in the logistics
performance index are examined and it is discussed why Turkey's ranking is behind in this regard. In the
research findings, it has been determined that the logistics performance index rankings between the union
countries that implement the same policies are different. In the research part of the study, the reports
published by the institutional transport structures of the European Union and Turkey, the publications of
international organizations and academic literature were scanned.
Keywords: Green Logistic, Supply Chain Management, Transportation Policies, EU, Turkey

INTRODUCTION
Logistics, its importance is increasing day by day, is defined as the process of planning, implementing,
transporting, storing and controlling the movement of all kinds of products, services and information from the
starting point to the end point in an effective and efficient manner in order to meet the needs of customers and
consumers. It can also be expressed as providing the right product in the right quantity, in the right form, at
the right time, from the right source, in the right way, at the right price and at a competitive price (Takım,
2015).
A significant majority of companies delegate their logistics operations to service providers who are
experts in this field in order to focus on their core activities, facilitate their transactions and use their resources
more efficiently. Logistics outsourcing is the use of a third-party service provider in a specified contractual
relationship in the execution of all or some logistics activities of the enterprise (Sohail et all., 2004). The
widespread use of outsourcing and the gradual increase in logistics activities have escalated the importance of
companies operating in this field day by day. Changes in the sector, new trends such as the effect of
outsourcing on the sector are exacerbating the existing competitive environment.
Along with the brisk competition environment, the rapid and uncontrolled economic growth in
industrialized and industrializing countries brings with it the environmental effects of urbanization. In the 20th
century, industrialization caused the natural environment to change rapidly and a new social environment to
emerge, because industrialization was not planned, only industrialization was targeted, and the environmental
factor was ignored (Guha, 2000). As a result, climate changes have emerged and natural resources have
started to deplete rapidly. The change of seasons, the melting of glaciers, the destruction of forests, the
pollution of the seas, the decrease in the air quality we breathe, the increase in noise, concrete and metal
pollution are among the biggest indicators of this process (Baykal and Baykal, 2008). Increasing

1
Necmettin Erbakan University, Konya, Turkey, [email protected]
2
Necmettin Erbakan University, Institute of Social Sciences, International Marketing and Logistics Management, Konya, Turkey,
[email protected]

247
environmental awareness with the emergence of such problems led businesses to reduce the rate of waste by
designing production systems that cause minimal damage to the environment. However, depending on the
increase in globalization and industrialization, the struggle for competition between businesses has started.
Businesses that want to gain competitive advantage have turned to environmental issues. The reason for this is
that businesses that want to gain superiority in the competition struggle want to make profit as a result of their
work with the awareness of protecting nature (Vachon and Klassen, 2007). Since the destruction of natural
resources and the environment will not allow long-term growth-enhancing activities, it has become necessary
to create a growth and development model that is sensitive to the environment and natural resources.
In this direction, the logistics performance index, which is announced every two years by the World Bank,
which deals with performance indices from different perspectives in the logistics sector, which is among the
sectors most affected by negative natural disasters, plays a role in the sector's updating itself and keeping up
with the requirements of the age. Performance results, on the other hand, can be said to be the distinguishing
power of competition among countries in the world economy. It has become mandatory to use performance
measurement techniques in the context of allocating a healthy competitive environment (Ener, 2010).
In order to be in the first place in the competition between countries, a strong infrastructure system,
planning, investments, education, regional developments, keeping up with changes and technological
developments are needed. The LPI report provides a significant benefit in terms of what steps should be taken
in order for countries to self-assess, strategically plan for the future, draw a roadmap and gain competitive
advantage (Cezayirlioğlu, 2012).
The study is structured as follows. The introduction part defines the boundaries of the article. The second
section reviews the literature that addresses logistics, green logistics, logistic performance index in context of
Turkey and EU countries. Section III describes logistics and green logistics concepts. Section IV explains the
factors affecting the green supply chain. Section V details the Global Logistics Performance Index (LPI).
Section VI discuss Turkey and EU within the frame of LPI. Finally, section VII summarizes the main
conclusions of the research.

LITERATURE REVIEW
Logistics and green logistics have been the subject of many studies in the literature due to the
disappearance of the borders of trade in our world, which is attributed as a global village, and its becoming a
very important competitive position through performance measurement indices. In this section, studies related
to the Global Logistics Performance Index in the context of Turkey and the European Union, rather than the
studies related to logistics and green logistics in general, are discussed in terms of the subject of the study.
According to Çevik and Kaya (2010), although the term logistics is a concept of military origin, its history
goes back thousands of years. Logistics, which is defined as a more advanced expression of the concept of
transportation, is to take the product from the point of production, to keep it in warehouses and to stock it, to
deliver it to the desired place as desired, and to carry out all these operations in a planned, efficient and fast
manner.
According to Köksal (2012), logistics should not be considered only as a transportation activity. Logistics,
one of the cornerstones of international trade covers core activities involving various public and private sector
institutions ranging from transportation, warehousing, cargo consolidation, border controls to in-country
distribution and payment systems.
Sbihi and Eglese (2007) argued that green logistics is concerned with the sustainable production and
distribution of products, considering environmental and social factors.
Srivastava (2007) defined green supply chain management as integrating environmental thinking into
supply chain management, and stated that this definition includes issues such as material procurement and
selection, product design, production processes, delivery of the product to the customer and product life cycle
management.
The World Bank (2010) conducted a survey with one thousand logistics employees to evaluate and
compare the logistics performance of countries. Logistics performance score was calculated according to six
questions in the study, in which a total of 32 questions were included and scoring was done according to the
answers received. Respondents were asked to give scores from 1 to 5 for the countries in which they do
business. The six main issues taken into consideration while evaluating the logistics performance were
determined as the efficiency of customs and border gate operations, the quality of the infrastructure related to
trade and transportation, the convenience of arranging shipments at competitive prices, the quality of logistics

248
services, the traceability of shipments, and timely delivery. Performance scores for 155 countries are obtained
by taking the weighted average of the scores given in subjects. The value found reflects the logistics
performance of the country as an index value.

LOGISTIC CONCEPT
The term logistics was used as “logistikos” in ancient Greek and “logisticus” in Latin, meaning
“calculation” or “mathematical logic” in both languages (Kayabaşı, 2007). The word logistics comes from the
French “logistique”. It was translated into English as "logistics" in the 19th century. Turkish Language
Institution is defined as the science of management, organization and planning of activities for the provision
of goods and services (TDK, 2021).
In the 1950s, when the term logistics did not have any equivalent in the industrial sector, it was used as a
military term. In parallel with the developments in management theory and information systems, the scope
and impact of logistics has begun to be seen in the private sector (Frazelle, 2002). In the 1970s, logistics
developed as an activity carried out by the production-related enterprises themselves, and in this period, it
began to separate from the manufacturer companies and become an independent field of activity carried out by
specialized companies (Levent, 2009).
The definition and scope of logistics services has been transformed over the years with technological and
economic developments (UNCTAD, 2006). According to the definition of the Council of Logistics
Management, the term logistics is the effective and efficient planning, implementation, transportation, storage
and control of the movement of any product, service and information flow in the supply chain from the point
of production to the final consumer in order to meet the needs of customers (Tutar et all., 2009).
Logistics has become very important in providing competitive advantage. It is observed that the local
logistics industry has also gained momentum. The logistics sector constitutes approximately 13% of the
national income in our country (U.L. Sektör Raporu, 2018).

GREEN LOGISTIC CONCEPT

Rodrigue et al. (2001) divided the expression green logistics into two parts and explained both words
separately. They explained the term logistics as the exact midpoint of transportation. Logistics is one of the
most important advances in the transportation industry. Green, on the other hand, has become important with a
number of environmental concerns and is generally regarded as positive. The combination of the words Green
and Logistics means environmentally friendly transportation and distribution systems.
Green logistics has been emphasized by many countries and has gained more and more importance with
the increasing interest of the public in environmental protection. (Tao, 2001).
Consumers' interest in environmentally friendly green products and packages has increased in recent
years. It has been observed that some consumers are willing to pay high prices for recyclable products and
demand more information about the content, disposal and recycling of these products. However, these reasons
are not sufficient for the implementation of green logistics (Nylund, 2012). In order for green logistics
activities to be successful, it is necessary to overcome obstacles such as cost, time, flexibility, network
approach, reliability, storage, electronic commerce. (Baltacıoğlu et all., 2008).
The reason why businesses turn to green is the increasing sensitivity of the business world to the
environment. It has been determined that the business has a competitive advantage by choosing green.
Consumers demand that businesses be greener with environmental awareness (Köksal, 2011).

FACTORS AFFECTING GREEN SUPPLY CHAIN MANAGEMENT

The sequence of activities involved in the materials management, production, and distribution of a product
using sophisticated technology to minimize environmental damage throughout the process is referred to as a
green supply chain. Furthermore, a green supply chain contributes to the creation of long-term value for
businesses by striking a balance between economic success and environmental protection (Dashore and
Sohani, 2013). There were five primary elements for analyzing supply chain success in the early days of green
supply chain implementation: design, operations, reverse logistics, waste management, and manufacturing.
(Scupola, 2003). There are many factors that affect each of these stages (Dashore and Sohani, 2013).
There is pressure from both the public and the authorities to use green management practices by
businesses. The green supply chain consists of processes that involve the use of resources and the

249
minimization of waste (Thun and Müler, 2010). The factors affecting the green supply chain are discussed in
the literature in three main groups as internal, external and environmental. Sub-headings of these three titles
are given in Table 1.

Table 1. Factors Affecting Green Supply Chain Management

External Factors Internal Factors External Factors
Suppliers Management Support and Total Amount of Energy Used
Government and Commitment Amount of Material Used
Policy Organizational Structure Energy and Raw Material Consumption Level
Regulations and Strategy Level of Wastes in the Water
Market / Cost Concentration of Harmful Substances Contained in the
Consumer Reverse Logistics Products
Competitors Environmental Design Level of Compliance with Environmental Regulations and
Social Factors (Eco Design) Standards
Supply Logistics Recovery of Investments Gas Emission Quantities
Shipping Organizational Learning Number of Administrative Penalties
Logistics Type of Fuel Used in the Distribution Process of the
Products
Source: Adapted from Yarlıkaş and Can (2020)

Today, the importance of logistics activities is gradually increasing in the context of competition and
sustainability. While countries and unions focus on infrastructure works in order to make trade easier, with the
increase in the carbon rate released as a result of the increase in logistics activities, it has also started to have a
direct effect on climate changes. For this reason, green logistics has become an important issue especially in
developed countries. In addition, it is widely accepted that logistics performance is highly effective in
economic growth and poverty reduction (T.C. Gümrük ve Ticaret Bakanlığı, 2012).

GLOBAL LOGISTICS PERFORMANCE INDEX (LPI)

In the 21st century, the perfectionist efforts of businesses continue rapidly and some changes are observed
both in the structure of enterprises and in their management understanding (Doğan, 2006). In today's
environment where there is intense competition and keeping up with this rapid change, the most accurate way
for companies to understand better their position and developments is to measure their performance (Bayır
and Yılmaz, 2017).
Performance measurement is of great importance as the evaluation of the quality of logistics services, the
outputs of the resources used in terms of efficiency, the perceptions of the customers about the service they
receive or the experiences during the delivery of the product or service, and the determination of success in all
similar process stages (Baki and Şimşek, 2004).
The Logistics Performance Index (LPI) was created by the World Bank to evaluate and rank the logistics
performance of countries. In the measurement of logistics performance, it is ranked by managing the questions
to the employees and managers of the logistics companies in each country and determining the answers given
to these questions according to their scores. The first measurement of the logistics performance of the
countries was made in 2007. In the measurement in 2007, primarily 7 areas were determined. However, in the
following years, measurements were made in only six areas in 2010, 2012, 2014, 2016 and 2018 (Arvis et all.,
2016). These 6 areas are shown in table 2 below.

Table 2. Sub-Criteria of the Logistics Performance Index and the Years Used
LPI Measurement Criteria Years
Customs 2007-2010-2012-2014-2016-2018
Infrastructure 2007-2010-2012-2014-2016-2018
International Shipments 2007-2010-2012-2014-2016-2018
Logistics Competence 2007-2010-2012-2014-2016-2018
Tracking & Tracing 2007-2010-2012-2014-2016-2018
Timeliness 2007-2010-2012-2014-2016-2018

250
 Local Logistics Costs 2007
Source: World Bank, (2018)

LPI, which has become a key tool to explain the relationship between international trade and logistics
infrastructure, has been effective in determining logistics-related problems and priorities, strengthened the
dialogue between the public and private sectors, and supported the development of trade and transportation in
different countries. LPI's top-ranked countries are those with broad distribution platforms and industries
specializing in logistics services. These countries tend to benefit from economies of scale and are at the
forefront of major technological innovations. At the bottom of the LPI are low-income countries, which are
often geographically isolated or experiencing internal political conflicts and economic instability (Raimbekov,
et all., 2017).
While the logistics performance of 150 countries was evaluated in the report of 2007, this number
increased to 155 in the report published in 2010 and 2012, and to 160 countries in 2014, 2016 and 2018. The
criteria used in performance measurement have changed over the years. Only the local logistics costs criterion,
which was among the evaluation criteria in 2007, was not evaluated in the following years.
The place of logistics in the global economy and trade is even better today than it was 10 years ago. With
the logistics performance indicator, it is also possible for companies and countries to see their existing
shortcomings and to provide competitive advantage by planning the investments that need to be made in this
field (Bayat and Özdemir, 2016).

GLOBAL LOGISTICS PERFORMANCE INDEX AND TURKEY

Due to its geographical location, it is one of the countries with a high probability of being a logistics base
by taking an important place in global trade, as it is close to Europe, the Middle East, Asia, Africa and the
Turkic Republics (Babacan, 2003). In addition, it has a high potential and a logistics infrastructure with its
young manpower potential, the suitability of its geographical location and its relatively affordable cost
structure (YASED, 2012).
Although the logistics sector is developing in Turkey, there is a lack of planning issues and this leads to
the inability to get enough efficiency from the investments made. Obstacles and problems in front of Turkey's
logistics sector can be listed as follows: deficiencies in legal infrastructure, bureaucratic obstacles brought by
legislation, lack of trained workforce in logistics, inadequacy of technological infrastructure and ineffective
use of technology, and excess vehicles and companies (Küçük, 2013).
There are 3 main factors effective in logistics being one of Turkey's strategic sectors. The first is that it is
surrounded by sea on three sides and the sea shores are very suitable for making ports, and the second reason
is the geopolitical position of the country and the wide plains of Continental Anatolia. The third reason why it
is strategically important is Turkey's logistics capability, which comes from its past and has become a culture
(Tunç and Kaya, 2016).
The importance of logistics is indisputable in the realization of 500 billion dollars export in line with
Turkey's 2023 vision, sustainability and innovation in exports. The logistics sector, which has a high growth
potential in Turkey, is the second sector that has the highest potential after tourism (Erkan, 2014).
Turkey has been included in all of the evaluations in the Logistics Performance Index reports carried out
by the World Bank between 2007-2018. According to the 2018 performance index, Germany ranks first with
4.20 points, Sweden ranks second with 4.05 points, and Belgium ranks third with 4.04 points. European
countries are always among the leading countries in the rankings. Despite its high potential, Turkey is not
among the top countries (World Bank, 2018).

251
 Table 3. Turkey's Score and Ranking According to the Logistics Performance Index Criteria for the Years
2007-2018

Infrastructure

International

Competence

Tracking &

Timeliness
LPI Score
LPI Rank

Shipment

Logistics
Customs

Tracing
Years

2007 34 3,15 3 2,94 3,07 3,29 3,27 3,38

2010 39 3,22 2,82 3,08 3,15 3,23 3,09 3,94
2012 27 3,51 3,16 3,62 3,38 3,52 3,54 3,87
2014 30 3,50 3,23 3,53 3,18 3,64 3,77 3,68
2016 34 3,42 3,18 3,49 3,41 3,31 3,39 3,75
2018 47 3,15 2,71 3,21 3,06 3,05 3,23 3,63
Cumulative 37 3,29 2,94 3,36 3, 19 3, 23 3,37 3,68
Source: World Bank, (2018)

Table 3 shows Turkey's ranking and scores according to each LPI criteria between the years 2007-2018. In
2007, Turkey ranked 34th among 150 countries with a score of 3.15, while in 2018, it ranked 47th among 160
countries with a score of 3.15. Timeliness, which was our most developed criterion in 2007, was the
benchmark with the best performance in 2018 with a score of 3.63. The criterion that Turkey should make the
most progress on was customs clearance with a score of 2.71. Customs clearance decreased to the lowest level
in 2018 when compared to other years. When analyzed by years, it is seen in the general table that Turkey
recorded a decline in sub-criteria after 2014. Although relatively progress was made in the ranking in 2012
and 2014, a significant regression is observed in 2016 and 2018 (World Bank, 2018).
In Figure 1, the ranking of Turkey according to the LPI criteria is presented graphically over the years.
Considering the customs clearance criteria, it is seen that there is a regression. While Turkey was in the 33rd
place in the customs clearance criteria in 2007, it declined to the 58th place in 2018. In the infrastructure
criterion, however, there has been a slight increase, and it is seen that it has risen from 39th to 33rd place. The
criterion of international shipments is one of the criteria that has regressed. While Turkey was in the 41st rank

Timeliness

Tracking & Tracing

Logistic Competence

International Shipment

Infrastructure

Customs

LPI

0190001900r1l
10190001900r1l
20190001900r1l
30190001900r1l
9190001900r2l
19190001900r2l
29190001900r2l
10190001900r3l

2007 2010 2012 2014 2016 2018

Figure 1. Ranking of Turkey by LPI Criteria by Years

252
in 2007, it regressed to the 53rd rank in 2018. There was also a decline in logistics quality and competence
criteria. Timeliness, one of the best criteria in 2018, showed an increase over the years and it is seen that an
increase was recorded from 52nd to 44th place. As can be seen from the figure, Turkey was able to show an
increase in only 2 of the LPI Criteria, while a decrease was recorded in the other 4 headings (World Bank,
2018).

GLOBAL LOGISTIC PERFORMANCE INDEX AND EUROPEAN UNION

The transport policy of the European Union (EU) represents EU norms in the transport sector as a whole.
While these standards aim to defend the social rights and freedoms of the citizens of the Union, they are also
of critical importance in terms of ensuring economic competitiveness and balanced and sustainable
development. The EU transport policy was created with the aim of increasing economic and social integration
while contributing to the smooth functioning and development of the single market.
The general principles of the common transport policy, which is important for economic development and
social cohesion in the EU, are based on the 1957 Treaty of Rome, which established the European Economic
Community for the first time. Among the articles of this agreement, there are also headings related to
transportation (Çelik, 2007). When the Maastricht Agreement was signed in 1993, the basic principles for the
Trans-European Transport Networks project were discussed and provisions were made and a legal basis was
given to the subject. The aim here is to establish missing interregional connections that may hinder socio-
economic integration within the framework of the development of the single market and to eliminate
bottlenecks in transportation infrastructure (European Commission, 2010). Among the objectives of the White
Papers published in 1985, 1992, 1995 and 2001 are changing the balance between transportation types,
removing the bottlenecks in the law, and placing users at the center of transportation policies (Commission of
the European Communities, 2001).
Another aim of the EU transport policy is to integrate the logistics and transport sectors into sustainable
development. The logistics sector is one of the sectors that most affect environmental and sustainability
policies. When we look at the freight transport within the EU, it is seen that while the share of the road is
44%, the share of the short-distance seaway is 39%, the share of the railway is 10% and the share of the inland
waterway is 3%. Increasing traffic congestion over time, deterioration in the quality of services, damage to the
environment, endangerment of security and isolation of some areas necessitated new activities in
transportation. For example, 25% of the amount of CO2 emissions in the EU originates from the activities of
the logistics sector. The fact that the total area of the EU has increased by 25% and the population by 20% as a
result of the last enlargement adds a new dimension to the issue and makes the problem more critical. For this
reason, it is important to transform into more environmentally friendly, smart and safe systems in
transportation and to bring sustainable quality to transportation models (METU, n.d.).
For this, efforts are made to prevent climate changes as a result of air pollution and greenhouse gas
emissions, to cope with the safety of resources and noise pollution. For this purpose, the Gothenburg
European Council, convened in June 2001, placed the link between economic growth and growth in
transportation at the center of its sustainable development strategy. In this framework, about sixty measures
have been put forward, including pricing, focusing on alternative modes of transport to highways, and
investment in trans-European networks (European Commission, n.d.)
These measures can be listed as; restructuring of highways, strengthening the position of railways in the
transportation sector, preventing delays and congestion in air transportation, increasing the efficiency of sea
and inland waterways transportation, providing sea-inland waterway-railway connections, encouraging
multiple transportation, accelerating the development of trans-European networks, road safety and regulation
of usage fee etc. (İKV, 2010). For these purposes, projects on a regional scale are created and large resources
are allocated. Despite these measures taken, the leading countries in the ranking are among the EU member
states.
The level of development and logistics performance of countries also differ. In other words, while the
logistics performance is high in high-income countries, it is seen that the logistics performance of the
countries decreases with the decrease in the income level (Ofluoğlu et all., 2018). The following table has
been created in order to compare the logistics performance index of Europe's industrialized, rich and powerful
countries with Turkey's logistics performance.

253
 Table 4. Scores and rankings of EU countries in the top 10 and Turkey in the 2018 LPI Ranking

Netherlands
Germany

Denmark
Belgium
Sweden

Finland
Austria

Turkey
UK
LPI Rank 1 2 3 4 6 8 9 10 47
LPI Score 4,20 4,05 4,04 4,03 4,21 3,99 3,99 3,97 3,15
Customs 4,09 4,05 3,66 3,71 3,92 3,92 3,77 3,82 2,71
Infrastructure 4,37 4,24 3,98 4,18 4,21 3,96 4,03 4,00 3,21
International 3,86 3,92 3,99 3,88 3,68 3,53 3,67 3,56 3,06
Shipment
Logistics 4,31 3,98 4,13 4,08 4,09 4,01 4,05 3,89 3,05
Competence
Tracking & 4,24 3,88 4,05 4,09 4,02 4,18 4,11 4,32 3,23
Tracing
Timeliness 4,39 4,28 4,41 4,25 4,25 4,41 4,33 4,28 3,63
Source: World Bank, 2018.

Looking at the 2018 data, the EU countries lagging behind Turkey are Bulgaria (52), Croatia (49), Latvia
(70), Lithuania (54), Malta (69), Romania (48) and Slovakia (53). If we count the UK, whose exit procedures
have not been completed yet, 8 EU countries are in the top 10 in the 2018 LPI ranking. The average LPI score
of the union has increased over the years. It is observed that a decline was recorded with 3,308 points in 2007,
3,321 points in 2010, 3,423 points in 2012, 3, 3,474 points in 2014, 3,534 points in 2016, and finally 3,461
points in 2018 (World Bank, 2018).

CONCLUSION
Logistics should not be considered only as a transportation activity. Logistics, which is one of the basic
elements of global trade, incorporates core activities involving various public and private sector institutions
ranging from transportation to warehousing, from border controls to inland distribution and payment systems
(Köksal, 2012).
Today, improving logistics performance is at the center of economic growth and competitiveness.
Performance measurements greatly contribute to logistics which is based on the policy framework measured
by countries or regional economic groups on infrastructure, legislation, service sophistication, and trade
facilitation practices (Köksal et all., 2014).
The World Bank aims to create a competitive environment between countries with its index study and
tries to determine the situation for countries to improve their logistics performance. The logistics performance
index provides detailed information about the logistics environments of countries, their basic logistics
processes and organizations, and their time and cost performances (Burmaoğlu, 2012)
Due to Turkey's geographical location, criteria such as being a country with experience in a market
economy, being surrounded by seas on three sides and having a suitable port infrastructure, a young dynamic
population, having important human resources, being close to Caspian oil, serving as a bridge country
between EU and Balkan countries should be taken into consideration. It is possible to say that Turkey has the
potential to become a base in the world logistics sector (Küçük, 2013). However, considering the stated
potentials and more, it is seen that Turkey is not in the position it deserves in the ranking.
When the six LPI reports published between 2007 and 2018 are examined, it is seen that the countries in
the first place are countries such as Germany, the Netherlands and Sweden. It is seen that European countries
are predominantly among the top ten countries in the ranking. Eight of the top 10 countries in the 2018 LPI
report are Union countries. Among the countries of the Union, Germany maintained its first place in the world
in 4 consecutive LPI rankings in 2010, 2014, 2016 and 2018. It is seen that the countries with high logistics
performance are effective in global trade and directing the trade (Yapraklı and Ünalan, 2016).

254
 Last but not least, the first green policies and trends affect LPI is sustainable procurement which is the one
of the main practices of a sustainable supply chain, thus businesses should engage in it (Delmonico et al.,
2018). Customers are willing to pay extra for green, therefore sustainable buying techniques are closely linked
to organizing competitively priced shipments in the LPI's shipments dimension. (Hazen et al., 2012).
Secondly, research have indicated that staff education and training are one of the most important elements in
enhancing LPI (Ekici et al., 2019). In addition, it is obvious that a green human resources approach is
necessary for a more sustainable supply chain. Third, researches have indicated that technological literacy
(Ekici et al., 2019) and info technologies are of vital importance in terms of LPI and sustainability in supply
chain. Technological innovations and information systems are especially important for tracing and tracking
shipments, customs clearance efficiency, and ensuring and supporting the on-time delivery. Fourth,
infrastructure improves the LPI ranking by allowing delivery on time (Ekici et al., 2019) while lowering
energy consumption therefore emission release (Khan et al., 2018). Lastly, top LPI performers stated that
shippers regularly seek eco-friendly options, according to the World Bank's LPI study (Arvis et al., 2018). In
this way, as countries that tend to green become more popular, they will tend to invest more in green.

REFERENCES
Arvis, J.-F., Saslavsky, D., Ojala, L., Shepherd, B., Busch, C., Raj, A. (2016). Trade Logistics in the Global Economy:
The Logistics Performance Index and its Indicators.
Arvis, J.F., Ojala, L., Wiederer, C., Shepherd, B., Raj, A., Dairabayeva, K., Kiiski, T. (2018). Connecting to Compete:
Trade Logistics in the Global Economy. World Bank.
Babacan, M. (2003). Lojistik sektörünün ülkemizdeki gelişimi ve rekabet vizyonu. Ege Akademik Bakış, 3(1), 8-15.
Baki, B., Şimşek, B. (2004). Lojistik Faaliyetlere Göre Performans Ölçütlerinin Belirlenmesi. Yöneylem
Araştırması Endüstri Mühendisliği XXIV Ulusal Kongresi, 15-18.
Baltacıoğlu,T., Yurt, Ö., Yumurtacı, I. Ö.(2008). Yeşil Pazarlama ve Yeşil Tedarik Zinciri Yoluyla Sürdürülebilirliğin
Sağlanması. Çukurova Üniversitesi İktisadi ve İdari Bilimler Fakültesi 13. Ulusal Pazarlama Kongresi Bildirileri, 108-
115.
Bayat, T., Özdemir, Ş. (2016). Yenı Bır Lojıstık Performans Endeks Oluşturmak İçın Gereklı Olan Krıterlerın
Belırlenmesı Üzerıne Araştirma. V. Ulusal Lojistik Ve Tedarik Zinciri Kongresi. Mersin.
Bayır, T., Yılmaz, Z. (2017). AB Ülkelerinin Lojistik Performans Endekslerinin AHP ve VIKOR yöntemleri ile
Değerlendirilmesi. 3(2).
Baykal, H., Baykal, T. (2008). Küreselleşen Dünya’da Çevre Sorunları. Mus f Kem l Ünive si esi Sos l Bilimle i
Dergisi, 8(9), 1–17.
Burmaoğlu, S. (2012). Ulusal İnovasyon Göstergeleri ile Ulusal Lojistik Performansı Arasındaki İlişki: AB Ülkeleri
Üzerine Bir Araştırma. Ege Akademik Bakış Dergisi, 12(2), 193-208.
Commission of The European Communities. (2001). White Paper European Transport Policy for 2010: time to decide.
Retrieved from website: http://ec.europa.eu/transport/strategies/doc/2001_white_paper/lb_com_2001_0370_en.pdf.
Çelik, C. (2007), AB Ulaştırma Politikasına Uyum Sürecinde Türkiye’de Kara Ulaşımı Trafik Güvenliği. Dokuz Eylül
Üniversitesi, Sosyal Bilimler Enstitüsü, Avrupa Birliği Yüksek Lisans Programı.
Çevik, S. Kaya, S. (2010). Türkiye’nin Lojistik Potansiyeli ve İzmir’in Lojistik Faaliyetleri Açısından Durum (SWOT)
Analizi, İzmir Ticaret Odası, AR-GE Bülten, 2010 Kasım-Sektörel, 22-28.
Dashore, K., Sohani, N. (2013). Green Supply Chain Management: A Hierarchical Framework for Barriers. International
Journal of Engineering Trends and Technology, 4(5).
Delmonico, D., Jabbour, C.J.C., Pereira, S.C.F., de Sousa Jabbour, A.B.L., Renwick, D.W.S., Thomé, A.M.T. (2018).
Unveiling Barriers to Sustainable Public Procurement in Emerging Economies. 134, 70-79.
Doğan, N. Ö. (2006). Veri Zarflama Analizi İle Belediyelerde Performans Ölçümü: Kapadokya Bölgesi Örneği. Kayseri:
Erciyes Üniversitesi Sosyal Bilimler Enstitüsü.
Ekici, Ş.Ö., Kabak, Ö., Ülengin F. (2019). Imroving Logistics Performance by Reforming the Pillars of Global
Competitiveness Index. Transport Pol, 81, 197-207.
Ener, T. (2010). Küresel Lojistik Performans İndeksi: Mersin'de Faaliyet Gösteren Lojistik Firmalarının Sektörel
Performanslarının İncelenmesi, Yayımlanmamış Yüksek Lisans Tezi, Mersin Üniversitesi Sosyal Bilimler Enstitüsü,
Mersin.Cezayirlioğlu, H. (2012). Ulusal Lojistik ve Tedarik Zinciri Yönetimi Kongresi. Ulusal Lojistik ve Tedarik
Zinciri Kongresi (s. 82). Konya: Necmettin Erbakan Üniversitesi.
Erkan, B. (2014). Türkiye’de Lojistik Sektörü ve Rekabet Gücü. ASSAM Uluslararası Hakemli Dergi (ASSAM-UHAD)
ASSAM International Refereed Journal, 1.
European Commission. (2010). Lightening the Load-Marco Polo Leads the Way, Retrieved from website:
http://ec.europa.eu/transport/marcopolo/files/publi/docs/brochures/bestof_2009_en.pdf.
European Commission. (n.d.). Mobility and Transport. Retrieved September 7, 2021, from website:
https://ec.europa.eu/transport/home.

255
Frazelle, E. (2002). Supply Chain Strategy: The Logistics of Supply Chain Management, The McGraw-Hill Companies,
USA.
Guha, R. (2000). Environmentalism: A global History. Longman.
Hazen, B.T., Wu, Y., Cegielski, C.G., Jones-Farmer, L.A., Hall, D.J. (2012). Consumer Reactions to the Adoptions of
Green Reverse Logistics. Int. Rev. Retail Distribution. 22(4), 417-434.
İKV (2010), Avrupa Birliği Ulaştırma Politikası. Retrieved from website: http://www.ikv.org.tr/pdfs/c10e9c58.pdf.
İstanbul Üniversitesi ve Sakarya Üniversitesi. (2018). Ulaştırma ve Lojistik Sektör Raporu 2018. Ulaştırma ve Lojistik
Kongreleri. Retrieved from https://www.ulk.sakarya.edu.tr/wp-content uploads 2018 05 Ulaştırma-ve-Lojistik-Sektör-
Raporu-2018.pdf
Kayabaşı, A. (2007). İşletmelerin Rekabet Gücünün Geliştirilmesinde Lojistik Faaliyetlerin Performansının Artırılması:
Üretim İşletmeleri Üzerine Bir Uygulama. Dokuz Eylül Üniversitesi, SBE, Doktora Tezi.
Köksal, M. Z. et all. (2014). Connecting to Compete 2014: The Logistics Performance Index and Its Indicators. Gümrük
ve Ticaret Bakanlığı Yayını, 424.
Köksal, M., Z. (2012), Connecting to Compete 2012: The Logistics Performance Index and Its Indicators, Gümrük ve
Ticaret Bakanlığı Yayını Yayın No: 138, Ankara.
Köksal, S., (2011). Müşterilerin Yeşil (Çevreci) İşletmelere Yönelik Kurumsal İtibar Algısı Üzerine Bir Araştırma.
Çukurova Üniversitesi, Sosyal Bilimler Enstitüsü, Yüksek Lisans Tezi.
Küçük, B. (2013). Türkiye’de Bölgesel Kalkınma İle Lojistik Performans Düzeyi Arasındaki İlişkinin İncelenmesi.
Doktora Tezi. Maltepe Üniversitesi Sosyal Bilimler Enstitüsü.
Küçük, B. (2013). Türkiye’de Bölgesel Kalkınma İle Lojistik Performans Düzeyi Arasındaki İlişkinin İncelenmesi.
Doktora Tezi. Maltepe Üniversitesi Sosyal Bilimler Enstitüsü.
Levent, T. (2009). Bir Hizmet Faaliyetinin Mekansallığı: Lojistik Sektörü İle İlgili İș Hizmetlerinin Mersin Kentindeki
Mekansal Gelișimi Üzerine Görgül Bir Değerlendirme, Planlama, TMMOB Şehir Plancıları Odası Yayını, 47, 63-81.
METU. (n.d.). Retrieved from website: http://abofisi.metu.edu.tr/cp-isbirligi.htm.
Nylund, S. (2012). Reverse Logistics and Green Logistics A comparison between Wärtsilä and IKEA (Thesis). Vaasan
Ammattikorkeakoulu University of Applied Sciences.
Ofluoğlu, Ö.N., Kalaycı, C. Artan, S., Bal, Ç. H. (2018). Lojistik Performansındaki Gelişmelerin Uluslararası Ticaret
Üzerindeki Etkileri: AB ve MENA Ülkeleri Örneği. Gümüşhane Üniversitesi Sosyal Bilimler Enstitüsü Elektronik
Dergisi, 9(24).
Raimbekov, Z. S., Syzdykbayeva, B. U., Mussina, K. P., Mussina, K. P., & Zhumataeva, B. A. (2017). The study of the
logistics development effectiveness in the in the Eurasian economic union countries and measures to improve it.
European Research Studies Journal, 20 (4B), 261.
Rodrigue, J., Slack, B., Comtois, C. (2003). The Paradoxes of Green Logistics. The Handbook of Logistics and Supply-
Chain Management, 339–351.
Sbihi A., Eglese R., 2007. Combinatorial Optimization and Green Logistics. Annals of Operations Research, 175(1),159-
175.
Scupola, A. (2003). The adoption of internet commerce by SMEs in the South of Italy: An environmental, Technological
and Organizational Perspective. Journal of Global Information Technology Management, 6(1), 52-71.
http://dx.doi.org/10.1080/1097198X.2003.10856343
Sohail, M. S., Austin, N. K., Rushdi, M. (2004). The use of third-party logistics services: Evidence from a sub-Sahara
African nation. International Journal of Logistics Research and Applications, 7(1), 45–57.
https://doi.org/10.1080/13675560310001619268
Srivastava, S. K. (2007). Green supply-chain management: A state-of-the-art literature review. International Journal of
Management Reviews, 9(1), 53–80. https://doi.org/10.1111/j.1468-2370.2007.00202.x
T.C Gümrük ve Ticaret Bakanlığı (2012). The Logis ics Pe fo m nce In ex n I s In ıc o s. Ankara, Yayın No: 138
Takım, A. (2015). Uluslararası Ticarete Giriş Kavram, Teori, Politika ve Uygulama, Bursa: Ekin Yayınevi.Vachon, S.,
Klassen, R. D. (2007). Supply chain management and environmental technologies: the role of integration. International
Journal of Production Research, 45(2), 401–423.
Tao, J. (2001). Researches on Establishment Model of Green Logistics System. Zhongyuan University of Technology,
China.
Thun, J., Müller, A. (2010). An Emprical Analysis of Green Supply Chain Management in the German Automotive
Industry. Business Strategy and the Environment, 19, 119-132
Tunç H., Kaya M. (2016). Türkiye’de Lojistik Sektörünün Gelişmesinde Dış Ticaretin Rolü Üzerine Bir Nedensellik
Analizi, Süleyman Demirel Üniversitesi Vizyoner Dergisi, 7(14), 58-64.
Tutar, E., Tutar, F., Yetişen, H. (2009). Türkiye’de Lojistik Sektörünün Gelişmişlik Düzeyinin Seçilmiş AB ülkeleri
(Romanya ve Macaristan ) ile Karşılaştırmalı Bir Analizi. KMU İİBF Dergisi, 11(17), 190-216.
Türk Dil Kurumu. (n.d.). Türk Dil Kurumu | Sözlük. Retrieved September 7, 2021, from sozluk.gov.tr website:
https://sozluk.gov.tr
UNCTAD. (2006). Trade and Development Aspects of Logistics Services, http://www
.unctad.org/en/docs/c1ahm1d2_en.pdf

256
World Bank. (2010). Connecting to Compete 2010: Trade Logistics in the Global Economy the Logistics Performance
Index and its Indicators, Washington, DC: The World Bank.
World Bank. (2018). Aggregated LPI | Logistics Performance Index. Retrieved from website:
https://lpi.worldbank.org/international/aggregated-ranking.
Yapraklı, T. Ş., Ünalan, M. (2016). Küresel Lojistik Performans Endeksi ve Türkiye’nin Son 10 Yıllık Lojistik
Performansının Analizi. Ankara Üniversitesi İktisadi ve İdari Bilimler Dergisi, 31(3), 589-606.
Yarlikaş, S., & Can, Z. V. (2019). Yeşil Tedarik Zinciri Yönetimini Etkileyen Faktörlerin Önem Sıralamalarının
SWARA ve Copeland Yöntemleri ile Belirlenmesi. Eskişehir Osmangazi Üniversitesi İktisadi ve İdari Bilimler Dergisi,
14(3), 899–924. https://doi.org/10.17153/oguiibf.519150
Yased, (2012). 2023 Hedefleri Yolunda Bilgi ve İletişim Teknolojileri.

257
 THE IMPACT OF COVID-19 ON E-COMMERCE: A STUDY OF SUPPLY
CHAIN NETWORK AND CARGO TRANSPORTATION

Muhammet Mustafa Akkan1, Tufan Yayla2

Abstract ⎯ The emergence and spread of the COVID-19 pandemic has accelerated digital transformations. E-
commerce has been one of the areas that has received the most attention and demand during the pandemic. As
quarantines become the new normal, businesses and consumers are increasingly digitized, offering and
buying more goods and services online. Despite the persistent differences between countries, the pandemic
has increased dynamism in the e-commerce environment between countries and expanded the scope of e-
commerce, including new firms, consumer segments and products. In many countries, e-commerce
transactions have shifted from luxury goods and services to daily needs that concern a large number of
individuals. This has led to the share of e-commerce in global retail trade rising from 14% in 2019 to about
19% in 2020. Such a trend of e-commerce will undoubtedly lead to significant changes in the supply chain
network, and especially in cargo transportation. Therefore, this study will examine the effects of the covid-19
pandemic on e-commerce, supply chain network and cargo transportation. These effects will be analyzed
using secondary data published by international organizations and future inferences will be made.

Keywords ⎯ COVID-19, e-commerce, supply chain network, transportation

INTRODUCTION
The COVID-19 pandemic has dominated global economic development throughout 2020. Movement
restrictions and other interventions to protect public health have reduced economic activity in many sectors
and countries, affecting production, distribution and consumption. The global economy shrank and there were
serious decreases in global goods trade. The pandemic and its associated economic downturns came at a time
of changing economic conditions and increasing economic challenges. Uncertainties regarding future trade
relations among major economies came to the fore. Significant strains have also erupted in the global supply
chain due to the worldwide disruption of commercial operations. There was a gradual effect spreading to
various sectors (buyers, distributors, suppliers) in the supply chain and the workflow of manufacturers and
logistics departments had to be stopped.
The pandemic has caused a sharp slowdown in economic activity for which economies are largely unprepared.
One of the phenomena affecting the supply chain in this process is e-commerce. The COVID-19 pandemic has
accelerated digital transformations in the world. Digital solutions are increasingly needed to maintain some
economic and social activities remotely during the restrictions imposed by the pandemic. Teleconferencing is
critical to keeping social bonds alive in times of restriction on telework and online education. The accelerating
trend towards e-commerce seen during the pandemic seems likely to continue during the recovery. The
acceleration of the trend towards e-commerce has also affected cargo transportation and operations. With the
increasing momentum, cargo demand has increased, and a period of intense work and transformation has been
entered to meet the increasing demand.

1
Muhammet Mustafa Akkan, KTO Karatay University, Commerce and Industry Vocational School, Department of Foreign Trade,
Konya, Turkey, [email protected]
2
Tufan Yayla, Bilecik Şeyh Edebali University, Bozüyük Vocational School, Foreign Trade Department, Bilecik, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

258
 E-COMMERCE AND COVID-19
E-commerce is the name given to the sale of products and services to consumers over the internet. E-
commerce, which has been a part of our lives since 1997, has become one of the most popular professions in
recent times (Bhatti et al, 2020). E-commerce has been defined by the Organization for Economic Co-
operation and Development (OECD) as "the sale or purchase over computer networks by methods specifically
designed for the purpose of receiving or placing goods or services" and is at the heart of the digital economy.
Payment and delivery can be online or offline, while goods and services are ordered through these networks.
This represents a complex and growing spectrum of economic activity (OECD, 2020). Businesses that sell
products through e-commerce sites generally sell with three different e-commerce models. These are
(UNCTAD, 2020a):
 B2B (Business to Business): It is an e-commerce model in which the business selling products and
consumers are businesses with e-commerce. In other words, with this model, businesses engaged in e-
commerce sell from business to business.
 B2C (Business to Consumer): It is an e-commerce model that sells from business to consumer.
 C2C (Consumer to Consumer): It is an e-commerce model that provides the opportunity to sell from
consumer to consumer.
Regardless of the definitional differences between different analysts, the growing importance and rapid
growth of e-commerce in global economic activity is clear. The share of e-commerce in global retail trade
increased from 10.4 percent in 2017 to 14.1 percent in 2019. The e-commerce sector in Southeast Asia is
known to have grown sixfold between 2015 and 2019, reaching US$38 billion in 2015 (UNCTAD, 2020a).
The COVID-19 pandemic has dramatically changed trading patterns, at least in the shortterm. During the
epidemic, the volume of global trade decreased and the balance of international trade changed. Traditional
supply chains have been disrupted and additional pressure has been placed on shipping, logistics and border
controls. It has spurred the growth of online shopping and services in domestic markets, where movement
controls and consumer concern deter physical transactions.
There has been a significant reduction in overall economic activity during the pandemic. The crisis has led to
growth in the commerce and retail market share taken up by e-commerce, particularly in domestic markets,
where restrictions on movement and consumer concerns about social interaction have spurred the use of e-
commerce. There were significant increases in online conferencing, gaming and entertainment, as well as in
merchandise. In short, more people have started to use electronic commerce more frequently. It is estimated
that the proportion of individuals using the internet worldwide increased from 29.3 percent in 2010 to 53.6
percent in 2019. Global Internet Protocol (IP) traffic, a proxy for data streams, has grown from around 100
gigabytes per second in 2002 to about 88,000 gigabytes per second in early 2020. This figure is now
estimated to exceed 100,000 gigabytes per second (UNCTAD, 2020b) as the pandemic spread throughout the
year with increased traffic from teleworking, video conferencing, digital entertainment and other applications.
Along with traditional businesses with enough resources to market online fast enough and boost their sales,
businesses with an established online presence have become better equipped to take advantage of it.
Communication networks have largely survived during the pandemic (Sangster, 2020). The extent to which
trends experienced during the crisis will continue during the recovery is uncertain and will depend, among
other things, on the pace of the recovery and the nature and extent of the measures taken to facilitate it. Many
consumers state that they will continue to spend more online after the crisis than before. Many employers plan
to blend office and work from home. Therefore, e-commerce platforms are likely to retain most, if not all, of
the market share gains they achieved during the pandemic (Brooks, 2021). The rate of participation in digital
technology and the Internet is a strong indicator of an economy's capacity to leverage e-commerce. The ability
of businesses to participate in international markets increasingly depends on the quality of digital connectivity
available to them, while citizens' online shopping or use of online business services depends on the
availability and affordability of reliable communication networks (UNCTAD, 2020a).
The COVID-19 pandemic has demonstrated the value of information and communication technologies in
keeping economic activity going despite the dramatic restriction in the movement of goods and people. While
office workers worldwide are required, encouraged, or preferred to work from home using the internet,
messaging, and video conferencing platforms to do work previously performed in offices, international
business travel has largely been replaced by video conferencing (United Nations, 2020). Digitization has thus

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

259
alleviated the economic damage caused by the pandemic, the continued reliance on digital technologies for
office workers once the crisis subsides, and the experience that could lead to lasting changes in their work
patterns.

THE EFFECT OF E-COMMERCE ON SUPPLY CHAIN AND CARGO TRANSPORTATION

Supply chain disruptions are defined as “unplanned and unexpected events that disrupt the normal flow of
goods and materials in the supply chain and consequently expose firms to operational and financial risks”. An
interruption in the supply chain can affect the functionality of other supply chain elements (upstream or
downstream) (Butt, 2021). Additionally, managing a highly interconnected supply chain is an increasing
challenge in today's competitive business environment. Higher levels of uncertainty in supply and demand,
shorter technology and product life cycles, market globalization, and increased use of distribution,
manufacturing and logistics partners all result in a complex international network (Hendrickson & Rilett,
2020). Given the complexity of many supply chains, disruption is considered inevitable by many
organizations.
In this context, it is known that the COVID-19 epidemic has interrupted all commercial operations worldwide.
The pandemic identifies as one of the most severe disasters since the last decade as it disrupted and disrupted
supply chains globally. In the manufacturing and services sector, the impact of COVID-19 has led to strict
measures such as the imposition of trade. Approximately 94% of Fortune 1000 firms reported experiencing
disruption from COVID-19 (Paramannand, 2021). The impact of COVID-19 creates a ripple effect in supply
chains spanning different industries. Also, COVID-19 triggers the most variations in both supply and demand.
On the other hand, the impact of COVID-19 is creating a cascading effect that spreads to various sectors
(buyers, distributors, suppliers) in the supply chain (Butt, 2021). In fact, businesses have had to suspend the
flow of raw materials due to a nationwide lockdown that has caused a major problem for manufacturers and
logistics departments. Also, reduced supply and significantly unaffected demand indicate an increase in costs
(Subramanya & Kermanshachi, 2021). In addition, COVID-19 has caused the suspension of production
activity and demand for goods has dropped significantly. Especially the agricultural sector has faced a
decrease of around 80-100% in sales due to the quarantine (Paramannand, 2021).
Amid slowing economic activity, COVID-19 has fueled a surge in e-commerce and accelerated digital
transformation. As quarantines become the new normal, businesses and consumers are increasingly
“digitalized,” offering and purchasing more goods and services online, increasing the share of e-commerce in
global retail trade from 14% in 2019 to nearly 17% in 2020. (Bhatti et al, 2020). E-commerce has great
potential for developing countries to diversify the scope and geographic reach of trade opportunities and
expand the scope of both established businesses and start-ups. It also plays an increasingly important role in
the supply and distribution of both goods and services in domestic markets. E-commerce is increasingly
understood by governments and development partners seeking new avenues of growth. Innovations in
technology and trading methods create opportunities for companies of all sizes to engage in local activities.
This transformation requires adapting supply chains, impacting international trade through lower trade costs
and expanded market access (UNCTAD, 2021).
During the pandemic, the number of initiatives has increased regionally to address deficiencies in supply
chains. For example, securing the supply of COVID-related personal protective and therapeutic equipment,
often traded online through e-commerce platforms and marketplaces and delivered by express or courier
services, was a priority at the start of the crisis (OECD, 2020). However, later on, this area has become
significantly used for many products. Restrictions Transport and additional security requirements at ports,
airports and land borders have disrupted international trade in goods and services, including many goods now
made possible by digital transactions, and small shipments often sent by passenger aircraft. The disruption of
international trade has increased awareness and understanding of the value of international e-commerce in
sustaining economic activity, including facilitating digital trade, and has highlighted the need for accelerated
efforts to promote digitalization throughout supply chains.
Disruption in supply chains means that many businesses lose access to regular suppliers and sometimes have
to look for alternatives to imports rather than local sourcing; this was costly, time consuming and required
inexperienced business managers to acquire new e-commerce skills (Deloitte, 2021). However, a
transformation has begun in the supply chain as the pandemic evolves throughout the year and e-commerce
businesses take the opportunity to make up for the disruptions in traditional trade regulations. For example,
consulting firm McKinsey reported that cross-border e-commerce volumes increased from the second quarter

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

260
of 2020 as economies adjust to pandemic realities and the initial logistical challenges are addressed (Brooks,
2021).
One of the most important aspects of delivering a great online shopping experience is ensuring that the
products shoppers need are available and delivered on time and undamaged. Therefore, many businesses are
looking for new ways to optimize their supply chains and shipping strategies (IATA, 2020a). In this context,
accelerating the production of critical products has come to the fore. Retailers and manufacturers are ramping
up production of popular products as resources become available to ensure consumers have access to
everything they need, including cleaning supplies, paper products, soups and other preserves. Again, Box
manufacturers and carton suppliers are expanding the production of boxes and other shipping materials of all
sizes to ensure that businesses have the means to truly pack and ship individual orders (United Nations, 2021).
As the rapid change in consumer behavior continues to shift from visiting physical stores to shopping online,
businesses are moving to quickly adjust their fulfillment and distribution plans to accommodate more "direct-
to-consumer" businesses. While e-commerce continues to grow at a phenomenal rate, there is certainly a huge
increase in package and LTL (less than truckload) shipments (IATA, 2021). The rise in business-to-consumer
demand and the greater need for fast home delivery requires companies to prepare their supply chains for the
future – today.
Retailers are learning to be more flexible and looking for ways to streamline their supply chains to fulfill what
seems like an endless barrage of online orders. Due to the great need for fast delivery, some businesses and
logistics companies are considering increasing the capacity of their fulfillment networks, including building
new distribution and sorting centers, renovating existing facilities, and implementing more automation
technology. The location of these facilities is also very important (UNCTAD, 2021). Many businesses are
localizing the production and distribution of their products, using smaller warehouses in multiple areas, and
moving to fulfillment centers closer to major consumption areas. To keep up with the demand driven by the e-
commerce boom and maintain inventory levels, retail businesses are turning to intermodal shipping for an
efficient, reliable and cost-effective way to deliver massive amounts of freight.

E- COMMERCE AND CARGO TRANSPORTATION

Cargo transportation (package distribution in the international literature) has become an independent sector by
coming to the fore in the transportation sector. The growth in e-commerce is closely related to the
transportation and logistics activities aimed at delivering the products directly to the consumers. One of the
most important factors affecting customer satisfaction is that the products are delivered to the buyer in the
fastest, least cost and robust manner. At this point, transportation and storage systems become the most
important component of e-commerce (IATA, 2020d).
In our age where online consumption is increasing with the developments in technology, it is possible to say
that the COVID-19 epidemic accelerated this increase. One of the most important actors of online
consumption is cargo services. It is thought that the importance of online consumption and cargo services will
increase exponentially in the COVID-19 period, when contact and social isolation are vital, and perhaps in the
future periods when we will have to learn to live with the virus (Ozden & Erkan, 2021). Ensuring customer
satisfaction in cargo services is very important for cargo service companies and product marketing businesses.
It is seen that cargo transportation has been significantly affected during the pandemic period. Cargo
transported by passenger planes was also affected, as passenger planes were shut down and air transport came
to a standstill. The shrinking carrying capacity and increased security measures have caused delays and pauses
in cargo transportation (IATA, 2021). In this context, the importance of the logistics system for the e-
commerce supply chain, which is growing day by day and is in serious demand due to COVID-19, is clearly
seen. This situation causes important retail companies in the e-commerce sector to establish their own logistics
companies and cargo companies to increase their investments in this direction. For this reason, new cargo
companies and warehousing investments specializing in e-commerce infrastructure are encountered in the
sector (Deloitte, 2021).
The sector that must first respond to the increasing e-commerce volume and adapt itself is the logistics and
cargo companies. At this point, many innovations are observed. Searches continue to increase efficiency and
speed, from distribution systems with drones and autonomous vehicles, to robotic systems, digital solutions
and flexible employee applications. The e-commerce industry also expects its logistics partners to keep up
with this development (Paramannand, 2021). Establishing their own logistics mechanisms by the big online
commerce players seems to increase the competition with traditional cargo and logistics companies. At this

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

261
point, it is imperative for cargo companies to invest in technology, R&D and innovative business models in
order to adapt to the retail trade models of the future.
E-Commerce has revolutionized the way of logistics and cargo transportation. In order to meet the increasing
demand and ensure customer satisfaction, it requires the development of operating models in order to
accelerate delivery and transportation. Air cargo is naturally suitable for this logistical challenge, as no ship,
train or truck is faster than an airplane. In December 2019, e-commerce represented 15% of air cargo volumes
(IATA, 2020b). This number is constantly increasing, and the trend has accelerated during and after the
COVID-19 pandemic. According to McKinsey, in the second quarter of 2020, e-commerce realized the
growth predicted for the next decade (Bhatti et al, 2020).
Global air cargo transportation performance exceeded 238 billion FTK in 2018 and is expected to grow by
over 4% annually through 2038 (Statista, 2021). In 2019, the annual value of global e-commerce merchandise
sales reached $2 trillion and is estimated to exceed $4.4 trillion by 2025. In 2020, more than 2 billion people,
almost a quarter of the world's population, shopped online (IATA, 2020c). It is estimated that there will be
2.14 billion online buyers by 2021. In addition, according to research, 73% of online shoppers during the
pandemic period stated that they will continue to maintain this trend after the pandemic (Brooks 2021).
Until December 2018, eCommerce represented only 14% of total retail sales. This means there is a lot of room
for growth as consumers continue to move online (United Nations, 2021). According to the report published in
May 2021, online sales increased by 15% in the US, 28% in the UK, 52% in China and 27% in South
KoreaThe worldwide growth to date has been 27.6%, representing $4,280 trillion. The e-commerce packaging
market also stood at US$ 27.04 billion in 2020. It is estimated that this figure will reach 61.55 billion USD by
2026. For example, in China, the order volume is expected to reach 127.5 billion packages from the current
figure of 67.1 billion packages in 2025 (IATA, 2021).

CONCLUSION

With the COVID-19 process and the increasing prevalence of e-commerce, it seems likely that consumer
behavior will change in the coming years. With the pandemic, diversity has been added to an already very
dynamic sector with many different players, large and small, and various business models. In this process, e-
commerce expanded to new companies, customers and product types. The increase in demand in e-commerce
has affected both supply chains and cargo transportation. In this process, most industries have been pushed to
focus on speed, digitization, automation and efficiency to ensure that demands can be met.
The necessity of reviewing the supply chains provides important opportunities for e-commerce companies on
a global scale to take their place in this trend. The unexpected epidemic environment of Covid-19 has revealed
the importance of meeting the urgent consumption and input needs of people and companies quickly and
without interruption. For this reason, it seems likely that production and supply centers will shift from the
cheapest production locations to the points where the fastest supply can be obtained. Among all these, a new
era awaits cargo transportation. The sector has entered into a transformation process towards adapting to the
changing conditions with the effect of the pandemic and e-commerce. Undoubtedly, time will tell what will
happen in this process.

REFERENCES

Bhatti, A., Akram, H., Basit, HM, Khan, AU, Naqvi, SMR, Bilal, M. (2020). E-commerce trends during
COVID-19 Pandemic, International Journal of Future Generation Communication and Networking, 13(2),
1449-1452
Brooks, AB (2021). E-commerce Boom: Welcome to the new reality,
https://www.logisticsmgmt.com/article/e_commerce_boom_welcome_to_the_new_reality
Butt, AS (2021). Strategies to mitigate the impact ofCOVID-19 on supply chaindisruptions: a multiple
caseanalysis of buyers and distributors, The International Journal ofLogistics Management, 32(3), 1-25
Deloitte (2021). COVID-19: Managing supply chain risk and disruption,
https://www2.deloitte.com/global/en/pages/risk/cyber-strategic-risk/articles/covid-19-managing-supply-chain-
risk-and-disruption.html
Hendrickson, C., Rilett, LR (2020). The COVID-19 Pandemic and Transportation Engineering, Journal of
Transportation Engineering, Part A: Systems, 146(7), 1-2
IATA (2020a). E-Commerce Monitor: COVID-19 accelerates consumer behavior.
https://www.iata.org/contentassets/d22340c37e0c4cfd8fc05ca6ebf6cc9f/e-commerce-monitor.pdf
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

262
IATA (2020b). The e-Commerce Logistics Landscape.
https://www.iata.org/contentassets/d22340c37e0c4cfd8fc05ca6ebf6cc9f/ecomm-logistics-models.pdf
IATA (2020c). An opportunity for transformation: Air cargo e-commerce and COVID-19 impact,
https://www.iata.org/contentassets/d22340c37e0c4cfd8fc05ca6ebf6cc9f/ecom-article-covid.pdf
IATA (2020d). The e-Commerce Impact on Air Cargo Oparation: Challenges and Opportunities for Logistics,
https://www.iata.org/contentassets/d22340c37e0c4cfd8fc05ca6ebf6cc9f/ecommerce-impact-challenges.pdf
IATA (2021). E-Commerce Monitor: One year after pandemic.
https://www.iata.org/contentassets/d22340c37e0c4cfd8fc05ca6ebf6cc9f/e-commerce-monitor-q2.pdf
OECD (2020). E-commerce in the time of COVID-19, https://www.oecd.org/coronavirus/policy-responses/e-
commerce-in-the-time-of-covid-19-3a2b78e8/
Ozden, AT, Erkan, C. (2021). Analyzing the service quality priorities in cargo transportation before and
during the Covid-19 outbreak, Transport Policy, 108, 34-46
Paramannand, RA (2021). Impact of Covid 19 on E-Commerce, International Journal of Research and
Analysis in Science and Engineering, 1(3), 48-53
Sangster, S. (2020). E-commerce growth and Covid-19 push to need to digitise,
https://www.aircargonews.net/technology/e-commerce-growth-and-covid-19-push-to-need-to-digitise/
Statista (2021). Worldwide air freight traffic from 2004 to 2021, statista.com/statistics/564668/worldwide-air-
cargo-traffic/
Subramanya, K., Kermanshachi, S. (2021). Impact of COVID-19 on Transportation Industry: Comparative
Analysis of Road, Air, and Rail Transportation Modes, ASCE International Conference on Transportation &
Development, January
UNCTAD (2020a).COVID-19 and e-commerce: a global review, https://unctad.org/system/files/official-
document/dtlstict2020d13_en_0.pdf
UNCTAD (2020b). ESTIMATES OF GLOBAL E-COMMERCE 2019AND PRELIMINARY
ASSESSMENT OF COVID-19 IMPACTON ONLINE RETAIL 2020, https://unctad.org/system/files/official-
document/tn_unctad_ict4d18_en.pdf
UNCTAD (2021). How COVID-19 triggered the digital and e-commerce turning point,
https://unctad.org/news/how-covid-19-triggered-digital-and-e-commerce-turning-point
United Nations (2020). COVID-19 and E-commerce: findings from a survey of online consumers in 9
countries, https://digitallibrary.un.org/record/3886558
United Nations (2021), Global e-commerce jumps to $26.7 trillion, fuelled by COVID-19,
https://news.un.org/en/story/2021/05/1091182

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

263
 EUROPE’S NEW TRADE DOOR OPENNING TO ASIA: NAKHCHIVAN
CORRIDOR

Tufan YAYLA1, Sinan Çizmecioğlu2

Abstract ⎯ Until the geographical discoveries, Europeans obtained the goods they needed from the Asian
continent using the routes called Silk Road and Spice Road. As these routes involve commercial risk sand
dangers, Europeans have started to provide the products they need with the least risk and danger by crossing
the ocean with ships in order to create new alternatives. As a result of this period called geographical
discoveries in history, European sailors formed the basis of today's economic life with the products and
commercial income they obtained from new geographies. Moreover, with the effect of developing
transportation Technologies and the newly established world order, these trade routes have lost their
importance over time, as a requirement of globalization, the world has become the only market and new
modes of transport have been developed. However, despite all these developments, countries that want to
trade globally as a result of trade wars and economic competition between countries have started to search
for a new route .As a result of this search, historical trade routes came back to the agenda. For this reason,
the Silk Road, which was used in the past, was redesigned under the leadership of China, taking the title of
Iron Silk Road. This development has again increased the importance of global road and rail transport. For
this reason, in accordance with the agreement between Azerbaijan and Armenia, the connection of Azerbaijan
and Nakhchivan by road through a corridor located on the territory of Armenia constitutes a new global trade
route alternative. This alternative is the activation of Europe, Central Asia and Far East road transport on the
global trade route with the Nakhchivan Corridor in the 21st century. In this study, the current trade roads and
routes between Asia and Europe and the data on trade volumes are examined in terms of rimland theory and
economic integration theories. Therefore, the main purpose of this study is to theoretically discuss the
commercial geopolitical effect that Nakhchivan corridor will create on trade between Europe and Asia with
current technological possibilities.

Keywords ⎯ International Logistics, International Trade, Nakhchivan Corridor, Global Integration

INTRODUCTION
According to the World Trade Organization data, the world trade volume has exceeded $19 Trillion (WTO,
2021). During the period from the Industrial Revolution to the unipolar period, the direction of world trade,
which was in the form of raw materials from east to west and product shipments from west to east, reversed.
The superior trade advantages of the West pass into the hands of the East, thanks to countries such as China.
While western companies bought eastern companies 50 years ago, today eastern companies have started to
buy western companies. This race between the west and the east of the world in trade both increased the trade
volume and brought new shipping routes to the agenda. The oldest of these shipping ways is the silk road. In
2013, China announced that it would revive this historical trade route. In this context, an expenditure of $ 40
billion is planned. In addition, it is anticipated that $750 million will be spent each year for investments to be
made after completion. The iron silk road aims to revive an old route connecting Central Asia, the Middle East
and Europe. With this project, China aims to reach rich Europe. According to TradeMap data, China exported
$463.97 billion to Europe and imported $278.33 billion in the same period, according to 2020 data. In
addition, with this project, 21 Trillion $ added value is expected to be created (Çalışkan,2018).

1
Tufan Yayla, Bilecik Şeyh Edebali University, Bozüyük Vocational School, Foreign Trade Department, Bilecik, Turkey, [email protected]
2
Sinan Çizmecioğlu, KTO Karatay University, Commerce and Industry Vocational School, Department of Civil Aviation and Cabin
Services, Konya, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

264
 CHINA'S TRADE WITH THE WORLD
With the start of the unipolar era, China was an endless market for the west. However, with the acceleration
and technological breakthroughs it has gained after the 1990s, China has become a seller rather than a market.
In an environment where the World Trade volume decreased by 17.5 trillion dollars with the effect of COVID
for 2020, China managed to increase its exports and increased it to 2.59 trillion dollars.
As a result of the Chinese dream that started 30 years ago, China has become a country with a foreign trade
surplus of $535 billion. However, only $185 billion of this comes from Europe. China regularly runs a foreign
trade deficit against only Germany among the European countries. In general, it is the party that gives foreign
trade surplus in its trade with Europe. With its new Silk Road project, China aims to both have more foreign
trade surplus in its trade with Europe and to reverse its trade relationship with Germany.
In addition, the World Trade Volume, which was 3.5 Trillion in 1991, has reached the level of 17.58 Trillion $
as of 2020. In the same period, China's trade increased from $71.91 Billion to $2.59 Trillion (WTO, 2021).
Therefore, in the 30-year period, China has been the locomotive of world trade and has proven to be a global
player. Despite the trade wars, China's most important trade allies are the United States and Canada.

THE NEW SILK ROAD PROJECT AND THE RELATIONSHIP WITH EUROPE

In our close geography, the idea of being united with the awareness of being independent in domestic affairs
but acting jointly in foreign affairs first came into being in the European Union, which was formed by the
European continent states where the First and Second World Wars took place. These countries, which have
fought each other throughout history over the distribution of wealth and trade, first established the European
Coal and Steel Community at the end of the second world war. As a result of the success achieved with the
European Coal and Steel Community, the Community turned into the European Economic Community with
the Treaty of Rome signed in 1957. With this transformation, it was agreed on the liberalization of the trade of
commercial products other than coal and steel among the member states. The success achieved as 6 member
states pushed other European countries to become members of this union, and as a result, a wave of
enlargement started. However, the main success in this process was the ratification of the customs union
agreement, which was established after the completion of the common trade policies, in 1968. Especially with
the unification of Germany and the disappearance of the Soviet influence from Europe, the European
continent entered a new era and on the path of political integration, the establishment of the European Union,
which first adopted monetary union and then security and foreign policy union, was approved with the
Maastricht Treaty in 1991. The European Union process, which started with 6 countries in 1951, has reached
the last stage of the 5-step integration stages, which has increased to 27 member countries today (Ministry of
Foreign Affairs, 2020).
As of today, the European Union is the second largest economy in the world after the United States with a
GDP of 17.1 trillion dollars. Again in the same period, the share of the European Union in world exports was
15.2% and its share in world imports was 14.7%. This success of Europe and its large market, China, have led
to the development of alternative routes for the European market. With the New Silk Road or Iron Silk Road
project, it is planning to transport non-perishable industrial products to Europe with rail systems, which is the
most suitable transportation route, and to buy industrial machinery and raw materials from Europe in return.
Thus, the transactions that started 30 years ago are reversed. While Europe is selling raw materials and
machinery, China has come to the position of selling finished products.

THE EFFECT OF THE NAHCHIVAN CORRIDOR

Despite the great role they played in history, the Turkish states in the same time period were completely
integrated into the strongest. It never thought of establishing a union by preserving its independence as it is
today. However, in today's world, developments and changes outside the Turkish geography have made it
necessary for the Turkish States to act together. Being aware of this, the representatives of the Turkish states
first signed the agreements that formed the basis of the Turkic Council with the summit held in Almaty in
2011 with the theme of "Economic and Commercial Cooperation" (Turkic Council, 2021). As a matter of fact,
as a continuation of the problems experienced, as a result of the provocation of the Armenian invaders in
Azerbaijan's occupied Karabakh lands and the attacks launched by the Armenian State on Azerbaijani lands,
Azerbaijan took action both to respond to these provocations and to take back the occupied Karabakh, a

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

265
member of the Turkic Council, especially Turkey. countries supported both politically and militarily. As a
result of this struggle and strong stance, Armenia, which could not get the support it expected from Russia,
both lost the war in a short time like 40 days and gradually withdrew from the Azerbaijani lands it had
occupied since 1992. In addition, Armenia has officially agreed to open a transit corridor on its own territory
between Azerbaijan and Nakhchivan, which is also Azerbaijani territory (SDE, 2020). This corridor, also
called the Nakhchivan corridor, has created a new alternative route that geographically connects Turkey,
which is a member of the Turkic Council, and Azerbaijan, as well as other Turkic Council members and Asia.
With this alternative route, known as the Nakhchivan corridor, a new alternative will be gained to the Baku-
Tbilisi-Kars railway, which was built to eliminate Armenia in the New Silk Road project. Thus, new
alternative routes will be developed against the disruptions that may occur in the project with a total cost of
more than $ 40 billion. Especially the fact that the Nakhchivan corridor is 340 km shorter than the Baku-
Tbilisi-Kars line will save time.
It is clear that the Nakhchivan corridor, as mentioned above, will first of all contribute to the development of
trade between Turkey and the Turkic Republics. Moreover, together with Turkey's geopolitical position, the
340 km shortening of the connection road will improve logistics opportunities. In Turkey, there will be an
opportunity to do twice as much trade depending on two different routes on the Iron Silk Road. It will also
serve Turkey's desire to improve its relations with Asia by changing its foreign policy (Esenbel and Atlı,
2013).

CONCLUSION

World trade creates a new situation under the leadership of China, which wants to strengthen its regional
power and increase the volume of trade with the implementation of projects such as the new silk road project.
Compared to 30 years ago, the country that steered the world trade is now China. The revitalization of the
New Silk Road essentially serves this purpose. However, the Nakhchivan Corridor, which was accepted as a
result of the Azerbaijan-Armenia war that started and ended unexpectedly, was an unexpected gain. The
distance, which is reduced by 340 km according to the route in the new silk road project, offers new
commercial opportunities for both China and Turkey. Although this road is the new alternative route of
Europe opening to Asia, it also constitutes the new alternative route of Asia opening to Europe.
In order to realize all the above, logistics investments especially between Azerbaijan, Nakhchivan and Turkey
should be planned and integrated and the integration of the newly built transportation network into the New
Silk Road should be completed. Thus, both the commercial activities of countries like Turkey will increase
and the trade volume between Asia and Europe will increase more than expected.

REFERENCES

Çalışkan, Z. (2018). Çin’in Yeni İpek Yolu Projesi. Erişim Adresi: https://insamer.com/tr/cinin-yeni-ipek-
yolu-projesi_1111.html. Erişim Tarihi: 25 Temmuz 2021.
Dışişleri Bakanlığı (2020). Avrupa Bi Temmuz rliği Tarihçesi. Erişim Adresi: https://www.ab.gov.tr/avrupa-
birliginin-tarihcesi_105.html. Erişim Tarihi: 15 Temmuz 2021.
Esenbel, S. ve Atlı, A.(2013). Turkey’s Changing Foreign Policy Stance: Getting Closer to Asia?. Erişim
Adresi: https://www.mei.edu/publications/turkeys-changing-foreign-policy-stance-getting-closer-asia. Erişim
Tarihi:20 Temmuz 2021.
SDE. (2020). Azerbaycn- Ermenistan Antlaşmasının Tam Metni. Erişim Adresi:
https://www.sde.org.tr/asya/azerbaycan-ermenistan-anlasmasinin-tam-metni-haberi-20029. Erişim tarihi: 01
Temmuz 2021
Trademap. (2021) Erişim Adresi:
https://www.trademap.org/Country_SelProductCountry_TS.aspx?nvpm=1%7c156%7c%7c%7c%7cTOTAL%
7c%7c%7c2%7c1%7c1%7c3%7c2%7c1%7c2%7c1%7c1%7c1. Erişim Tarihi: 30 Temmuz 2021
Trademap. (2021). Dünya Ticaret Verileri Erişim Adresi:
https://www.trademap.org/Country_SelProductCountry_TS.aspx?nvpm=1%7c156%7c%7c%7c14719%7cTO
TAL%7c%7c%7c2%7c1%7c1%7c3%7c2%7c1%7c2%7c1%7c1%7c1. Erişim Tarihi: 30 Temmuz 2021
Trademap. (2021). Avrupa İle Çin Arasındaki Ticaret Erişim Adresi:
https://www.trademap.org/Country_SelProductCountry_TS.aspx?nvpm=1%7c156%7c%7c%7c14719%7cTO
TAL%7c%7c%7c2%7c1%7c1%7c3%7c2%7c1%7c2%7c1%7c1%7c1. Erişim Tarihi: 30 Temmuz 2021

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

266
Trademap. (2021) Çin’in Avrupa’ya İhracatı. Erişim Adresi:
https://www.trademap.org/Country_SelProductCountry_TS.aspx?nvpm=1%7c156%7c%7c%7c14719%7cTO
TAL%7c%7c%7c2%7c1%7c1%7c2%7c2%7c1%7c2%7c1%7c1%7c1. Erişim Tarihi: 30 Temmuz 2021
Trademap. (2021) Çin’in Avrupa’dan İthalatı. Erişim Adresi:
https://www.trademap.org/Country_SelProductCountry_TS.aspx?nvpm=1%7c156%7c%7c%7c14719%7cTO
TAL%7c%7c%7c2%7c1%7c1%7c1%7c2%7c1%7c2%7c1%7c1%7c1. Erişim Tarihi: 30 Temmuz 2021
Türk Keneşi. (2021). Türk Keneşi Tarihçesi. Erişim adresi: https://www.turkkon.org/tr/organizasyon-tarihcesi.
Erişim tarihi: 01 Temmuz 2021
WTO. (2021). Erişim adresi: data.wto.org. Erişim tarihi: 30 Temmuz 2021

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

267
GREEN LOGISTICS PRACTICES IN AVIATION INDUSTRY: A RESEARCH
ON ANTALYA AIRPORT
Ayşegül Toy1, Abdullah Oktay Dündar2

Abstract ⎯ The rapid growth of international trade in recent years has increased the importance of logistics.
On the other side, increase in the importance of speed factor in competition has caused a rise of the use of
airline transport in logistics operations. Airports carry out a significant function in airline transportation. While
airports create many economical values, they also negatively contribute the environmental pollution due to
reasons such as the use of fossil fuel by airplanes and waste produced in the terminals. In order to minimize the
environmental pollution and harmful emissions caused by airports, sustainable and green logistics approaches
are recommended. The purpose of this study is to present the practices of green logistics in the aviation industry,
to review the “Green Airport” project by Directorate General of Civil Aviation (DGCA) and “Carbon-free
Airport” project by General Directorate of State Airports Authority (GDSAA) in Turkey, and to examine green
logistics practices being carried out in Antalya Airport. In Antalya Airport, temporary storage of batteries,
vegetable oil and engine oils, electronic devices and dangerous wastes is being implemented in accordance with
related regulations. Antalya Airport performs greenhouse gas measurements complying with standarts of TS
EN ISO 14064-3:2007. Greenhouse gas emission amount in Antalya Airport is 28.691,84 ton CO2 equivalent
in 2017, and DGCA has certificated Antalya Airport as green airport.

Keywords ⎯ Airline Transportation, Green Logistics, Green Airport

1. INTRODUCTION
Increasing commercial activities world wide lead to many environmental problems. Therefore, today there
is an increasing sensitivity regarding the environmental issues such as using natural resources consciously. As
a result of this, "green lojistics" practices, which are based on integration of environmental factors to lojistics
activities, come to the scene and organizations are forced to put serious efforts on this area.
Growing international trade have increased the demand for lojistics activities. Among all transportation
modes, usege of airway transportation, which becomes prominent with the advantage of speed, shows an
increase world wide in recent years. While this increase in the volume of air transportation contributes to the
economy, fuels used by airplanes on cruise, departing and landing cause CO2 emmisons. In addiiton to this,
activities being carried on in airports casue environmental problems such as noise, air and water pollution, and
solid wastes. According to IPPC (Intergovernmental Panel on Climate Change), air transportatin constitutes
3.5% of total effect on global warming caused by emissions by human activities (Kumaş, 2019).
Various policies at both national and international levels are being developed in order to challenge
environmental problems caused by airports. Annex 16 – Environmental Protection, published by ICAO contains
advisories on aircraft noise, emissions and land usage; European Union Aviation Safety Agency (EASA) sets
standards for noise, air quality and climate change; General Directorate of State Airports Authority (DHMI)
implements "Carbon-free Airports Project" to reduce the negative environmental effects of airports; Directorate
General of Civil Aviation (DGCA) aims to encourage organizations operating in airports on environmental
conscience with its project "Green Airports". Organizations which comply with certain criteria are benefited
from various discounts with their title of "Green Organization" (Canöz, 2020).
This study aims to present green lojistics practices in the air transport industry, to review "Green Airport

1Necmettin Erbakan University, Institute of Social Sciences, International Marketing and Logistics Management, Konya, Turkey,
[email protected]
2Necmettin Erbakan University, Faculty of Applied Sciences, Department of Logistics Management, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

268
Project" by DGCA, and "Carbon-free Airport Project" by DHMI, and to examine green lojistics implementations
being carried out in Antalya Airport. Accordingly, the first part of the study is introduction. In the second part,
the concept of green lojistiscs is described, and its development and implentations are mentioned. The third part
is consists of the presentation of green logistics examples in air transport industry from around the world. In the
fourth part, "Green Airport" and "Carbon-free Airport" projects carried out by state authorities are reviewed and
Antaya Airport is examined in the context of "green airport". In the concluison section, effects of green logistics
practices in the air transport industry is evaluated.

2. GREEN LOGISTICS
"Kyoto Protocol", which has been signed in 1997 within the context of challenging global warming and
climate change has brought up environmental issues such as global warming, carbon emissions, greenhouse
gases and their recognition has increased in the following years. The concept of green lojistics and its
implementations first took place in the literature in 1950s. Today, green logistics has become an important
concept for organizations not only for complying the standards set and forced by state authorities, or reducing
costs but also in order to establish a respectable and impressive corporate image (Üstünbaş, 2018:11).
Özdemir (2019) states that green logistics, which also can be named as ecological logistics is the process of
understanding the effects of logistics to the environment and of reducing this effect to the minimum. In
conventional logistics activities, while economical profit is being made environment is hardly a concern. On the
contrary, green logistics aims sustainable economic growth and social development (Özdemir. 2019).

2.1. The Concept of Green Logistics

Green logistics concept that was defined in mid 1980s means that transportation and supply systems which
are efficient and compatible with the environment. In another definition, green logistics is conducting the
production and distribution of goods and services in a sustainable manner that takes into consideration of
environmental and social factors (Üstünbaş, 2018).
Green logistics practices provide benefits such as an efficient management of distribution channels,
improving the environmental performance, minimising the waste, and cost reduction (Küçük and Orhan, 2012
as cited in Üstünbaş, 2018:12).

2.2. Development of Green Logistics

Logistics industry has a significant portion of greenhouse gas emissions to the atmosphere. Percentage of
total greenhouse gas emissions from logistics industry world wide is about 24%. The concept of "green" that
aims to minimise the environmental problems caused by transportation industry has come to the scene in the
end of 1980s, and a public awareness of the issue has started to arise. Report publisehed in 1987 by the World
Commission on Environment and Development has procured significant political and financial support to the
issue. And in the begining of 1990s, important studies has performed regarding to integration the environment
to the logistics industry (Banister and Buton, 1993; Zengin and Ocalır Akunal,2017).
Wtih the integration of green into logistics, the concept of "green logistics" has emerged. Saving energy and
minimising and recycling waste have gained importance. Practices that carry out production and distribution
activities in a sustainable manner by taking into account social and environmental factors has started to attain a
place in the sector (Zengin and Ocalır Akunal,2017).
In 1990s, a perspective that integrates the environment and logistics has started to be adopted. One of the
important steps intended to challenge climate change and to protect the environment was Kyoto Protocol
(Üstünbaş, 2018). Signed in December 11, 1997, Kyoto Protocol involves reducing greenhouse gas emissions
to the atmosphere. In this context, countries are required to reduce amount of the emssions down to the levels
in the year 1990. The protocol which was signed in 1997, could take effect with the participation of Russia in

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

269
February 16, 2005 (Yesilaski, 2021).
Factors effecting the development of green logistics (Üstünbaş, 2018):
– Raising the environmental consciousness in the society
– Public's force to the local authorities
– Environmental organizations' activities
– Disclosing harms to the nature in a clear manner

Green logistics practices increase environmental consciousness, preserve nature, decrease factors that are
harmful to the health. From the businesss' perspective, when they adopt green logistics, on one hand they are
able to minimise harms to the environment, on the other hand they attract the consumers who have
environmental sensitivity (Üstünbaş, 2018).
Özdemir (2019) states that raising awareness about environment and becoming widespread of low carbon
economy are very effective on development of green logistics. Accordingly, it is indicated that with green
logistics, preventing the environmental pollution and reducing resources usage are being aimed (Özdemir,
2019).

3. GREEN LOGISTICS IN AVIATION INDUSTRY

In 1987 Brundtland Report, sustainability defined as the development that meets the needs of today without
comprimising the future generations' capability of meeting their own needs (Wikipedia, 2021a). According to
this definition, economical development should be performed by considering the environmental factors.
Since aviation industry grow rapidly in a global scale it causes environmental problems such as climate
change, air pollution, emissions, damage to natural life, noise, land use, energy and water consumption and
waste (Güngör, 2020).
Public authorities, organizations and companies make provisions on various subjects from airport design to
construction and management of them, in order to minimise the effects of these environmental issues (Oto,
2010). Preferring the use of some specific fuel types, executing special taxes, offering subventions for
alternatives like high-speed trains can be given as examples of precautions taken in air transportation industry
against climate change. In addition to this, carbon tax and emission policies are among the precautions (Forsyth,
2011 as cited in Güngör, 2020).
Environment reports of European Airways can be example for many companies in developing countries.
While in developed countries big investments can be made about environmental reporting, in less-developed
countries it is recommended that cooperating with public authorities and professional organizations in order to
reduce these investment costs (Güngör, 2020)

3.1. Practices in the World

Aviation industry is facing with an increasing pressure about climate change. In addition to airplanes
consuming great amounts of fossil fuels, airports might have negative effects on the environment as well. As
the number of airports are increases and as they are expanding in size, the more energy consumption is needed
for construction of new terminals and necessary infrastructre. This results in increasing the carbon foot print
(Ekoyapidergisi, 2021).
Alongside with the increase of the volume of air transportation, airport capacities are increasing as well for
the need of more aircraft in the sector. Rising construction capacities might cause some raised eyebrows because
of harms to the environment. Therefore many airports in the world adopt an environment friendly approach in
both design and operation and management levels, and carry out practices that would reduce carbon foot print.
Moreover, international organizations also provide airports support for reducing the carbon foot print. Within
this scope, Airports Council International (ACI), with airprot carbon accredition programme, provide support

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

270
to over 200 airports for emission management in order to accomplish the objective of carbon-neutr in airports.
Following are some of the airports that has been designed with an environment friendly approach
(Ekoyapidergisi, 2021):
Boston Logan International Airport, USA
Boston Logan International Airport have the first airport terminal that has got the Accreditation of Energy
and Environment Design which is one of the highest quaility standard in the area of green building certification.
There are roof panels and surface platings over the A Terminal of the airport, which was designed for refrect the
heat from the building. Also, low-flow faucet armtures was used for water saving.
3% of the energy being used in operations is produced by wind turbines located on the roof of office
buildings.
Galapagos Ecological Airport, Galapagos Islands
Galapagos Ecological Airport, established in 2012, is the first total green airport which operates solely with
the wind and sun energy in the world. 65% of its energy consumption is met by wind turbines, and 35% is met
by the solar panels set on pedestrian walkways. Moreover, 80% of the materials used in the infrastucture of the
airport is recycled materials. For the similar purposes, it has been used reclaimed woods and metals in most
parts, and furnitures were produced in an environment friendly manner.
In addition, there is a salt removing system that purifies the sea water for the water need of the terminals.
Waste water is resending to the same system for reusage.
Oslo Airport, Norway
Oslo Airport management has focused on the energy efficiency in the new design while increasing its
capacity. For this purpose, they use recycled and natural materials in the expansion phase. For example, the area
which was constructed with an environment friendly concrete made up from volcanic ash, was covered with
recycled steel and timber. In addition, the snow ploughed from the runways in the winter time is used as coolant
in the summer time. Oslo Airport has won the degree of "exellence in sustainability" with this practice.
Stockholm Arlanda Airport, Sweden
Arlanda Airport is the first airport which has achieved the carbon neutrality in 2009. It is also the only airport
which has set a maximum level for carbondioxide emission. It is required that the emission caused by operations
in terminals and airplanes must not exceed the level in the year 1990.
In the airport, terminals, hangars and buildings in the air side are heated with bio-fuel for saving electricity.
The water has been used in the ventilation system in the summer time, is heated with a system operating with
bio-fuel. Additionally, the same system is being used in the preventing icing around the gates and ramps. With
this practices and precautions in the airport, energy consumption has been reduced by a rate of one third between
2005 and 2012.

3.2. Literature Search

In this section, studies on green logistics and green logistics in the aviation sector are presented as a
summary of the literature.
In her study, Üstünbaş (2018), aimed to evaluate the green logistics practices of logistics companies
operating in the Marmara Region in terms of their contribution to environmental sustainability. In order to
analyze the applications of the companies, the survey method was applied and as a result of the study, it was
determined that the companies were aware that green logistics contibuted positively to the brand image. In
addition, it has been stated that factors such as high investment and technology costs and lack of information
are barriers to green logistics applications.
Zengin and Akunal (2017), used the data obtained from the literature review, face-to-face interviews with
companies, and a questionnaire presentation with questions about activities in their study aiming to reveal green
logistics practices in Turkey. As a result of the study, there is no activity in Turkey on creating limited areas for
logistics applications, determining low emission zones, coordinated transportation systems and renewable
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

271
energy indicators, legal regulations, measures to reduce energy consumption, and measures to reduce emissions
and carbon footprints are the most important ones. It has been observed that there are very common applications.
In his study, Mete (2020), aimed to investigate the effect of the development in logistics activities on
greenhouse gas and carbon emissions. In this context, the samples of 26 European Union member countries and
Turkey were analyzed with the panel regression model. As a result of the study, it has been seen that the
development in logistics activities has a reducing effect on greenhouse gas emissions.
In their study, Canöz and Ertek (2020), subjected the web pages of the certified organizations to content
analysis in order to evaluate the contribution of the green organization certificate to the business image and to
reveal the use of the certificate. As a result of the study, it has been seen that the majority of businesses with
green organization certificate do not include the green airport logo and green organization certificate on their
web pages.
In the study conducted by Durdağ (2020), it was aimed to determine whether the green airport certificate
has an effect on the international trade of the airport. In this context, in-depth interviews were held with the
authorities of two different airports with green airport certificates. As a result of the study, it was observed that
airport operators had difficulty in establishing a relationship between the direct impact of having a green airport
certificate on the airport's trade, and it was stated that service price, accessibility, speed were the priority issues
that provided advantage.
In the study conducted by Oto (2010) it was aimed to reveal the environmental effects of the aviation sector
and airports by considering the environmental effects of airports, environmentally friendly airport planning,
implementation and operation principles and the approaches that affect this process, and Esenboğa Airport was
examined. In this context, environmental friendly approaches and sustainability practices at the airport were
examined and compared with various airports in the world and airports in our country. As a result of the study,
it is predicted that the civil aviation sector will grow steadily with environmentally friendly airport approaches
and sustainability practices. It has been stated that sustainability practices will increase efficiency in air transport
and airport activities.

4. GREEN AIRPORT
With the change in the Civil Aviation Law in 1983, private sector was given the right to operate airway
companies, airports and ground services in Turkey. Starting this date, Turkish air transportation sector begun to
grow rapidly. With this growth, many aviation companies largened their fleets by purchasing new airplanes. Air
traffic and amount of fuel consumed has increased parallelly. Since increasing the amount of fuel consumed
means that increasing the polluters which would be emitted to the atmosphere, it is important for companies to
consider the environmental problems of their new airplane purchase decisions (Şekertekin, 2017).
Emissions stem from aviation activities caused a high percentage of air pollution (Durmaz et al. 2007; as
cited in Şekertekin, 2017). Environmental effects of mentioned emissions are long-distance air pollution, effects
to ozon layer, and green house effects. Polluters like SO2 and NOx emitted by airplanes cause acid rains that
are harmful for plants, animals and various structures (Şekertekin, 2017).
There are efforts in Turkey as well as in the world to reduce the negative environmental effects arising from
aviation industry. Some examples of these efforts are projects conducting by Ministry of Environment and
Urbanisation which are dealing with air quality, wastes and noise pollution; "Carbon-free Airport Project" by
General Directorate of State Airports Authority; "Green Airport Project" by Directorate General of Civil
Aviation aiming to encourage organizations operating in airports to conduct their businesses in an
environmentally friendly manner (Canöz & Ertek, 2020).
This study examines DGCA's "Green Airport Project" and GDSAA's "Carbon-free Airport Project" and
Antalya Airport in this context.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

272
 4.1. DGCA and "Green Airport" Project
Along with the growth in the aviation industry in Turkey, new airports has built and many new airway
companies has entered to the sector. And consequently increasing air traffic has caused a number of negative
consequences on the environment. Directorate General of Civil Aviation has started a project called "Green
Airport" in order to minimise these negative effects in airports. In this context, a "Green Organization" certificate
is awarded to the organizations which manage to establish an environmental operation system complying to set
of regulations set by DGCA. Organizations which have got this certificate can use the "Green Organization"
logo, and benefit from 20 % discount on payments for licence documents. When both the airport operator
organization and all of the other organizations operating in the airport comply with the regulations and
specifications, the airport is titled as "Green Airport". When this happens, discounts to the organizations is
increased to 50% (Canöz & Ertekin, 2020).

4.2. GDSAA and "Carbon-free Airport" Project

The purpose of "Carbon-free Airport Project" conducted by GDSAA is to accomplish a sustainable airport
managemenet by controlling the factors such as resource usage, air pollution, noise, waste, waste water,
chemical usage and carbon emission along with the all shareholders of an airport.
This project includes management processes of environmental outcomes of operations of airports operated
by GDSAA and/or terminal operators as part of KÖİ; and ground services in air side, maintenance, fuel and
catering facilities and other necessary service areas (Antalya Airport Chief Directorate, 2020).

4.3. Antalya Airport

According to Airport Council International's 2018 data, Antalya Airport is 14th busy airport in terms of
pessenger movement in the Europe. Also, accroding to GDSAA's 2019 end of December data it is the second
busy airport in Turkey after İstanbul Airport (Wikipedia, 2021b).
When examined green activities of the airport, we see following practices as part of the project "Zero Waste
Management" (Antalya Airport Chief Directorate, 2020):
Waste batteries are harmful to environment and public health because they contain heavy metals, and they
need to be collect and recycle separately. Specially designed collecting equipments for waste batteries are
located around terminals and isles in a way that passengers can access and use them easily.
Waste vegetable oil cause narrowing and blockings in the waste water systems, and this results in soil and
underground water pollution. Thus they need to be collect separately as well.
Waste electrical and electronic equipment are harmful to environment and public health because they
contain heavy metals, and they need to be collect and recycle separately. Moreover, their contents, which are
silver, gold and copper, are valuable to the economy when they recycled. Flourescent lamps collected in the
airport are stored in temporary storage containers.
Hazardous wastes are flammable, burner, carcinogen, explosive, irritant and poisonous wastes (toner
cartridges, contaminated packages, contaminated filters, pressurized containers and so on). Waste electrical and
electronic equipment and hazardous wastes in the airport are collected by the personnel of "Waste Department"
and are stored in temporary waste storages separately.
In this study, GDSAA Antalya Airport Directorate's 2018 Greenhouse Gas Calculating Report has been
examined. The purpose of this report (GDSAA Antalya Airport Chief Directorate, 2019); is to calculate the
greenhouse gas emissions and disposals from all activities in the responsibility of GDSAA; and preparing the
greenhouse gas statement in accordance with ISO – 14064-1 standard; and to provide data to the "Green
Organization Project" of DGCA.
By controlling the activities that may cause greenhouse gas emission, it is aimed reducing the total
Greenhouse Gas Emission per aircraft that uses the airport in the context of Green Airport Project by policies

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

273
such as reducing the harms to the environment, reducing greenhouse gas emission, maintaining and improving
energy management, reducing the consumption of energy and natural resources from personnel, suppliers and
outsources. With purpose of reducing and preventing the greenhouse gas emissions, Greenhouse Gas
Management Plan has been prepared.
For the purpose of reducing greenhouse gas in the airport, 10 pieces of 90 W sodium vapored armatures was
replaced with 60 W led armatures in the fire station outer spaces; 8 pieces 36 W armatures was replaced with 4
pieces of 20 W led flourescent lamps in the 1st floor of the New Technical Block building.
Direct emissions within the airport boundaries are stated as Content 1 in the greenhouse gas report, and they
are as follow: fuels used for heating, fuels used in firefight drills, fixed and mobile generators, propane cylinders,
fuels used by on-road and off-road vehicles, gas leakages from firefighting drills, cooling systems, power
distribution units.
Content 2 in the report consists of indirect emissions that stem from the electricity purchased by the GDSAA
from outside. This electricity is provided by terminal operator ICF.
Content 3 in the report is other indirect emissions which stem from landing and departing aircrafts to the
airport and their APU usage.
In choosing the emission factor, it has been used international factors (Tier 1) when the national factors are
not sufficant.

CONCLUSION
In parallel with the development of commercial activities, the demand for air transport that stands out among
the transport modes with its speed and safety advantage has increased, and the capacity of air transport has
grown. In order to meet the increasing air traffic, new airline companies are starting their activities while new
airports are being built. In addition to the economic gains it provides, airports cause environmental problems as
a result of fossil fuel use of aircraft and the activities at the airport. In order to reduce the negative environmental
impacts of aviation activities, “Green Airport” projects were developed by DGCA, and “Carbon-free Airport”
projects were developed by GDSAA. In the study, these projects were evaluated and the green logistics activities
carried out at Antalya Airport were examined. With the Green Airport project, it is aimed to minimize the
damage to the environment as a result of the activities of the organizations operating at the airports and to carry
out aviation activities in a sustainable way. The project obliges businesses to meet certain criteria within the
scope of the TS EN ISO 14001 standart. Businesses that create their environmental management systems by
meeting the specified criteria are given a “Green Organization” certificate and within this scope, businesses can
benefit from a certain discount in the fees they pay to the airport. With the green establishment certificate of all
establishments at the airport, that airport gains the title of “Green Airport”. In the “Carbon-free Airport” project
carried out by GDSAA, it is aimed to ensure the sustainability of the airports by fulfilling the conditions
determined in order to control the environmental effects of the activities at the airports. In line with these
projects, the green logistics activities carried out at Antalya Airport were examined. Temporary storage of
batteries, vegetable and motor oils, electronic goods and hazardous wastes generated at the airport is carried out
in accordance with their qualifications. Organic wastes are reintroduced to the economy by making compost. In
order to reduce greenhouse gas emissions at the airport, a Greenhouse Gas Management Plan has been prepared
with policies such as ensuring energy management, strengthening the environmental sensitivity of employees,
suppliers and other stakeholders through trainings. Antalya Airport carries out greenhouse gas measurements in
accordance with TS EN ISO 14064 - 3: 2007 standard. According to the 2017 measurements, the greenhouse
gas emission amount at Antalya Airport was 28,691.84 tons of CO2 equivalent and was certified as a green
airport by DGCA.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

274
 REFERENCES
Antalya Airport Chief Directorate (2020). Zero Waste and Environmental Awareness Training Document.
Banister, D., & Button, K. (1993). Transport, the Environment and Sustainable Development. London: Routledge.
Canöz, N., & Ertek, A. (2020). Contribution of Green Organization Certificate to Green Image Formation: A Research on
Turkish Civil Aviation. International Journal of Aeronautics and Astronautics, 1(1), 23-32
Durdağ (2020). A Qualitative Study On The Effect of Green Airport Certificate On The International Trade of Airport.
International Symposium on Air Transport and Regional Development Statement Book (pp. 42-47). Mersin: Toros
University.
General Directorate of State Airports Authority Antalya Airport Chief Directorate (2019). 2018 Greenhouse Gas
Calculation Report
Ekoyapidergisi (2021, May 30). Sustainable Airports from the World. Retrieved from
https://www.ekoyapidergisi.org/dunya-dan-surdurulebilir-havalimanlari.
Güngör, A. (2020). Environmental Sustainability in Airport Transportation. International Symposium on Air Transport and
Regional Development Statement Book (pp. 90-99). Mersin: Toros University.
Kumaş, K., İnan, O., Akyüz, A., & Güngör, A. (2019). Determination of the Carbon Footprint Originating from the Planes
of Muğla Dalaman Airport. Academic Platform Journal of Engineering and Science, 7(2), 291-297.
Mete, E. (2020). Green Logistics in the Scope of Sustainable Development: The Example of the European Union and
Turkey. Black Sea Journal of Social Sciences, 12(23), 383-396
Oto, N. (2010). Environmental Impacts of Airports, Environmentally Friendly Airport Planning, Implementation and
Operation Principles: Example of Esenboğa Airport. 1st Project and Construction Management Congress: (pp.
1203-1220). Ankara.
Özdemir, S. S. (2019). The Future of the Logistics Industry: Green Logistics and C - Footprint. Irdıtech 2019 International
R&D, Innovation and Technology Management Congress Proceedings (pp. 127-144). İstanbul: Okan University.
Şekertekin, Y. Y. (2017, February). Determination of Emissions from Turkey Domestic Flights. Master Thesis. Bülent
Ecevit University Institute of Science and Technology, Zonguldak.
Üstünbaş, N. N. (2018, May). Green Logistics Applications of Logistics Companies in Marmara Region. Unpublished
Master Thesis. İstanbul Commerce University The Institute of Social Sciences, İstanbul.
Wikipedia (2021b, June 11). Antalya Airport. Retrieved from https://tr.wikipedia.org/wiki/Antalya_Havaliman%C4%B1.
Wikipedia (2021a, May 30). Our Common Future. Retrieved from
https://tr.wikipedia.org/wiki/Ortak_Gelece%C4%9Fimiz.
Yesilaski (2021, August 19). What is the Kyoto Protocol and Why is it Necessary. Retrieved from
https://www.yesilaski.com/kyoto-protokolu-nedir-nicin-gereklidir.html.
Zengin, E., & Ocalir Akunal, E. (2017). Green Logistics Applications in Turkey. Journal of Management, Marketing and
Logistics, 4(2), 116-124.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

275
 BIG DATA TECHNOLOGIES: APPLICATIONS AND EVALUATIONS IN
THE LOGISTICS SECTOR
Faruk TAŞKIRAN1
Abstract: The transformation of business is being witnessed. From banking to retail, health to production, this
structure transforms every industry, with Big Data underneath it. The logistics sector is also the top of the most
affected sectors. Every transaction in the digital world leaves a data trail. Big Data from sensors that surround
our environment is an entry into today's world. Big Data from sensors that surround our environment is an entry
into today's world. While retailers use large data to predict trends and consumer behavior, technology companies
are developing direct marketing systems. Using large data intelligently and accurately, businesses have changed
their way of doing business with innovative, competitive ideas and have created new business areas. Companies
such as Amazon, Walmart, and Netflix have differentiated their business with Big Data usage. The whole world
is in this transformation. The main purpose of the study is to explain the benefits Big Data Technologies provides
businesses and the barriers to Big Data for businesses. The use of Big Data in the logistics sector analyzes the
summary printouts of important reports, providing recommendations for the Turkish Logistics sector, and
providing light to the future.
Keywords: Logistics, Big Data, Supply Chain, Digital Transformation

BÜYÜK VERİ TEKNOLOJİLERİ: LOJİSTİK SEKTÖRÜNDE

UYGULAMALAR ve DEĞERLENDİRMELER
Özet: İş dünyasının dönüşümüne tanıklık ediliyor. Bankacılıktan perakendeye, sağlıktan üretime her sektörü
dönüştüren bu yapının altında Büyük Veri yer almaktadır. Lojistik sektörü de bu değişimlerden en çok etkilenen
sektörlerin başındadır. Dijital dünyada yapılan her işlem bir veri izi bırakmaktadır. Çevremizi saran sensörlerden
sağlanan Büyük Veri, günümüz dünyasının bir girdisi konumundadır. Perakendeciler büyük veriyi trendleri ve
tüketici davranışlarının tahmin etmek amacıyla kullanırken, teknoloji şirketleri direkt pazarlama sistemleri
geliştirmektedir. Büyük Veriyi akıllı ve doğru şekilde analiz ederek kullanan işletmeler, yenilikçi ve rekabetçi
fikirlerle iş yapış şekillerini değiştirmiş ve yeni iş alanlarını ortaya çıkarmıştır. Amazon, Walmart, Netflix gibi
işletmeler Büyük Veri kullanımı ile işletmelerini farklılaştırmıştır. Tüm dünya bu dönüşüm içerisinde yer
almaktadır. Çalışmanın temel amacı Büyük Veri’nin işletmelere sağladığı avantajları ve işletmeler açısından
Büyük Veri’nin önündeki engelleri açıklamaktır. Lojistik sektöründe Büyük Veri kullanımı, önemli raporların
özet çıktıları analiz edilerek, özellikle Türk Lojistik sektörü için öneriler ve tavsiyeler sunularak geleceğe ışık
tutulmaktadır.

Anahtar Kelimler: Lojistik, Büyük Veri, Tedarik Zinciri, Dijital Dönüşüm

1
Research Assistant Faruk Taşkıran, Istanbul Esenyurt Uni., Logistics Management Department, Istanbul, Turkey,
[email protected]

276
 INTRODUCTION
As researcher author and educator David Buckingham pointed out, ‘Data is new oil. We need to find it, extract
it, develop it, distribute it, and make money from it’ (Issa, 2016 :10-13), (Uldrich, 2013: 25). Today's new mine,
data, can be found not only in the soil, but in the pocket of every individual in today's World. From mobile
phones to the refrigerator at home, to the cargo that comes home from the website shopped, to the internet
content that watched, to the public transport that used, contributed to data production every minute as an
individual. Businesses, businesses' tools and tools, our world digitized by social networks, became a data slot.
This huge data slot is called Big Data in the literature.
Today, Big Data can be more valuable than oil. Oil is a limited and depletable mine, whereas today's Big Data
is growing by feeding on 5.2 billion people with mobile phones every day, 4.6 billion people connected to the
Internet, and 4.2 billion people actively using social media (Kemp, 2021: 5). In addition, British Airways, flying
from London to New York, MSC, Ford, making automobile, Amazon running an E-commerce platform,
produce and store data every minute in gigabytes, petabytes, zettabyte sizes. This huge network can become
even more valuable than oil.
Businesses that recognize this change are investing in Big Data technology, industries are becoming digital, and
every day these huge data are analyzed to ensure operational efficiencies, and new sales strategies are being
developed by reviewing customer behavior, and many new projects are being implemented. Governments are
developing new security protocols, and the medical world is able to analyze treatment processes faster. The
power behind this change is derived from data, sensors, signals, social media tools in structured and unstructured
types. While Netflix develops intelligent algorithms to explore our likes, Amazon analyzes our shopping
behavior. While General Motors is performing efficiency and supply chain network optimizations, the UPS is
analyzing more road travel with less fuel. Today, all sectors are looking for more market share by developing
new business models thanks to Big Data technology.
In the first section of this study, Big Data Technology will be examined. The benefits and disadvantages that
Big Data Technology has brought to the business world will be discussed. The second section will examine the
literature on Big Data Technologies in Logistics and Supply Chain. The third section describes the usage areas,
benefits and industry innovations of large data in the Logistics and Supply Chain industry by providing examples
of businesses. The conclusion and recommendations section interprets important information from work on Big
Data and offers recommendations on the investment of Big Data technology in Turkey’s Logistics Industry.

WHAT IS BIG DATA TECHNOLOGY?

Over the past two decades, data production has increased dramatically (Warden, 2011: 32). This data can be
collected from any digital location in our lives, such as social networks, sensors, etc. Big data is described as a
large amount of data sets that cannot be analyzed and controlled with customary data processing tools (Ohlhorst,
2012: 42). In other words, almost every set of data has the potential to be big data because the information
provided by these data for many organizations has a high value. As large companies adapt quickly to this data,
small companies expect some disadvantages. At the beginning of these disadvantages, software engineering
skills are required to use this data. Secondly, some software languages need to be dominated, and a third but
major challenge is that the cost of investing in this area is high.
Today, data is collected and stored in very different formats and stored in a structured, non-structural and semi-
structural manner. In order to obtain information from data in large data technology, multiple servers must run
in parallel. Due to the increasing economic value of Big Data Analysis systems, many industry pioneers are
moving their investments into this area. Facebook, Yahoo, Amazon, Google, LinkedIn, GE, Zynga, Netflix, etc.
World-famous businesses are making big investments in Big Data Analysis systems (Davenport & Dyché,
2013:1-31). In 2012, the business rate that invested in big data technologies was at 58%, and in 2013, it increased
to 64%. According to a study conducted in 2013, 91% of business executives stated that they are working on
big data. 60% of these businesses have completed at least one big data initiative and 32% have benefited from
big data in production. These numbers show that big data in businesses has become a big reality (Davenport &
Dyché, 2013: 1-31).
Another issue that provides the capital functionality of large data is that it allows new business models to be
developed. Big data reveals a great economic value with detailed analysis of customers, organizations, products,
markets and all other business elements. It has become possible to do predictive market segmentation with large

277
data, use of complex analysis techniques related to consumer decision-making processes, and to consult new
methods in development/improvement of goods or services (Gürsakal, 2021: 35). As a company that transforms
big data into value in our country, it can be given as an example of a Yemeksepeti. The use of Big Data explained
that it gained a value added to the company by 4.5 million Turkish Liras in 2015. The Yemeksepeti CEO said
the company was structured with Big Data technology and that they could achieve meaningful results even from
a small amount of stored data, such as 1% (Marketing Türkiye, 2016: 1).
It is defined in the literature where the big data has three key characteristics. These features, defined as Volume,
Variety and Velocity are located in the source in 3V. In some recent studies it is important to note that
Verification and Value additions are also defined as 5V.
Volume
The Volume is used to indicate the size of the data. Today's data has reached petabyte, zettabyte levels. The
digital world contains almost 2,7 zettabytes of data. On social media, Facebook 100 Terabyte data upload takes
place, 230 million tweets are posted daily. 2,9 billion e-mails are sent to the workplace. High technology
company like Google, Microsoft, and Amazon have own big data solutions, but are not open source code and
medium-sized organizations are not easy to access (Lam, 2010:10-20). Digital data is very large in terms of
volume and diversity. In an article published in Economist magazine, Wal-Mart is 167 times the number of
books in the United States Library of Congress, where more than 1 million customers are doing online
transactions and the digital data generated from these transactions is more than 2.5 petabytes (Cukier, 2010:5-
20).
Variety
The diversity that represents the source of the data types indicates that large data is not the same type and
structure. Data can be collected from a wide range of data sources such as mobile phones, sensors, sensors, and
social media. Data that varies as a source is separated as a structure. Data structures are divided into three
structures, structural data, semi-structural data, and non-structural data. Examples of structural data, user
feedback, e-mail semi-structural data, images, audio, or text files in the relational database, or data held on
specific forms are examples of non-structural data. 95% of today's data consists of non-structural data (Cukier,
2010:5-25).
Velocity
Velocity is one of the key features of data from various sources. This feature is not only related to the speed of
incoming data, it also includes the storage speed of incoming data. The speed of data produced over the past
day is increasing, and this increase increases the number and diversity of transactions that need data. The speed
of data is one of the most focused components in the research world (Satyanarayana, 2015:115).
Value
One of the most important features of big data is to create value. Value is treated by Oracle as a fourth feature
in addition to other features of big data. As a result of the components described earlier, large data must be
generating a plus value for the business after the data production and processing phases. The fact that businesses
can instantly access data in the decision-making process and that data is effective in making the right decision
is a sign that big data adds value to businesses (Kaisler, Armour, Espinosa, & Money, 2013:50-60).
Verification
Another feature added to the 3V main component of large data in the following years. Today, data is huge and
validation of data is a key feature. Data must be monitored, privacy protected, and stored safely during data
streaming. Huge data stacks include big data technology, analyzing the reliability of data and among the main
components (Walker, 2015:181-183).
What's Hadoop?
The most used technology in Big Data Technologies is called Hadoop. It is a technology that was first launched
by Google and developed by Hadoop open-source code software, which has been implemented in many different
areas. Hadoop technology is an open-source code technology designed to process data of structural and non-

278
structural terabyte, petabyte, etc. Hadoop technology works as cluster structures created by combining ordinary
servers. It is an open-source code library developed with Java programming language (Hurwitz, Nugent, Halper,
& Kaufman, 2013:130-136). Unlike conventional database technologies, scanning on Big Data is designed to
produce fast results because it is a distributed structure. With Hadoop technology, the Yahoo can handle 40
petabytes of data per day on 40.000 clusters. Many international companies use Hadoop technology to create
new models and new applications using data that is too large. Examples of companies that develop models and
applications are Amazon, Facebook, Twitter, Rolls Royce and many of the most important companies with Big
Data technology (Zikopoulos, and others, 2012: 10-20).
In addition to Hadoop technology, it can be list other technologies used by many businesses in Big Data analytics
as follows (Henschen, 2014:1):
• Action Analytics Platform • Microsoft Bigdata
• Amazon Web Service • MapR
• Cloudera Enterprise Bigdata • Tableau Software bigdata
• Hortonworks Data Platform • Oracle Bigdata Analytics
• IBM Bigdata Analytics, • Teradata Bigdata Analytics
• HP Bigdata • Pivotal Bigdata
• SAP Bigdata Analytics • Google BigQuery
Benefits of Big Data Technology
NASA (US National Aviation and Space Agency), NOAA (National Ocean and Atmospheric Organization),
pharmaceutical companies, large-scale energy companies use, analyze and interpret large data to make decisions
with consistent and accurate data on a daily basis. NOAA uses big data for climate, ecosystems and weather
forecasts, while NASA uses Big Data for aerospace and space research. As pharmaceutical, space, and energy
companies conduct feasibility and profitability research, the New York Times, the largest media organization,
utilizes Big Data for text, content analysis and web mining (Ohlhorst, 2012:19-21).
In a research conducted at MIT (Massachusetts Institute of Technology), businesses running data-based
decision-making processes reported that they increased profits from around 5% to 6% (Brynjolfsson, Hitt, &
Kim, 2011:1-33). In companies that use data very efficiently, there is a very strong link between the effective
management of data and financial performance. In a research report by McKinsey Global Institute (MGI), the
transformative effect of big data clearly shows that it has an impact on all sectors, from healthcare to retail and
manufacturing to political campaigns (Manyika, and others, 2011:47-50).
On the basis of the growth of the big data market; there are contributions made by Big Data Analysis in the
fields of R&D, marketing, sales, operations management and supply chain management. The key values
generated by Big Data Analysis can be shown as follows (Grover, Chiang, Liang, & Zhang, 2018:50-60):
• Big Data leads the creation of added value by integrating existing data, contributing to improving and
improving service quality in businesses.
• Big Data produces more positive results by increasing links between different data.
• Big Data contributes to the development of new options through process analysis in areas dominated by
uncertainty.
• Big Data contributes to the capabilities and infrastructure of the organization, increasing capacity and
creating great innovations and opportunities.
Data is indispensable in all businesses. McKinsey's report states that there is no industry left in developed
countries such as the United States with data under 200 turbs. McKinsey Global Institute has listed some
value-enhancing benefits of Big Data use as follows (Manyika, ve diğerleri, 2011:47-50):
• Big data adds value to businesses by making information more frequently and transparently available.
• As businesses digitally store the data they produce more frequently, they can get more detailed and
accurate information. It helps to improve business performance through analysis of data in various
structures.
• It makes provide personalized products and services by making the customer layers even more obvious.
With big data usage, businesses can easily access potential customers by reviewing consumer behavior.
• Big data makes decision-making processes significantly easier by analyzing data from a variety of data
sources, such as mobile devices, radio frequencies and websites.

279
• The next generation of products and services can be developed. It enables the after-sales services to be
increased from the data provided through sensors in manufactured products.
• Big Data science can create new jobs and job fields.
• Big data integrates both structured and unstructured data, making it easy to detect, prevent, and recheck
financial fraud.
The UPS company, which provides a productivity boost as a result of Big Data Analysis, made it possible in
2011 to add GPS to all of its vehicles and, as a result of analysis of data from GPS, all of the drivers traveled
the most appropriately, providing 48.300.000 km less. This savings enabled the company to reduce emissions
of 3 million tons of fuel and 30.000 metric tons of carbon dioxide, and transformed the benefits of big data into
a value for the company (Mayer-Schönberger & Cukier, 2014:132-133).
Barriers to Big Data Technology
Given that the big data is new to organizations, they need to learn how to use new technologies as soon as they
get into the market. It creates a competitive environment for these businesses. Big Data creates many innovative
solutions, business models, and application areas for businesses, but there are some challenges to big data, such
as collecting and analyzing it for businesses. The next generation of traditional information technologies may
be inadequate in storing, analyzing and interpreting data in large, varied and complex structures. New
technologies need to be created to manage complex and high-volume data that is not configured, semi-
configured (Hu & Kaabouch, 2013:12-17).
Problems that arise from the lack of Big Data analysts in businesses are often coming to us today. Businesses
recently need personnel with data science skills. Davenpont and Dyche describe these staff as individuals who
can perform video and image analysis, perform natural language and text mining, have visual analytical skills,
and are able to use programs such as Python, Pig, Hive. These features are among the most sought-after features
that can report data and adapt to technology. It is important that new data scientists combine their data scientist
skills with traditional data management skills (Davenport & Dyché, 2013:15-17).
Here are some of the disadvantages in front of businesses (Chen, Mao, Zhang, & Leung, 2014:25-35):
• The biggest hurdle to Big Data technology is the need for IT professionals. In the report published by
MGI, there is a need for more than 190.000 analytical experts in the United States and more than 1.5
million data literacy principles (Manyika, and others, 2011: 16). It is a disadvantage because businesses
that are weak in Information Technologies expertise are either training their own staff or moving on to
hiring experts will increase investment costs.
• Another disadvantage of big data is privacy and security. The fact that data is in complex structures is a
huge cost to store and store them for businesses. International mergers should be noted that there are many
legal issues regarding privacy.
• The lack of technological infrastructure is another problem facing businesses to effectively use large data
and to conduct real-time analytics.
It can be given businesses an example of IBM's work for those who are having difficulty integrating Big Data.
In talks with 524 senior company managers from 56 countries, 35% of managers stated that they had difficulty
integrating Big Data sources, 30% said they didn't know how to do the analytical approach and 22% said that
data access was insufficient (Bloomberg Businessweek Türkiye, 2014:2).
Although Big Data technology has some barriers to it, it's still worth it for many businesses. Businesses can
focus on creating value by exceeding the barriers identified by investments to be made.

LITERATURE
Rogers and Tibben-Lembke (2001) indicate that logistics costs are an important place in the United States.
About 9.9% of the American economy is logistical costs. The study states that the reverse logistics costs of the
companies examined account for 4% of the total logistics costs. In this work, it emphasized the importance of
data in reducing reverse logistics costs, and artificial intelligence said machine learning technology would
reduce costs (Rogers & Tibben-Lembke, 2001: 129-148). In the literature, it is possible to find studies on the
frequent benefit of Big Data in the fleet management systems in transportation of air, land, sea and railways.
Big Data Technology and analysis monitoring and momentary monitoring of the positions of vessels in maritime
transport (Fabbri, and others, 2012: 2404-2410), reducing costs in airway transport, regular control of aircraft

280
maintenance etc. (Zhang, and others, 2015: 1-8), from road transport to patient transport in many different areas,
improving efficiency, saving time, optimizing rotalization, transportation safety, Big Data technology is utilized
(Ilocca, and others, 2017: 1899-1910), (Gogoulos, and others, 2013: 49-54), (Stancel & Surugiu, 2017: 861-
867), (Johanson, Belenki, Jalminger, M., & Gjertz, 2014: 736-741), (Gowda & Gopalakrishna, 2015: 417-421),
(Bélanger, Kergosien, Ruiz, & Soriano, 2016: 216-229). Examples of large data technology are available for
cloud-based fleet management and vehicle charging energy optimization for electric vehicle fleets (Ostermann
& Koetter, 2016: 1-11).
Channels such as sensors, devices, websites, and social media enable a wide range of data to be collected and
analytics can be achieved by using the Internet of things, Big Data technologies (Zhong, Xu, Klotz, & Newman,
2017: 616-630). One of the many concepts that have been re-defined with Industry 4.0 has also discussed
transportation logistics and data logistics in Hofmann and Rüsch's work (Hofmann & Rüsch, 2017: 23-34).
These discussed concepts should work in line with smart factories (Baretto, Amaral, & Pereira, 2017: 1245–
1252). Witkowski described the technologies that should be used in its study in collaboration with industry 4.0
as big data analysis, augmented reality, autonomous and green tools, the Internet of objects, robotic studies,
drones, and Big Data said that it allows information from different databases to be combined and new
information to be derived (Witkowski, 2017: 763-769). Many researchers categorized strategies for using Big
Data technology in logistics and supply chain areas (Wang, 2016: 1-18). The digital supply chain selection was
carried out by the multi-criteria decision-making method in the Buyuközkan and Guler study and determined
supplier criteria in accordance with industry 4.0 (Büyüközkan & Güler, 2019: 21-47). Ivanov and Dolgui
conducted a dynamic algorithm proposal with Big Data analysis techniques for short-term supply chain charting
for use in smart factories (Ivanov & Dolgui, 2020: 1-15). Within Industry 4.0, Big Data analysis, in workflows,
production, supply chain, managing marketing-sales processes has contributed greatly to its optimization
(Ardito, Petruzzelli, Panniello, & Garavelli, 2019: 323-346). Big Data technology helps reduce costs and prevent
risks (Govindan, Cheng, Mishra, & Shukla, 2018: 114). It can be designed to plan big data, prepare effective
inventory and create dynamic logistics plans (Linh & Huong, 2020: 27).

281
 BIG DATA IN LOGISTICS IN SUPPLY CHAIN

Figure 1. Big Data in Logistics (Jeske, Grüner, & Weiß, 2013: 16).
Big Data technologies relate directly to the increase in internet usage. This led to a new job site. This new
business site is third-party businesses that provide storage and analytics services for using Big Data
technologies. Big Data, which is vital in today's digital world, is a tool used in creating strategy, in the fields of
supply chain and logistics, in decision-making processes such as production, human resources, sales, marketing
and finance.
Once popular, ERP systems were used to gather customers, distribution centers, suppliers and production on a
platform. ‘Who is the right customer for us?’ Its question led to the emergence of CRM systems. With these
systems, the right product or service made it easier to offer the right customer, the right channel, the right place
and time at the right price. This means that the product cycle has now started according to the customer, not the
customer by product. The concept of 'Big Data' in academic literature; It is the process of collecting, storing,
cleaning, visualizing, analyzing and generating meaningful results of data that comes in large volumes, high
diversity and quickly from all kinds of data producing tools (Gürsakal, 2021:15).
Logistics is defined as the stage of the supply chain process that plans, executes and controls the storage of
goods, services and information to ensure accurate, effective, efficient flow from the production stage to the
final phase of consumption to meet customer requirements (Larson & Halldorsson, 2004: 17-31). Logistics
industry pioneers have managed to provide efficient and efficient operations from production to consumption
with Big Data technology at every stage. Logistics businesses that provide services in the international arena
with the size jump of globalization must effectively control all their processes to succeed. Businesses that have
been successful in their logistics processes have been able to have a large part of the market (Witkowski, 2017:
768).
Digitization has changed business models in society and business areas. The Logistics and Supply Chain has
been significantly affected by this change. As with all stakeholders serving in logistics, third- and fourth-party
logistics companies must benefit from Big Data in order to provide cost benefit, demand estimation,
predictability, better service. The industry needs to adapt to changing dynamics as quickly as possible. It's
important to develop different applications to provide customers with an uninterrupted brand experience, deliver

282
personalized marketing services for customers, digital experience, and rich payment options to survive (PWC,
2016: 1-20).
In 2018, the World Economic Forum report identified the technologies that would transform the logistics sector
as follows (Schwab & World Economic Forum, 2018: 1-50):
• Information Services: The network of active information services will contribute to optimizing costs while
maintaining operational efficiency.
• Logistics Services: Logistics services to be digitally efficient will contribute to the growth of the trade
network.
• Delivery capabilities: With the development of delivery paths to end-user, drones, autonomous vehicles
will contribute to development in many areas with 3D printing.
• Conversions Economy: Sustainability, reducing air pollution, waste management and logistics will be able
to prevent environmental damage.
• Sharing Economy: Shared logistical services and areas such as storage, transport lines will contribute to
transformation.
The competitive logistics industry must set up their own data chains to keep pace with this transformation, and
develop their ability to utilize Big Data Flow to improve efficiency and satisfaction (Schwab & World Economic
Forum, 2018: 1-50).
The transformation of the logistics sector is a sign of this change, where the large capital and virtual companies
appear instead of investment. Transport vehicles, small capital owners without tanks, can be matched to more
efficient and competitive platforms that provide customer and logistics services, resulting in more economic
and efficient results (PWC, 2016:1-20).
At this point, it can be show Uber as an example, which is not a freight company. Uber offers UberCargo and
UberRush services in Hong Kong and open service for online retailers. The US-based dolly company has
managed to connect freight engine owners and cargo transmitters to a platform for urban cargo delivery. We
see new initiatives and investments in the logistics sector thanks to businesses that digitally deliver many
transactions such as e-customs, route planning, augmented reality, warehouse management, real-time data-based
tracking systems, paperless operations flow, dynamic pricing (PWC, 2016:1-20). In addition, it can be consider
GETİR's activities in Turkey, the UK and the Netherlands, which benefit from Big Data technologies, as a retail,
customer-connected enterprise in the logistics sector.
Benefits of Big Data in Logistics
1. Real time Route Optimization: Big Data enables dynamic calculation with data such as traffic status, order
of receivers. Cost savings are achieved with automatic roasting (Wegner & Kückelhaus, 2013:18).
2. Strategic Network Planning: Analysis from historical data to realize strategic investments, and long-term
network mapping and deployment planning can be achieved through Big Data.
3. Operational Capacity Planning: Big Data Technologies are utilized to measure the impact of external
impacts and internal factors and to build predictive models while performing short- and medium-term
operational planning (Wegner & Kückelhaus, 2013:21).
4. Customer Relationship Management: Customer database analyzes and customer interaction and operations
data and performance and prediction models can be performed. As the customer touchpoints become a
wide range of contacts, Big Data is used to measure the satisfaction of their customers.
5. New Product and Service Manufacturing: Customer requests and expectations in the Logistics sector are
an important part of the business. Therefore, it is important to increase quality and diversity based on
customer feedback. With Big Data, customer requests can be responded to momentarily.
6. Risk Planning: Supply chain cuts have become important for the world, as the Covid-19 outbreak has
shown. Big Data is essential to ensure processes are flexible, including all supply chain processes, as well
as to develop case forecasts and alternatives that will cause disruptions. Predictive risk assessments can be
used to respond quickly to customer requests. Data is obtained from many sources such as politics,
economy, health (for example, social media data, blog data, weather forecast data, website data, stock
data) and then the collected data is analyzed. Most of this information stream is not configured and is
constantly updated. Therefore, big data analytics ensures that input is obtained which makes sense in the
supply chain and especially in the determination of risks in logistics (Wegner & Kückelhaus, 2013: 1-50).
7. New and Actual Market Information: Thanks to big data, it contributes to the production of market
information reports for medium and small businesses without strong logistics infrastructure. As long as

283
 shipping records data and postal identity are kept confidential, logistics providers can review this data to
verify existing external market research. With regression analysis, sensitive information in the shipping
database can significantly increase the sensitivity of traditional demand and supply estimates.
8. Financial Demand and Supply Chain Analysis: Contributes to the consolidation of global supply chain
data with economic data in improving the rating and investment decisions of financial institutions. It also
contributes Big Data to creating more efficient, efficient, cost-effective processes with analysis of data
collected from the supply chain processes (Wegner & Kückelhaus, 2013: 1-50).
9. Environmental Intelligence: Digital urban systems and data from many different areas are provided. In
addition, it enables the collection of data, such as traffic density, noise, etc., via sensors connected to the
vehicles. With the data collected, it is possible to ensure that both the proper structure of urban
transformation and the analysis of factors that affect logistical performance (Wegner & Kückelhaus, 2013:
1-50).
In companies that can use big data, they can make investments in logistics. This framework is also an example
of the Amazon Company. Seeking to improve storage capabilities in the delivery area, the business has
purchased a warehouse automation company that is part of the Amazon robotic Business Unit. It has rented 20
planes to gestation his own shipment, and is conducting 30-minute delivery experiments with the Drons (PWC,
2016:1-20). The Chinese Alibaba company provides quality service through Big Data using the Cainiao
Company, which manages transport, storage, port logistics operations (PWC, 2016:1-20).
As an example of Big Data usage, the Big Data work of the American retail chain Target. By reviewing their
customer databases, the business that performs customer analysis has increased sales by sending brochures,
coupons, e-mails to customers based on pregnancy months from the records of their customers in the baby's
department. Target's 2002 revenues increased by $44 billion, and in 2010, it increased by $67 billion (Duhigg,
2012: 30).
Big Data and marketing and retail activities have come in a much different dimension than conventional
meaning. Online platforms companies like Amazon, Alibaba, E-bay have a lot of customer data. With analysis
of this data, there are studies to improve profitability in activities such as various sales-marketing techniques
and cost optimization warehouse location selection (Kumar & Jothimani, 2017: 1-24).
As with other areas of logistics, a lot of data is produced in aerospace. In addition to the data provided by the
sensors on the aircraft, the personal data of the occupants collects a wide range of data such as weather data, air
traffic data, aircraft maintenance and repair data. The average Atlantic flight provides a data production of 1000
gigabytes. Large volumes of data generated are an important resource for developing management, security,
cost, and sales-marketing with accurate analytics (Sumathi, S, Ramana, Muralitharan, & Kamal, 2017: 129-
132). It can be sort out some examples of airlines that are trying to use the benefits of big data.
• The ‘collect, detect, act’ application provides personal, customer-focused marketing service by reviewing
customer historical preferences through the Big Data analytics of United Airlines and analysis of 150
variables. With this application, revenue has increased by 15% (Williams, 2016: 16-20).
• With the ‘Know Me’ application, British Airways offers personalized offers to customers. Feedback from
customers identified positive results (Gonzalez-Feliu, 2019: 47-50).
• The Easyjet Airline has developed its algorithm, which charges aircraft seats and amenities with artificial
intelligence, by being duped by Big Data technology. This algorithm indicated that it has increased 20%
profit by implementing a dynamic rate policy based on line, tariff, passenger density, weather, etc. data
(Gonzalez-Feliu, 2019:47-50).
• Southwest Airlines makes it easy to create measures by using AI algorithms to analyze anomaly states using
Big Data for airline security in accordance with NASA's agreement. With his agreement with General
Electric’s Aviation, he has recently calculated flight data, such as humidity, wind, temperature, aircraft
weight and speed, and saved fuel for his aircraft (Gonzalez-Feliu, 2019:47-50).
• Delta Airlines has developed an aircraft maintenance algorithm using Big Data technology. Since 2003,
SmartSignal has been able to prevent 1.200 potential annual flight cancellations (Gonzalez-Feliu, 2019:47-
50).
• According to IBM's report, the Asian airline is reviewing its "revenue leak" application and other airline
customers, delivering special offers for flights that are not made with it. The business, which provides 2%
return on the app, has indicated that it is providing $9 million in revenue. With Big Data, airline businesses
can analyze competitor behavior, seasonal, market opportunities, and more than $20-30 million in revenue
potential (International Business Machines Corporation (IBM), 2014: 1-48).

284
 Figure 2. Data from the Airline (Andronıe, 2015: 154)
The disadvantages of Big Data Technology in the logistics sector are that of being industry-based, and that of
the core components of Big Data Technology. The negative aspects in the ‘Disadvantages of Big Data
Technology’ section of the article apply within the industry. In addition to these disadvantages, because the
industry is very dynamic and varied, it can be add a rapid adaptation process to technology. It should be taken
into account that small capital sector stakeholders have a cost disadvantage in adopting technology and creating
new models.

285
 CONCLUSION

The overall evaluation of Big Data Technology described above in this section consists of examining data from
particularly noted reports and recommendations for the Logistics Sector in Turkey.

Figure 3. Social and Technological Trends in Logistics (Jeske, Grüner, & Weiß, 2013: 11)

Research undergoing a review of technologies and social trends that will affect DHL's published Logistics sector
will have a high impact on Big Data and Big Data service delivery technology, which will take place over 5
years. Other technological and social trends that will influence the logistics sector can be reviewed in Figure 3.

Figure 4. Corporate Big Data Strategy (Jeske, Grüner, & Weiß, 2013:16)

When asked whether businesses have yet to create and create a strategy on the Big Data Strategy, Figure 4 in
the same report shows the results and shows that 63% of businesses are not a Big Data technology strategy. In

286
this international study, it is important that the current situation be established as soon as possible by working
on businesses operating in Turkey.

Figure 5. Sectoral Digitalization Rate (MÜSİAD Müsatakil Sanayici ve İşadamları Derneği, 2020:21)

The report published by Müsiad has investigated the level of digitalization of the sectors. The digitalization
level, which is contained in the Big Data technology, is also included in the report, and is rated at 57.1% in the
Transport and Logistics sector. In the Logistics sector, where technology and data are produced very intensively,
it is important for the future to move this level to higher levels.

Figure 6. Technologies Used In Companies (MÜSİAD Müsatakil Sanayici ve İşadamları Derneği, 2020:76)

Although there are many technologies in the Müsiad report to ask participants which technologies you are using
in your company, Big Data technology has not been able to find itself. Considering all sectors, it can be thought
that there are big gaps in the awareness and usability of Big Data Technology in Turkey and that it is a weak
trend in the future. In particular, businesses need to take quick action on the processing, analysis and creation
of new projects, which are more valuable than oil today.

287
 Figure 7. Barriers to Digitization in Companies (PWC, 2016:15)
Asked in another study by PWC, ‘What are the biggest challenges in building/developing digital capacity in the
business?’ The lack of Digital Culture and Education is a close to 50% in the responses to the question. High
financial investment and lack of talent are among the other main reasons. Logistics and Transit are stated to be
the biggest competition area in the next period. For Turkey, the situation is not different from the world, but is
particularly important to delay investments in digital transformation and Big Data Technology in order to
achieve competitive advantage and sustainable benefits.

Figure 8. The Importance of Big Data Technology in Business (Accenture Analytics, 2014:3)

288
The study conducted by Accenture looked at the importance of Big Data Technology for organizations. On
average, 59% of the large data in businesses has been found to be significant. Businesses with a property size
of more than $5-$10B and $10 show 60%. The larger businesses are, the Big Data investments are, the more
they can be removed from here. Compared to businesses in Turkey and these businesses, the rates in Turkey are
considered to be even lower.

Figure 9. The Challenges of Big Data Projects Within the Business (Accenture Analytics, 2014:5)

Figure 10. Solutions in Business for Big Data Projects (Accenture Analytics, 2014:5)

Another important point in the Accenture report, what is the biggest challenge in the business ahead of Big Data
Projects. In reviewing the report results in Figure 9, the Security and budget headings were identified as the root
cause of the business problem. Another major challenge is that there is a lack of talent to use Big Data
technology. The criteria set out as the main challenges for businesses investing in Big Data technology in Turkey
are unspoken. In particular, finding the skills to use security and the Big Data Technique, the major challenges
in attracting the business can be considered the biggest obstacle to the Turkish business.
Expert support from outside source, workshops, independent research has been stated to be used significantly
to address the challenges outlined in Figure 10. It can be assumed that these methods will be valid for businesses
operating in Turkey. In particular, infrastructure and talent transfers from suppliers directly in Big Data
technology can be stated as an important consideration in passing the rapid adaptation process.

289
 Figure 11. Big Data Investment in Business (Accenture Analytics, 2014:9)

Another important outcome in Accenture research has been the question asked over Data Scientist employment
in the coming years. When the answers are reviewed, the vast majority of businesses will be employing Data
Scientist jobs over the coming years, while nearly all of them have indicated to their businesses that they will
be employing Data Billet within the next 5 years. From here, it is important that businesses in Turkey quickly
develop and employ Data Science departments in these areas. In particular, the Data Scientist stack is an
important issue to increase the availability and effectiveness of Big Data technology.

Figure 12. Potential for Big Data (Accenture Analytics, 2014:10)

A key question in the Accenture study, when asked in the next 5 years what area Big Data technology would
impact, a large majority of the answers said would be effective in Customer relationships. The second impact is
product development and third, responses to changes in operations. Considering these impact levels, businesses
providing services to the Logistics sector, especially in customer relationships and product development should
quickly adapt to Big Data technology and participate in initiatives to make improvements and investments in
this topic.

290
 Figure 13. Who Runs The Change Process? (MÜSİAD Müsatakil Sanayici ve İşadamları Derneği, 2020:80)

According to the answers to the question of who is conducting the digitalization processes in the company in
the Müsiad research, it is observed that the Firm Partners/Owner response in Turkey has received the highest
percentage. In the institutional framework, this percentage is expected to be in more digital transformation,
R&D or Corporate Strategy units. The process of managing the process of digitization and adaptation to new
technologies in Turkey is a subject that needs to be reviewed again.
First of all, the results of the research on Big Data Technology awareness in Turkey could not be shared because
sufficient data could not be reached. With an application where participation is high and more data is accessed,
researching Big Data Technology awareness and determining the results in the Logistics sector in Turkey is
presented as a suggestion for future research.
It provides significant contributions to the Logistics sector as well as many sectors of Big Data Technology. It
brings significant benefits to many functions, especially in examples of airline businesses. It is clear that the
cargo business, warehouse businesses and supply chain network offer a significant competitive advantage
worldwide for the business in every ring. Considering this scale, it is important that the Logistics sector in
Turkey participate in competition with a rapid adaptation process.
These technologies should be dedicated to the establishment and support of departments such as Digital
Transformation, Data Science, and support of new projects by providing budgets. The main barriers of Big Data
technology should be analyzed well and the goal of achieving the world's competitive level by establishing
future perspective strategies with state, private sector and university cooperation on solution proposals.
Opportunities for businesses to use Big Data Technology, provide financial support to projects, and provide
incentives to grow and employ competent employees should be increased. In future studies, it is recommended
that Logistics sector and stakeholders operating in Turkey be determined and developed specifically on the level
of awareness and availability of Big Data Technologies.

REFERENCES
Accenture Analytics. (2014). Big Success With Big Data Executive Summary. Accenture Big Success with Big Data
Survey, 1-14.
Andronıe, M. (2015). Airline Applications of Business Intelligence Systems. Incas Bulletın, 153-160.
Ardito, L., Petruzzelli, A. M., Panniello, U., & Garavelli, A. C. (2019). Towards Industry 4.0: Mapping Digital
Technologies For Supply Chain Management-Marketing İntegration. Business Process Management Journal,
323-346.
Associated Press. (2021, June 06). The Retailer's Guide To Big Data. Entrepreneurshipinabox:
https://www.entrepreneurshipinabox.com/images/infographic-the-retailers-guide-to-big-data/
Ayvaz, S., & Salman, Y. B. (2020). Türkiye’de Firmalarin Büyük Veri Teknolojileri Bilinirliği Ve Kullanimi Analizi.
European Journal of Science and Technology, 728-737.
Baretto, L., Amaral, A., & Pereira, T. (2017). Industry 4.0 Implications in Logistics: An Overview. Procedia
Manufacturing, 1245–1252.
Bélanger, V., Kergosien, Y., Ruiz, A., & Soriano, P. (2016). An Empirical Comparison Of Relocation Strategies İn Real-
Time Ambulance Fleet Management. Computers & Industrial Engineering, 216-229.

291
Bloomberg Businessweek Türkiye. (2014). Müşterinin ve Büyük Verinin Yükseldiği Dönemde Pazarlama Üst
Yöneticilerini Neler Bekliyor? Özel Ek. İstanbul: Bloomberg Businessweek Türkiye, 16-22.
Brynjolfsson, E., Hitt, L. M., & Kim, H. H. (2011). Strength in Numbers: How Does Data-Driven Decisionmaking Affect
Firm Performance? SSRN Electronic Journal 1, 1-33.
Büyüközkan, G., & Güler, M. (2019). A Methodological Approach for Logistics 4.0 Technology. Journal of
Entrepreneurship and Innovation Management, 21-47.
Chen, M., Mao, S., Zhang, Y., & Leung, V. (2014). Big Data Related Technologies, Challenges and Future Prospects.
USA: SpringerBriefs in Computer Science.
Cukier, K. (2010). Data, Data Everywhere. The Economist.
Davenport, T. H., & Dyché, J. (2013). Big Data in Big Companies. International Institute For Analytics, 1-31.
Duhigg, C. (2012). How Companies Learn Your Secrets. The New York Times Magazine, 30.
Fabbri, G., Calenne, F., London, M., Boccaletti, C., Cardoso, A., & Mascioli, F. (2012). Development Of An On-Board
Unit For The Monitoring And Management Of An Electric Fleet. 20th International Conference on Electrical
Machines, 2404-2410.
Gandomi, A., & Haider, M. (2015). Beyond The Hype: Big Data Concepts, Methods, And Analytics. International Journal
of Information Management, 137-144.
Goepfert, J., & Shirer, M. (2019, April 04). IDC Forecasts Revenues for Big Data and Business Analytics Solutions Will
Reach $189.1 Billion This Year with Double-Digit Annual Growth Through 2022. Business Wire: A Berkshire
Hathaway Company: https://www.businesswire.com/news/home/20190404005662/en/IDC-Forecasts-Revenues-
for-Big-Data-and-Business-Analytics-Solutions-Will-Reach-189.1-Billion-This-Year-with-Double-Digit-
Annual-Growth-Through-2022
Gogoulos, F., Kaklamani, D., Venieris, I., Anadiotis, A., Inden, U., & Naimark, S. (2013). An Intelligent Trucking
Operations Management System. IFAC Proceedings Volumes, 49-54.
Gonzalez-Feliu, J. (2019). ata Production for Urban Goods Transport Planning and Management: How to Manage “No
Data” Situations in the Big Data Era. USA: IGI Global.
Govindan, K., Cheng, T. C., Mishra, N., & Shukla, N. (2018). Big Data Analytics And Application For Logistics And
Supply Chain Management. Transportation Research Part E Logistics and Transportation Review, 114.
Gowda, V., & Gopalakrishna, K. (2015). Real Time Vehicle Fleet Management And Security System. IEEE Recent
Advances in Intelligent Computational Systems , 417-421.
Grover, V., Chiang, R. H., Liang, T.-P., & Zhang, D. (2018). Creating Strategic Business Value from Big Data Analytics:
A Research Framework. Journal of Management Information Systems.
Gürsakal, N. (2021). Büyük Veri. Bursa: Dora Yayıncılık.
Halper, F., & Krishnan, K. (2013). TDWI Big Data Maturity Model Guide İnterpreting Your Assessment Score. TDWI
Benchmark Guide, 1-20.
Henschen, D. (2014, January 30). 16 Top Big Data Analytics Platforms. Informationweek:
https://www.informationweek.com/big-data/big-data-analytics/16-top-big-data-analytics-platforms/d/d-
id/1113609?image_number=2
Hofmann, E., & Rüsch, M. (2017). Industry 4.0 And The Current Status As Well As Future Prospects On Logistics.
Computers in Industry , 23-34.
Hu, W. C., & Kaabouch, N. (2013). Big Data Management, Technologies, and Applications. IGI Global.
Hurwitz, J., Nugent, A., Halper, D. F., & Kaufman, M. (2013). Big Data For Dummies. USA : For Dummies.
Ilocca, P., Farris, S., Angius, S., Argiolas, R., Obino, A., Secchi, S., & Mozzoni, S. (2017). Managing Data And Rethinking
Applications İn An İnnovativet. Transportation Research Procedia, 1899-191.
International Business Machines Corporation (IBM). (2014). Performance and Capacity Implications for Big Data.
Redpaper, 1-48.
Issa, T. E. (2016, May 16). Data Science: How İs İt Good For Business? Wamda.Com:
https://www.wamda.com/memakersge/2016/05/data-science-how-is-it-good-for-
business?utm_content=buffer6cbca&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer
Ivanov, D., & Dolgui, A. (2020). A Digital Supply Chain Twin For Managing The Disruption Risks And Resilience İn The
Era Of Industry 4.0. Production Planning and Control, 1-15.
Jeske, M., Grüner, M., & Weiß, F. (2013). Bıg Data In Logıstıcs A Dhl Perspective On How To Move Beyond The Hype.
DHL Customer Solutions & Innovation, 16.
Johanson, M., Belenki, S., Jalminger, J., M., F., & Gjertz, M. (2014). Big Automotive Data: Leveraging Large Volumes
Of Data For Knowledge-Driven Product Development. IEEE International Conference on Big Data, 736-741.
Kaisler, S. H., Armour, F., Espinosa, J. A., & Money, W. (2013). Big Data: Issues and Challenges Moving Forward. System
Sciences (HICSS).

292
Kemp, S. (2021). Digital 2021: Global Overview Report. Hootsuite.
Kumar, A., & Jothimani, D. (2017). Big Data: Challenges, Opportunities and Realities. 1-24.
Lam, C. (2010). Hadoop in Action. USA: Manning.
Larson, P., & Halldorsson, A. (2004). Logistics Versus Supply Chain Management: An International Survey. International
Journal of Logistics, 17-31.
Linh, P. N., & Huong, N. T. (2020). The Supply Chain And Logistics Of Vietnam İn The Context Of International
Economic Integration. International Business Research, 27.
Manyika, J., Chui, M., Brown, B., Bughin, J., Dobbs, R., Roxburg, C., & Byers, A. H. (2011). Big Data: The Next Frontier
for Innovation, Competition, and Productivity, , 2011. Mckınsey Global Instıtute, 1-20.
Marketing Türkiye. (2016, January 22). Yemek Sepeti, “Büyük Veri” Kullanımı Ile 4,5 Milyon Tl Değer Kazandı!
Marketing Türkiye Web Sitesi: https://www.marketingturkiye.com.tr/haberler/yemek-sepeti-buyuk-veri-
kullanimi-ile-45-milyon-tl-deger-kazandi/
Mayer-Schönberger, V., & Cukier, K. (2014). Big Data: A Revolution That Will Transform How We Live, Work, and Think.
New York: Eamon Dolan/Mariner Books.
MÜSİAD Müsatakil Sanayici ve İşadamları Derneği. (2020). Dijital Dönüşümün İş Süreçlerine Etkileri. İstanbul:
MÜSİAD Araştırma Raporları.
Ohlhorst, F. J. (2012). Big Data Analytics: Turning Big Data into Big Money. USA: Wiley.
Ostermann, J., & Koetter, F. (2016). Energy-Management-As-A-Service: Mobility Aware Energy Management For A
Shared Electric Vehicle Fleet. 5th International Conference on Smart Cities and Green ICT Systems, 1-11.
PWC. (2016). Shifting Patterns The Future Of The Logistics İndustry. Pwc’S Future İn Sight Series, 1-20.
Rogers, D. S., & Tibben-Lembke, R. (2001). An Examınatıon Of Reverse Logıstıcs Practıces. Journal of Business
Logistics, 129-148.
Satyanarayana, L. V. (2015). A Survey on Challenges and Advantages in Big Data. International Journal of Computer
Science And Technology , 115.
Schwab, P. K., & World Economic Forum. (2018). The Global Competitiveness Report. World Economic Forum, 1-50.
Stancel, I., & Surugiu, M. (2017). Fleet Management System for Truck Platoons . Generating an Optimum Route in Terms
of Fuel Consumption. Procedia Engineering, , 861-867.
Sumathi, N., S, G., Ramana, S. K., Muralitharan, R., & Kamal, C. (2017). Application of Big Data Systems to Airline
Management. International Journal of Latest Technology in Engineering, Management & Applied Science, 129-
132.
Uldrich, J. (2013, March 26). Big Data: The New "Oil" of the Future? Jackuldrıch.Com:
https://jackuldrich.com/blog/business/big-data-the-new-oil-of-the-future/
Walker, S. J. (2015). Big Data: A Revolution That Will Transform How We Live, Work, and Think. International Journal
of Advertising, 181-183.
Wang, M. (2016). The Role Of Innovatıon Capabılıty In The Australıan Courıer Industry. International Journal of
Innovation Management, 1650070-1-18.
Warden, P. (2011). Big Data Glossary. USA: O'Reilly Media, Inc.
Wegner, y. M., & Kückelhaus, D. M. (2013). Logistics Trend Radar. Delivering insight today. Creating value tomorrow! .
Powered by Solutions & Innovation. Trend Research, 1-50.
Williams, S. (2016). Business Intelligence Strategy and Big Data Analytics: A General Management Perspective . USA:
Morgan Kaufmann.
Witkowski, K. (2017). Internet of Things, Big Data, Industry 4.0 – Innovative Solutions in Logistics and Supply Chains
Management. Procedia Engineering, 763-769.
Zhang, G., Wang, J., Yang, Y., Su, H., Yao, Q., Huang, Q., & Huang, J. (2015). A İntegrated Vehicle Health Management
Framework For Aircraft. A Preliminary Report, Ieee Conference On Prognostics And Health Management, 1-8.
Zhong, R. Y., Xu, X., Klotz, E., & Newman, S. T. (2017). Intelligent Manufacturing in the Context of Industry 4.0: A
Review. Engineering, 616-630.
Zikopoulos, P., deRoos, D., Parasuraman, K., Deutsch, T., Giles, J., & Corrigan, D. (2012). Harness the Power of Big Data
The IBM Big Data Platform. USA: McGraw-Hill Education.

293
 CRITERIA PRIORITIZATION FOR ALLOCATION OF ELECTRIC
VEHICLE CHARGING STATIONS BY GROUP DECISION MAKING BASED
INTUITIONISTIC FUZZY AHP APPROACH
Mehmet Hanifi Aydın1, Mücahid Günay2

Abstract ⎯ Electric vehicle is indispensable for reduction of carbon emission caused by traditional gasoline
vehicle. Considering even more comprehensive factors such as environmental sustainability, decreasing fossil
fuel reserves, foreign dependency in energy and carbon emissions, it is expected that electric vehicles will
replace fossil fueled traditional gasoline vehicles in the near future. Optimal site selection of electric vehicle
charging station is thus crucial to fulfill needed demand at different locations for charging electric vehicles to
maintain the balance between demand and supply. In this study, prioritization of established criteria to select
best possible sites for allocation of electric vehicle charging station has been studied by applying a group
decision making based MCDM approach. The proposed method is considered as an intuitionistic fuzzy model
that can easily resolve the uncertainty, therefore, Intuitionistic Fuzzy Sets-based Analytical Hierarchy Process
(AHP) is integrated, due to its effective utility in modeling uncertain or imprecise information. With this study,
Institutions and companies will be capable of seeing which criteria they should focus on primarily and will be
able to plan the necessary actions accordingly. Evaluation criteria is determined by an extensive literature
review and expert opinions. Institutions and companies should take the outcome of the study into account in
order to adopt effective site selection of electric vehicle charging station. Sensitivity analysis and comparison
with available techniques are also presented to confirm the validity of the result obtained under varying
conditions.

Keywords ⎯ Criteria Prioritization, Electric Vehicle, Electric Vehicle Charging Stations, Group Decision
Making, Intuitionistic Fuzzy AHP

INTRODUCTION
Location allocation of electric vehicle charging station (EVCS) is a procedure to decide on the most appropriate
selection with candidates defined for the construction of EVCS. The task of choosing a location for EVCS is
extremely important for harmonious and sustainable development. Because of the swift development of the
economy and urban burst, environmental contamination and energy scarcity have created very serious problems
for Turkey's sustainable development. Vehicles contribute to 15-20% of the total greenhouse gas production in
Turkey and therefore, reducing greenhouse gases is an extremely urgent dilemma for the transportation sector
(Soruşbay, 2007). Electric vehicles (EV), a type of ecologically pleasant transportation, are deemed as an
encouraging solution to energy consumption and carbon emission problems. Electrification of vehicles is a
promising approach to addressing oil and air pollution problems. The construction of charging facilities, as an
energy provider for EV, is gaining huge momentum in terms of ensuring the synergetic development of
technology (Alhazmi, Mostafa, & Salama, 2017). Due to the uncertainty and ambiguity in location allocation,
and multiple-criteria nature, choosing the best location for EVCS can be deemed as a complex Multi-Criteria
Decision Making (MCDM) problem. In this research, an extended method is proposed for optimum location of
EVCS under the Intuitionistic Fuzzy Sets (IFS). This model is easy to implement and is compatible with the
human judgments. By the use of this model, while selecting the places where charging stations should be
installed, prioritization of established criteria to select best possible sites for allocation of electric vehicle
charging station has been studied by applying a Group Decision Making (GDM) based MCDM approach.

In the 1960s, MCDM methods were developed to assist decision-making. MCDM methods be composed of
methodologies and approaches attempting to attain the finest appropriate/ possible solution that meets multiple
conflicting criteria. MCDM approaches applies intricate weight data to unravel difficult challenges with

1
Mehmet Hanifi Aydın, Information Systems MSc, Kahramanmaraş Sütçü İmam University, Kahramanmaraş, Turkey, [email protected]
2
Mücahid Günay, Department of Computer Engineering, Kahramanmaraş Sütçü İmam University, Kahramanmaraş, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

294
inconsistent attributes. Initially, candidates and attributes are identified. Subsequently, the assessment of every
option is achieved according to all criteria. Next, the priorities are designated to all criteria. The aggregated
criteria weights and the values of criteria evaluation are fused to calculate the overall values of the candidates.
Lastly, analyses of sensitivity are applied, and concluding recommendations and assessments are revealed.
MCDM approaches are developed on the choice of the most fulfilling aims/objectives of the Decision Maker
(DM) from a substantial amount of possible choices (Büyüközkan, Göçer, & Feyzioğlu, 2018). Methods such
as Maximin, Maximax, TOPSIS, ELECTRE, Analytic Hierarchy Process, DEMATEL, Analytical Network
Process, PROMETHEE, and VIKOR are various instances of MCDM techniques.

Like many other industries, EV services is currently facing tremendous transformation. Most companies around
the world apply big data analytics to realize their operational strategies and facilitate location selection. In this
study, we will discuss several key areas of location selection for EV charging stations in the context of big data.
In order to choose an optimal location with cost-effectiveness, transportation costs must be reduced and site
selection should be simplified. The location problem arises because of these needs, and we know that the Hub
concept accumulates flows and makes networks more reliable. With this information in mind, we need to
determine where our focus will be. Thus, we have integrated a MCDM tool to prioritize the criteria to select an
optimal site for electric vehicle charging. Here, the decision criteria are determined by expert views and the
proposed integrated MCDM method is considered to be integrated under GDM steps using the Analytical
Hierarchy Process (AHP) to measure criterion weights and evaluate the performance of decision criteria in the
IFS environment. IFSs have shown definite advantages in handling uncertainty and ambiguity over clear, fuzzy,
or type-2 fuzzy sets, which allow for more representative decision-making by depicting experts' assessments
with a richer structure.

The study consists of four chapters. Firstly, the introductory part where the purpose of the study is explained.
In the second part, review of extant literature on the given topic is explored. Third part presents the information
about the methods to be used in the paper. In the fourth part, the proposed approach is utilized with the results
obtained by performing an application. The results are laid out in this section to finalize and interpret the
findings, and final evaluations are made in the conclusion section.

LITERATURE REVIEW
Since it is a very new and developing subject, the number of scientific articles written in this field is very few.
During the literature search, EBSCO, Sciencedirect, SCOPUS and Web of Science databases were searched.
The search was made with the keywords "Electric Vehicle Charging Stations", "Optimal Site Selection", "Fuzzy
Sets", "Multi criteria decision making" and studies on these subjects were examined. In addition, the National
Thesis Center of the Council of Higher Education, Google Academic and Dergipark databases were scanned
with Turkish keywords to examine the studies conducted in Turkey. In recent studies, it is seen that MCDM
methods and mixed approaches have come to the fore (Aktepe & Ersöz, 2014). Table 1 describes the key points
drawn from the researched articles and shows that there is a gap in the literature regarding the study topic we
explore.

Table 1. Relevant Literature

Author(s) Objective Method(s) Type GDM Area
Charging station location selection
AHP,
is investigated by considering Effects of electric
(Demirtaş et al., ArcGIS
ranges of electric vehicles and Empirical - vehicles on the
2021) Pro,
charging capacities in intercity environment.
SWOT
transportation.

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

295
 Author(s) Objective Method(s) Type GDM Area
To create a sustainable evaluation
index for charging station site
Effects and results
evaluation based on a literature
of MCDM
(Feng, Xu, & Li, review using the LEW method LEW and
Empirical - methods on electric
2021) and fuzzy axiomatic design FAD
vehicle charging
(FAD) methods for electric
location selection.
vehicle charging station site
evaluation.
Effects of electric
Positioning the charging stations at
(Nurmuhammed & TOPSIS, vehicle charging
the most suitable point by reducing Review -
Karadağ, 2021) AHP stations on the
their impact on the energy grid.
electricity grid.
To minimize travel costs for Ca, a
coastal city in southern China, by Effects of electric
positioning an electric vehicle K-Means Case buses on factors
(Wu et al., 2021) -
charging station on the bus Clustering Study such as carbon
terminal network at the most cost- emission and cost.
effective possible cost.
Determining the Periodic Technician
Technician Routing and Charging Case Routing and
(Erdem, 2020) CPLEX -
Station Location by Electric Study charging station
Vehicle for Samsun Province. positioning.
Effects and results
To locate electric vehicle charging AHP, of MCDM
(Kaya, Tortum,
stations with the most suitable VIKOR, Case methods on electric
Alemdar, & Çodur, -
range for electric vehicles in PROMET Study vehicle charging
2020)
Istanbul. HEE. station location
selection.
The effects of
elements for
Selecting the most suitable instance service
electric vehicle charging station network reliability,
location by considering elements Dematel, emergency
(Liu, Yang, Zhou,
for instance service network MULTIM Empirical - charging mileage,
& Tian, 2019)
reliability, emergency charging OORA soil locations,
mileage, soil locations, investment cost,
investment cost, and user demand. and user demand on
the charging station
location selection.
The effects of
To select the most suitable
social,
location for electric vehicle
technological and
charging stations for the Nanshan Case
(SUN, 2020) GIS-AHP - environmental
district of Shenzhen City, taking Study
factors on electric
into account social, technological
vehicle charging
and environmental factors.
station positioning.
To ensure that electric vehicles are
at a minimum distance from
Reliability and
charging stations at any time. In
(Arayıcı & Cplex control of the
addition, while providing this, the Empirical -
Poyrazoğlu, 2018) Turkish energy
electricity charging station
system.
installations are planned in such a
way that the lowest cost is created.
Selecting geographic information The effects of
(Karolemeas, system (GIS) ideal charging geographical
Tsigdinos, Tzouras, station locations for electric features on the
GIS-AHP Empirical -
Nikitas, & vehicles across the Netherlands selection of electric
Bakogiannis, 2021) and charging vehicles in 20 vehicle charging
minutes or less. stations.

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

296
 METHODOLOGY
Preliminaries
This subsection presents the basic definitions regarding to the IF set theory.
Fuzzy set theory is represented by A in 𝑋 as:

A = {< 𝑥, 𝜇𝐴 (𝑥) > |𝑥 ∈ 𝑋}, (1)

̃ in 𝑋 as:
And IF set theory is represented by A

̃ = {< 𝑥, 𝜇Ã (𝑥), 𝑣Ã (𝑥) > |𝑥 ∈ 𝑋},

A (2)

Where 𝜇Ã (𝑥): 𝑋 → [0,1] and 𝑣Ã (𝑥): 𝑋 → [0,1] represents the membership and non-membership degrees of 𝑥 ∈
𝑋, respectively. The IF sets should have the following condition:

0 ≤ 𝜇Ã (𝑥), 𝑣Ã (𝑥) ≤ 1, (3)

𝜋Ã (𝑥) = 1 − 𝜇Ã (𝑥) − 𝑣Ã (𝑥). (4)

̃ in X ={x} is fully defined with the form of Equation

𝜋Ã (𝑥) indicates the hesitancy degree. Hence, an IF set A
(Eq.) (2) where

μà (x) → [0,1]; νà (x) → [0,1]; πà (x) → [0,1].

Let à and B
̃ be two IF set in 𝑋 and 𝜆 > 0. Some of the basic arithmetic operations in IF sets are defined as
follows:

̃ B
A ̃ = {[x, μà (x) + μB̃ (x) − μà (x) ∗ μB̃ (x), νà (x) ∗ νB̃ (x)] | x ∈ X}, (5)

̃ B
A ̃ = {[x, μà (x) ∗ μB̃ (x), νà (x) + νB̃ (x) − νà (x) ∗ νB̃ (x)] | x ∈ X}, (6)

̃ = {[x, 1 − (1 − μà (x))𝜆 , νà (x)𝜆 ] | x ∈ X},

𝜆∗A (7)

IF-AHP Methodology
The steps of the proposed methodology are presented in this subsection. The methodology used in prioritizing
the decision criteria for allocation of EVCS is shown in Figure 1.

Step 1: Define hierarchy, criteria and alternatives

The alternative set 𝐴𝑖 (𝑖 = 1, 2, … , 𝑚) is assessed for criteria set 𝐶𝑗 (𝑗 = 1, 2, … , 𝑛).
𝑤𝑗 = {𝑤1 , 𝑤2 , … , 𝑤𝑛 } vector indicates the weight of n criteria, 𝑤𝑗 ≥ 0, 𝑗 = 1, 2, . . . , 𝑛, ve ∑𝑛j=1 𝑤𝑗 = 1.

Step 2: Calculate the weights of each DM

Let the DMs (𝑘 = 1, 2, … , 𝐾) having individual weight 𝜆𝑘 with ∑𝐾
k=1 𝜆𝑘 = 1. By the Eq. (8), DMs’ priorities
are calculated as adapted from (Boran, Genç, Kurt, & Akay, 2009). Linguistic terms are presented in Table 2.

µ 𝑘 ]]
[µ𝑘 +𝜋𝑘 [1−𝜋
𝑘
𝜆𝑘 = µ𝑘
, ∑𝐾
𝑘=1 𝜆𝑘 , (8)
∑𝐾
𝑘=1[µ𝑘 +𝜋𝑘 [1−𝜋 ]]
𝑘

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

297
 Defining the EVCS

Determination of Evaluation Criteria

Determination of Criterion Weights Using the IF-AHP Technique

No Confirm
(C.R. < 0.10)

Yes

Prioritizing Criteria using the IF-AHP Technique

Figure 1. Applied Methodology for Criteria Prioritization for Allocation of EVCS

Table 2. Linguistic terms used in the weight calculation of decision makers (Boran et al., 2009).

Table 3. Linguistic Terms used for Pairwise Comparison (Abdullah & Najib, 2016)

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

298
Step 3. Set Up a decision matrix for all DMs
A scale for the IF-AHP linguistic term is utilized to formulate the pairwise comparison matrix. Table 2 presents
verbal terms in the structure of IF set for pairwise comparison. After the hierarchy is formed, the verbal terms
are converted into an IF set matrix.

Step 4: Create aggregated decision matrix

The individual assessments are fused by the intuitionistic fuzzy weighted averaging (IFWA) aggregation
operator in Eq. (9) to form the GDM matrix. The IFWA operator (Xu, 2011) is applied to combine individual
views into a group view. The specific pairwise matrix of all DMs is first clustered and then fused with the GDM
𝑅𝑖𝑗 . As a result, the 𝑅𝑖𝑗 matrix obtained for each criterion represents the IF AHP weights, and for each
alternative, the matrix obtained on each criterion represents the weights of the alternatives.

(𝑘) 𝜆𝐾 (𝑘) 𝜆𝐾
𝐼𝐹𝑊𝐴(𝐴̃1 , 𝐴̃2 , … , 𝐴̃𝑘 ) = (1 − ∏𝐾
𝑘=1 (1 − µ𝑖𝑗 ) , ∏𝐾
𝑘=1 ((𝜐𝑖𝑗 ) )) (9)

𝑟11 ⋯ 𝑟1𝑛
𝑅𝑖𝑗 = [ ⋮ ⋱ ⋮ ] (10)
𝑟𝑚1 ⋯ 𝑟𝑚𝑛

Step 5: Apply Consistency Check

A reliability test is applied to make sure evaluations of pairwise comparisons are not arbitrary and inconsistent.
Eq. (11) is applied to calculate the Consistency Ratio (C.R.).
The C.R. for the IF AHP weights are estimated using the random index (RI) (Thomas L. Saaty, 1977) value
presented in Table 4.
∑ 𝜋𝑖𝑗 (𝑥)
𝑅𝐼−
𝑛
C.R. = (11)
𝑛−1

n is the quantity of matrix elements and 𝜋𝑖𝑗 (𝑥) is the hesitancy value. C.R is satisfactory if less than or equal to
0.10. Otherwise, the decisions are not consistent and once again evaluation is gathered from the decision makers.

Table 4. Random Index

Step 6: Combine Multilevel Weights

Multiply the IF AHP criterion weights and the IF AHP sub-criterion weights to find the overall criterion weights,
from the lowest level to the highest in the hierarchy. Using the overall criterion weights (𝑤 ̃𝑗 ) and the total weight
of each alternative on the sub-criteria (𝑤̃ 𝑖𝑗 ), total weight of each alternative on the objective (𝑊̃𝑖 ) is calculated
by Eq. (12).

̃𝑖 = ⨁𝑛𝑗=1 (𝑤
𝑊 ̃𝑗 ⨂𝑤
̃𝑖𝑗 ) (12)

Step 7: Prioritization of Criteria

Overall weights listed using the Entropy weights defined in Eq. (13) (Abdullah & Najib, 2016). The smaller the
𝑤
̅ 𝑖 value, the greater the ranking.
1
𝑤
̅ 𝑖 = − 𝑛𝑙𝑛2 [µ𝑗 𝑙𝑛µ𝑗 + 𝑣𝑗 𝑙𝑛𝑣𝑗 − (1 − 𝜋𝑗 ) ln(1 − 𝜋𝑗 ) − 𝜋𝑗 𝑙𝑛2] (13)

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

299
 APPLICATION
Several academics have created tools to identify and handle EVCS evaluation criteria. Most study aims merely
studying some part of the EVCS system, such as location selection, electric consumption, or environmental
issues. Existing studies talk about the importance of these decision criteria in EVCS applications or how to
increase the success in these applications. However, it is seen that almost no research has been done for criteria
prioritization in EVCS. As far as we know, there is no study in the literature on which of these suggested decision
criteria will make EVCS management more effective. The research in this article seeks to cover this gap and
investigate chances for prioritizing EVCS strategies from a sustainability perspective. Accordingly, the main
objectives of this study can be achieved by the prioritization of EVCS Evaluation Criteria as displayed in Table
5.

Table 5. EVCS Evaluation Criteria and Their Brief Descriptions

Main Criteria Sub Criteria Description Type Reference

The Impact on the

EVCS considers vegetation and landscape
Landscape of the Cost
destruction in site selection
City

EVCS considers wastewater discharge and

Environmental
Waste Discharge the impact of battery waste throughout the Cost
Feasibility
use process

Considers decrease of ecological

Greenhouse Gas
contaminants and particulate emission using Benefit
Emission
EV

Considers the simplicity of future expansion

Ease of Expansion Benefit
efforts

Ease of
Considers the simplicity of rebuilding Benefit
Reinstallation

(Demirtaş et al.
Project, equipment, demolition, and land
Economic 2021; Feng et al.
Construction Cost investment costs etc. measures the cost of Cost
Feasibility 2021; Kaya et al.
construction, which includes
2020)
(Demirtaş et al.
Operation and It includes electrical load, worker wages and 2021; Feng et al.
Cost
Maintenance Cost daily machine protection costs etc. gauge 2021; Kaya et al.
2020)
(Demirtaş et al.
2021; Feng et al.
Annual Profit Considers future revenues of EVCS Benefit
2021; Kaya et al.
2020)
Deems the challenge and stability of the
Reliability location chosen for the EVCS against many Benefit
external conditions

Providing
Technological Seizes the opportunity to provide convenient (Erdem, 2020; Wu
convenient services Benefit
Feasibility servicing to the chauffer at EVCS et al., 2021)
to drivers

Safety and ability to

It implies dealing with the emergency and (Erdem, 2020; Wu
deal with an Benefit
maintaining the EVCS in the future et al., 2021)
emergency

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

300
 Main Criteria Sub Criteria Description Type Reference

It convenes the needs of the layout of the city

Ease of access to (Erdem, 2020; Wu
transport network and is appropriate for car Benefit
public transport et al., 2021)
owners to plan their trip

Impact on people's Deems the effect on the existences of the (Erdem, 2020; Wu
Benefit
daily lives people during site selection for EVCS et al., 2021)

Establishment scale
EVCS measures enterprise range and (Erdem, 2020; Wu
and environmental Benefit
dissemination aspects of charging needs et al., 2021)
population density

Deems the real space between two nearby (Erdem, 2020; Wu

Service Radius Benefit
EVCS et al., 2021)

Social Impact on urban (Erdem, 2020; Wu

Considers the city's future planning Benefit
Feasibility growth et al., 2021)

Considers key route, entrance and exit,

City electricity grid (Erdem, 2020; Wu
suburban, metropolitan places and Benefit
planning et al., 2021)
continuous electrical power supply

Proximity within walking distance to the

(Demirtaş et al.
closest school or university, hospital or health
Proximity to Social 2021; Feng et al.
center, public administration building, Benefit
Areas 2021; Kaya et al.
recreation and entertainment center and
2020)
transportation hub/station
(Demirtaş et al.
Proximity to intersections, major roads,
2021; Feng et al.
Easy access substation, gas station and other EV charging Benefit
2021; Kaya et al.
stations
2020)
(Demirtaş et al.
Extent of designated or regulated parking
2021; Feng et al.
Parking Capacity spaces or share of family unit without private Benefit
2021; Kaya et al.
parking
2020)

Numerical Results
MCDM approaches are used to unravel challenges that involve choosing from a limited number of candidates.
MCDM methods list the alternatives and help the decision maker to determine the best alternative. All judgment
tables in MCDM approaches have 4 major components: (1) candidates, (2) attributes, (3) comparative
significance of all attributes, and (4) performance criteria of candidates by characteristics. One of the key
features of EVCS is its uncertainty in various facets. The ambiguities arise during each activity of EVCS. Since
the criteria prioritization problem involves a limited number of candidate alternatives and a large number of
conflicting criteria, MCDM approaches could be applied efficiently to unravel this kind of challenge. The AHP
enables decision makers to structure complex problems in a simple hierarchical manner and to evaluate
numerous aspects in a logical manner. The AHP is a robust and accommodating decision-making procedure that
helps administrators to establish weights and get the finest result while quantitative and qualitative criteria are
taken into account together (T.L. Saaty, 2000).

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

301
 Tablo 6. Individual Effect Level for Each DM
𝑫𝑴 Preferences [ 𝝁(𝒙), 𝒗(𝒙) ] 𝝀𝒌
𝐷𝑀1 - VI I 0.654 0.335 0.392
𝐷𝑀2 EI - I 0.585 0.397 0.262
𝐷𝑀3 I I - 0.500 0.450 0.346

Tablo 7. Individual Evaluations of DMs on Sub-Criteria

𝑫𝑴 𝑪𝟏𝟏 𝑪𝟏𝟐 𝑪𝟏𝟑 𝑪𝟐𝟏 𝑪𝟐𝟐 𝑪𝟐𝟑 𝑪𝟐𝟒 𝑪𝟐𝟓 𝑪𝟑𝟏 . 𝑪𝟒𝟖 𝑪𝟒𝟗
𝐷𝑀1 EP 1/IVIV IVIV IVIV IVI 1/RP RP SP 1/IVIV . SP IVI
𝐶11 𝐷𝑀2 EP 1/ SP IVIV SP IVIII SP IVII IVI 1/IVIV . RP SP
𝐷𝑀3 EP 1/IVIV IVIV IVIV IVI IVIII HP IVII 1/IVIV . IVIV IVII
𝐷𝑀1 . . . . . . . . . . . .
𝐷𝑀2 IVI HP IVI IVIII IVI 1/RP SP IVI IVIII . RP EP
𝐶49 𝐷𝑀3 SP 1/IVIV SP HP 1/HP IVI IVIII 1/HP RP . IVIV EP
𝐷𝑀1 IVI IVII 1/RP 1/IVII 1/IVII SP IVIII SP HP . 1/IVII EP

A group of experts, consisting of three DM, came together to evaluate four main and twenty sub criteria, and
Table 6 presents the linguistic expression of individual weights for each DM and their corresponding values. In
addition to the current literature review, the original evaluation criteria summarized in prior section are created
to evaluate and prioritize the criteria determined by the experiences of DMs. Each DM gives its assessments of
each criterion in the form of linguistic terms. Table 7 shows a small sample of the pairwise comparison matrix
for each DM based on the linguistic scale in Table 3. In order not to surpass the page layout, the values given
are scaled and presented as an example. For criteria evaluations, the linguistic expressions of each DM are
combined by the operator in Equation (9) and used to form the pairwise comparison matrix. Table 8 shows the
intuitive fuzzy values obtained as a result of the combined pairwise comparison matrix and the corresponding
results.

Tablo 8. Criteria Weights of IFS Values and Crisp Weights of Criteria

𝒘
̃𝒋 𝒘
̃ 𝒋𝒋 ̃ 𝒋 𝒘
𝒘 ̃ 𝒋𝒋
Main [𝝁, 𝒗] Sub [𝝁, 𝒗] [𝝁, 𝒗] 𝒘𝒋 Rank
𝑪𝟏𝟏 0.553 0.295 0.320 0.473 0.0486 19
𝑪𝟏 0.578 0.387 𝑪𝟏𝟐 0.269 0.291 0.155 0.471 0.0487 18
𝑪𝟏𝟑 0.601 0.302 0.347 0.477 0.0488 17
𝑪𝟐𝟏 0.638 0.405 0.445 0.517 0.0506 6
𝑪𝟐𝟐 0.569 0.396 0.397 0.511 0.0502 10
𝑪𝟐 0.697 0.352 𝑪𝟐𝟑 0.670 0.451 0.467 0.550 0.0513 2
𝑪𝟐𝟒 0.598 0.395 0.417 0.510 0.0504 8
𝑪𝟐𝟓 0.660 0.445 0.460 0.545 0.0511 3
𝑪𝟑𝟏 0.650 0.440 0.380 0.571 0.0508 5
𝑪𝟑 0.585 0.405 𝑪𝟑𝟐 0.383 0.280 0.224 0.479 0.0483 20
𝑪𝟑𝟑 0.525 0.336 0.307 0.508 0.0493 15
𝑪𝟒𝟏 0.630 0.322 0.331 0.536 0.0500 12
𝑪𝟒𝟐 0.535 0.445 0.281 0.602 0.0509 4
𝑪𝟒𝟑 0.625 0.268 0.328 0.512 0.0499 13
𝑪𝟒𝟒 0.495 0.522 0.260 0.649 0.0515 1
𝑪𝟒 0.525 0.453 𝑪𝟒𝟓 0.543 0.349 0.285 0.550 0.0501 11
𝑪𝟒𝟔 0.558 0.393 0.293 0.573 0.0505 7
𝑪𝟒𝟕 0.649 0.346 0.341 0.548 0.0503 9
𝑪𝟒𝟖 0.588 0.254 0.309 0.507 0.0490 16
𝑪𝟒𝟗 0.642 0.241 0.337 0.501 0.0495 14

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

302
Using Step 5 in the proposed method, the consistency ratio is checked for the combined IFS matrix. Since the
CR values calculated for all matrices are less than or equal to 0.10, the evaluation matrix appears to be consistent
so that the AHP criterion weights shown in Table 8 are established based on all expert opinions. The priorities
of EVCS criteria shown in Table 8 are calculated using the steps of the proposed methodology. As can be seen
from this table, the most prior criterion is the service radius (𝐶44). Providing convenient services to drivers (𝐶32 )
came last in the ranking. This means that among EVCS criteria, the highest priority over others should be given
to establishing a suitable service radius first. In the second place is the construction cost, and in the third place
is the annual profit to provide its sustainability. When the obtained rankings are discussed with DMs, it is
concluded that the results are significant, and it is stated that the results of the proposed method are reliable.

CONCLUSION
Charging station location issues are often resolved with optimization models. If the number of nodes and/or
constraints in the formulation increases, the problem becomes more complex to solve. This complexity led the
researchers to formulate the problem as an MCDM model. The main benefit of MCDM compared to
optimization approaches is that MCDM can accept contradictory and typically subjective parameters in decision
making processes. Finally, the selection methods adopted in the studies in the current literature are not very
suitable for evaluating the location selection problem of EVCS. Therefore, this research contributes to the
literature by presenting the following innovations: first, the presentation of an evaluation basis for PVCS
selection criteria, prioritization of established criteria to select best possible sites for allocation of EVCS, and
another contribution is the application of IFS-based AHP in the evaluation of criterion weights by the use of a
GDM environment. IFSs can model and solve complex problems more appropriately. To the best of the authors'
knowledge, this is the first study to extend the integration of a GDM approach based on the IFS environment,
and the AHP method for prioritization of PVCS evaluation criteria. The last but not least important contribution
of this study is the establishment of evaluation criteria with the feedback of real industry experts. Overall, the
process of decision-making is complicated in nature. Some data in decision process are fuzzy, uncertain, and
undefined. Uncertainty can be triggered by problems with people, machines, or systems. Decent decision
process must tolerate uncertainty. Opinions of people, containing experience and preference, are mostly vague,
and the DMs intuitively anticipate those. Additionally, DMs may or may not be too reluctant to designate
important arithmetic values to evaluate decisions. Fuzzy set is a theory established to integrate uncertainty into
a decision model. Future research topic could be the investigation of a combination of data mining and machine
learning techniques. Appropriate application of big data can bring great benefits to EVCS location selection.
The various characteristics of big data such as reality, variation, velocity, and volume render existing data
analysis methods ineffective.

ACKNOWLEDGMENT
The authors thank all experts for their support and commitment to the completion of the case study.

REFERENCES
Abdullah, L., & Najib, L. (2016). Sustainable energy planning decision using the intuitionistic fuzzy analytic
hierarchy process: choosing energy technology in Malaysia. International Journal of Sustainable Energy,
35(4), 360–377.
Aktepe, A., & Ersöz, S. (2014). AHP-VIKOR VE MOORA Yöntemlerinin Depo Yeri Seçim Probleminde
Uygulanması. Endüstri Mühendisliği Dergisi, 25(1–2), 2–15.
Alhazmi, Y. A., Mostafa, H. A., & Salama, M. M. A. (2017). Optimal allocation for electric vehicle charging
stations using Trip Success Ratio. International Journal of Electrical Power & Energy Systems, 91, 101–
116.
Arayıcı, K. C., & Poyrazoğlu, G. (2018). Optimal Placement of Electric Vehicle Charging Stations. Güc
Sistemleri Konferansı, 81–84.
Boran, F. E., Genç, S., Kurt, M., & Akay, D. (2009). A multi-criteria intuitionistic fuzzy group decision making
for supplier selection with TOPSIS method. Expert Systems with Applications, 36(8), 11363–11368.
Büyüközkan, G., Göçer, F., & Feyzioğlu, O. (2018). Cloud Computing Technology Selection Based on Interval
Valued Intuitionistic Fuzzy COPRAS. In J. Kacprzyk, E. Szmidt, S. Zadrożny, K. T. Atanassov, & M.
Krawczak (Eds.), Advances in Intelligent Systems and Computing (Proceedings of: EUSFLAT- 2017)
(Vol. 641, pp. 318–329).

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

303
Demirtaş, M., Yılmaz, E., Ünal, C., Küçükpehlivan, T., Ağaçsapan, B., & Aksoy, T. (2021). Elektrikli Araçlar
ve Şarj İstasyonlarının Konumlandırılması. GSI Journals Serie C: Advancements in Information Sciences
and Technologies, 4(1), 11–32.
Erdem, M. (2020). Elektrikli Araçla Periyodik Teknisyen Rotalama ve İstasyon Yeri Seçim Problemi. European
Journal of Science and Technology, Special Is, 16–27.
Feng, J., Xu, S. X., & Li, M. (2021). A novel multi-criteria decision-making method for selecting the site of an
electric-vehicle charging station from a sustainable perspective. Sustainable Cities and Society, 65,
102623.
Karolemeas, C., Tsigdinos, S., Tzouras, P. G., Nikitas, A., & Bakogiannis, E. (2021). Determining Electric
Vehicle Charging Station Location Suitability: A Qualitative Study of Greek Stakeholders Employing
Thematic Analysis and Analytical Hierarchy Process. Sustainability, 13(4), 2298.
Kaya, Ö., Tortum, A., Alemdar, K. D., & Çodur, M. Y. (2020). Site selection for EVCS in Istanbul by GIS and
multi-criteria decision-making. Transportation Research Part D: Transport and Environment, 80, 102271.
Liu, H.-C., Yang, M., Zhou, M., & Tian, G. (2019). An Integrated Multi-Criteria Decision Making Approach
to Location Planning of Electric Vehicle Charging Stations. IEEE Transactions on Intelligent
Transportation Systems, 20(1), 362–373.
Nurmuhammed, M., & Karadağ, T. (2021). A Review on Locating the Electric Vehicle Charging Stations and
Their Effect on the Energy Network. Journal of Science Part A: Engineerıng and Innovation, 8(2), 30–
45.
Saaty, T.L. (2000). Fundamentals of Decision Making and Priority Theory with the Analytic Hierarchy Process.
In Analytic Hierarchy Process Series (Vol. 6). Pittsburgh.: RWS Publications.
Saaty, Thomas L. (1977). A scaling method for priorities in hierarchical structures. Journal of Mathematical
Psychology, 15(3), 234–281.
Soruşbay, C. (2007). Karayolu Ulaşımından Kaynaklanan Karbondioksit Emisyonlarının Çevreye Etkisi ve
Kontrolü. Mühendis ve Makina Dergisi, s.564(c:48), 22–26.
SUN, L. (2020). Site selection for EVCSs by GIS-based AHP method. E3S Web of Conferences, 194, 05051.
Wu, X., Feng, Q., Bai, C., Lai, C. S., Jia, Y., & Lai, L. L. (2021). A novel fast-charging stations locational
planning model for electric bus transit system. Energy, 224, 120106.
Xu, Z. (2011). Intuitionistic Preference Modeling and Interactive Decision Making.

© 19th International Logistics and Supply Chain Congress

October 21-21, 2021, Gaziantep, Turkey

304
 EVALUATION OF SUPPLIER SELECTION CRITERIA IN THE MARBLE
SECTOR THROUGH THE DEMATEL METHOD

Tuğrul Bayat1, Önder İnce2, Betül Kocatürk3

Abstract ⎯ Supplier selection is one of the most important steps of supply chain management and also it is a
strategic decision for the business. The ability of businesses to carry out their activities successfully is directly
related to the their suppliers’ efficiency. Properly selected suppliers make a significant contribution to the
competitive capacity of enterprises. Multi-criteria decision-making methods are used extensively in the right
supplier selection. In this study, interviews were conducted with the purchasing managers of five companies
operating in the sector in order to determine the criteria and importance levels in the selection of suppliers in
the marble sector. Marble sector is very important for Afyonkarahisar and makes significant contributions to
the economy of the region and the country. During the interviews, information about the sector was obtained
and the criteria that play a role in supplier selection were evaluated. The analysis of the evaluations was carried
out with the DEMATEL method, which is one of the MCDM methods. According to the results, it was seen that
the criterion with the highest weight value was quality, followed by reliability and delivery speed, respectively.

Keywords ⎯ DEMATEL, Multi Criteria Decision Making (MCDM), Supplier Selection, Marble Sector, Supply
Chain

1 Tuğrul Bayat, Afyon Kocatepe University, Bolvadin Faculty of Applied Sciences, Department of Logistics Management,
Afyonkarahisar, Turkey, [email protected]
2 Önder İnce, Afyon Kocatepe University, Department of Business Administration, 100/2000 YÖK Doctoral Scholarship Student,

[email protected]
3 Betül Kocatürk, Afyon Kocatepe University, Department of Business Administration, 100/2000 YÖK Doctoral Scholarship Student,

[email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
1

305
 INTRODUCTION
According to Christopher (2016) “Supply chains compete, not companies”. Depending on this
ambitious discourse, supplier selection is defined as one of the most important processes in production,
purchasing, marketing and supply management activities (Ghodsypour & Ó Brien, 1998; Kaufmann
et al., 2010) and stands out as an important responsibility of the administration (Golmohammadi &
Mellat-Parast, 2012). Supply chain management is a relatively difficult management task, and the
significant gap between theory and practice leads this process even more difficult (Storey et al., 2006).
Borders are disappearing due to the ever-expanding global markets, and the distances between
production and consumption points are increasing. This makes supplier selection decisions even more
complicated. As a result, it becomes more important to cooperate with the right suppliers. Supplier
selection provides businesses with both cost advantages and the opportunity to increase their
profitability, as it causes a significant amount of revenue expenditure for purchasing in most industries.
Hereby, supplier selection is critical for the competitiveness of organizations (DeBoer et al., 2001).
Supplier selection has positive or negative effects on both business and operating performance of a
firm (Kannan & Tan, 2002; Koufteros et al., 2012; Thornton et al., 2013). With the increasing trend
towards outsourcing, choosing the right suppliers in various industries is becoming more and more
critical to a firm's business success (Wetzstein vd., 2016; Spina vd., 2013).
Due to its geological structure, Afyonkarahisar is an important city with its potential in the marble
sector for Turkey. An important part of the enterprises serving in the marble sector was established in
Iscehisar District of Afyonkarahisar in Turkey. Most of the marble is exported as blocks or processed
by the enterprises. The sector makes a significant contribution to the regional and national economy
(Çakır, 2014).
According to Mineral Research and Exploration (MTA) reports, it is estimated that Turkey has 40%
of the world's marble potential with 5.2 billion m3 (13.9 billion tons) of probable marble reserves.
However, only 1% of this potential is used. Turkey is ranked as the 7th in marble production and 8th
in marble exports in the world, and has some site-specific marble varieties that are acknowledged in
international markets. Approximately 23.6% of the marble quarries operating throughout Turkey are
located in Afyonkarahisar. It is the province with the most intense marble quarries after Balıkesir
(27.00%) (Cetin, 2003: 243-246). In 2020, it is seen that the amount of 6,468,473,922 kg and the
worth of 1,736,574,684 dollars exports were realized considering the natural stone export which is the
product group that includes the export of marble in block or processed form in Turkey (IMIB, 2021).
The supplier selection is very important for the improvement of all processes of marble enterprises.
Because the sector has high production, sales and export rates.
In recent years, supplier selection has become a dynamic subject that has been researched in different
sectors (automotive, food, manufacturing, service-hotel, pharmacy, etc.) with different methods (AHP,
ANP, TODIM, TOPSIS, MOORA, DEMATEL, Fuzzy DEMATEL, etc.) due to its strategic
importance. In the researches, especially in order to select the optimal supplier, multiple criteria are
analytically compared with each other (Davras and Karaatlı, 2014, Şimşek et al., 2015, Dweiri et al.,
2016) or focusing on various mathematical optimization approaches. For example, Özçakar and Demir
(2011) tackle supplier selection in food manufacturing with the TOPSIS method, Simsek et al. (2015)
discuss supplier selection in the service-hotel sector with TOPSIS-MOORA methods, Bai et al. (2019)
tackle the supplier selection in the automotive industry with the TODIM method. Nevertheless, no
studies were found regarding supplier selection in the marble sector through the DEMATEL method.
The contribution of the study to the literature can be summarized in four points;
i) Recognition of the enterprises operating in the marble sector,
ii) Perspectives of marble enterprises on supplier selection,
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey
2

306
 iii) The criteria taken into consideration by the marble enterprises in the selection of suppliers,
iv) Importance levels of criteria used in supplier selection are considered.
In this study, supplier selection, which is a strategic management practice for businesses to reduce
costs, increase financial resources, and gain competitive advantage by providing higher quality and
flexibility, is tackled specific to marble enterprises. In this context, the criteria taken into consideration
in the selection of suppliers and the importance levels of these criteria are tried to be determined for
five companies operating in the marble sector in Afyonkarahisar. The DEMATEL method, which is
one of the MCDM methods, is used in the analysis of the evaluations.
The study consists of five sections. In the first section, the concept of supply chain is explained. The
second section includes the literature on supplier selection. The third section provides information on
the methodology. The fourth section includes the findings and discussion section. The last section
contains conclusions and recommendations.
1.SUPPLY CHAIN AND SUPPLIER SELECTION
According to TDK (2008), the concept of supply is “unraveling, obtaining, providing”. In a broad
terms, supply is a process in which all kinds of activities are included such as the procuring of
consumables, raw materials, materials, semi-finished products, ready parts, etc., necessary for the
production or consumption of the enterprise from suppliers, and agreements, plans and programs,
possible collaborations or mergers made for this. (Nebol et al., 2015: 82).
The supply chain is a whole of businesses that includes all the sub-industrialists, suppliers, producers,
manufacturers, wholesalers, transporters, distributors, warehousing, retail and sales (including after-
sales services) businesses in the chain, including the activities after a product is transformed from the
initial raw material to the final products and delivered to the final consumer as a product or service
(Christopher, 1998; CSCMP, 2009).
Supply chain management (SCM) is the design and management of seamless, value-added processes
across organizational boundaries to meet the actual needs of the end customer (Fawcett et al., 2007:
6). In another definition, SCM covers the strategic and systematic coordination of the functions and
plans of the enterprises aforementioned, including all enterprises in the chain, in order to increase the
long-term performance of the supply chain and all enterprises within this chain (TZYPK, 2009;
Tanyaş, 2002).
SCM is the key to success for any business organization in today's competitive global environment
and is an important component of competitive strategy to increase organizational productivity and
profitability (Punniyamoorthy et al., 2013; Gunasekaran et al., 2004). In this context, SCM focuses on
how effectively they use their suppliers' business processes, talents and technologies in order to provide
high quality materials and components at competitive prices in order to provide competitive advantage
and to what extent they achieve success in the coordination of production, material management and
logistics functions in their enterprises (Farley, 1997; Davis et al., 1999: 382; Tan et al., 2002: 615). A
successful SCM is based on developing strategic partnerships with commercial partners in the chain
(Başyazıcı, 2010: 57; Şahin, 2003: 19).
Businesses have started to search for ways to improve their activities as a member of their supply chain
since the 1990s (Stadler and Kilger, 2000: 8). The most important reason for this is the decrease in the
ability of enterprises to compete independently, and in the past the competition between enterprises is
replaced by the competition between supply chains nowadays (Ungan, 2011: 318).
Supplier is used to mean “anybody who provides the necessary material” (TDK, 2008). Supplier
selection is a strategic decision that affects all processes for businesses, and is the result of a complex

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
3

307
process. Evaluating alternatives on the basis of various criteria and choosing the best supplier provides
businesses with a significant gain and competitive advantage in the short and long term (Dağdeviren
and Eraslan, 2008: 69). When this issue is evaluating in terms of today's aspect, the fact that the
competition is not between companies but between supply chains makes the supplier selection more
important (Christopher, 2016). Because supply chain management is relevant to the improvement of
all processes from the point of production to the point of consumption. Supplier selection, which is an
important part of supply chain management, also has a serious impact on all processes and the
importance of a strong integration within the chain is increasing day after day (Ungan, 2011: 318).
According to Eleren (2008), the issue of which of the many suppliers to work with is a complex multi-
criteria decision-making (MCDM) problem. At this stage, first of all, the criteria to be used in supplier
selection should be determined. In this direction, in order to determine the criteria, the priorities of the
sector, customer demands, competitive conditions and the strategies applied in supply chain
management are determined and the criteria are weighted according to the level of importance. The
criteria used commonly can be considered as in Table 1 (Eleren, 2008: 36):

Table 1. Criteria Used in Supplier Selection

Quality Performance Cost/Price Advantage
Delivery Performance Knowledge and Technology
Level
Participation Performance Experience and References
Integration Performance Reliability
Capacity Level Willingness
Financial Condition Flexibility
Scores from Past Quality Tests Quality Certificates Owned
Number of Complaints About Products Proximity to Business
Continuity and Consistency Corporateness

Geographical Location

Reference: Eleren (2008: 36)

The general purpose of the supplier selection process is to reduce the risk of purchasing, maximizing
the total value for the buyer, and to establish closeness and long-term relationships between buyers
and suppliers (Chen et al., 2006). In this context, choosing the right supplier can minimize procurement
cost, increase procurement quality and reliability, and thus increase the company's profit margin by
minimizing the risk of upstream supply chain (Mukherjee, 2017: 1). In addition, wrong supplier
selection can cause great losses in operational or financial aspects (Güner and Mutlu, 2005: 473).
Choosing the right supplier makes it possible for businesses to make profits by reducing their costs,
while not compromising on quality and customer satisfaction. Therefore, the primary objective to
consider each criterion that improve all processes together and to provide all of them at the required
rate.
2.LITERATURE REVIEW
The first study in which supplier selection was discussed and given a point of view was made by
Dickson in 1966. In his study, Dickson evaluated 23 criteria together, including quality, cost, delivery
performance and flexibility as the main criteria (Özdemir, 2010: 62). Studies on supplier selection have
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey
4

308
concentrated after the 1990s. It is seen that MCDM methods are used extensively in determining the
criteria that are effective in supplier selection and in revealing the relations between them. Information
on the studies conducted on supplier selection is summarized in Table 2.

Table 2. Criteria Used in Supplier Selection in the Literature

Geographical
Competence
Technology

Production
Price- Cost

Flexibility

Condition
Productivity

Technical
Financial

Capacity
Location
Delivery

Reliability
Quality

Service
Study Sector Model
Dickson
(1966) X X X X X
Dağdeviren &
Eren (2001) AHP X X X X
Güner (2005) Marble AHP X X X X X X X X X
Demir (2010) Food TOPSIS X X X X X X
Özçakar &
Demir (2011) Food TOPSIS X X X X X X
Chang et al. Fuzzy
(2011) Electronic DEMATEL X X X X X X X X
Antalya-
Akyüz (2012) Furniture VIKOR X X X X X X X
Rouyendegh
& Erkan
(2012) University AHP X X X X
Davras &
Karaatlı
(2014), Hotel AHP-BASP X X X X X X
Rouyendegh
& Saputro Fertilizer Fuzzy
(2014) Production TOPSIS X X X X X X X X
Kar & Pani India- 8
(2014) sector AHP X X X X X X
Şimşek et al. TOPSIS &
(2015) Hotel MOORA X X X X
BAAS &
Özbek (2016) VIKOR X X X X X X X X X X
Saracoğlu &
Dağistanlı Fasteners Fuzzy-AHP
(2017) Production ve VIKOR X X X X X X
Sureeyatanap
as et al. Egg
(2018) Production TOPSIS X X X X X
Leblebicioğlu
& Keskin Fuzzy
(2021) Pharmacy DEMATEL X X X X X X X X
(Note: The yellow-filled expressions in the table are the criteria with the highest priority in the related studies)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
5

309
In consequence of the literature review, it is seen that the most used criteria in supplier selection are
"Quality, Price / Cost, Delivery, Service, Technology / Technical Capability, Reliability, Geographical
Location, Flexibility and Financial Condition".
3.METHODOLOGY

In this section, the data and the method to be used in the study will be demonstrated. In this context,
the implementation steps of the method will be explained and thereafter the analysis phase will be
started.

3.1.Method
MCDM methods are used in problems that have more than one criterion and need to be decided.
Decision-making problems according to the policy of the decision-maker, and the overall purpose of
the decision, may require the choice of an alternative solution, the ranking of the alternatives from the
best to the worst ones (Zopounidis, 2002: 227).
The subject of supplier selection is generally discussed in the literature in a way that aims to obtain
results based on the opinions of one or more experts (Supçiller and Deligöz, 2018). The data obtained
at this point are subjective data. There are MCDM methods that evaluate a large number of subjective
data, and the DEMATEL method is one of the most important of these methods (Wu and Lee, 2007).
The DEMATEL method operates on of the basis of reciprocal comparison criteria. The DEMATEL
method is used in order to weight the criteria that are effective in supplier selection by comparing them
with each other, to determine the order of importance and to determine the affecting and affected
criteria (Chang et al. 2011).
The DEMATEL (Decision Making Trial and Evaluation Laboratory) method was first introduced by
Gabus and Fontela in 1972 (Gabus and Fontella, 1972: 3). The method is an effective method that
analyzes the cause and effect relationships between factors in complex problems by means of matrices
(Si et al., 2018: 20; Bakır et al. 2018: 326). Due to the DEMATEL method, the strength and level of
the relationships between the criteria can be determined. In this method, as in some other MCDM
methods, the evaluation of the criteria is based on expert opinions and experiences. In the DEMATEL
method, the complex relationships between the criteria are visualized in the relationship diagram
(Tseng, 2008). The relationship diagram allows decision makers to visually see the direction and
intensity of the relationships between the criteria, allowing for easier interpretation of complex
relationships between criteria (Ecer, 2020). In this way, decision makers can analyze how they can
develop criteria in the affected criteria group to make changes to other criteria. In addition, the method
determines criteria that have a high impact on the solution, allowing decision makers to focus on fewer
criteria (Chang et al., 2011; Lin, 2013; Tsai et al., 2015; Baykaşoğlu et al. 2013).
DEMATEL method, which is one of the multi-criteria decision making (MCDM) methods, is basically
used to determine the importance levels by weighting the criteria. The DEMATEL method performs
the analysis of subjective data that consists of the opinions of experts on the issue (Hsu et al., 2013).
The DEMATEL method, which is based on the graph theory, offers the opportunity to solve the
problems in which the relevant criteria are included in order to determine the relationships formed
within the scope of causality (Aksakal and Dağdeviren, 2010).
When the literature is examined, it is seen that many studies have been carried out on the DEMATEL
method and the priority order of the relevant criteria is determined by this method. (Fontela and Gabus,
1976; Wu and Lee, 2007; Hu et al. 2009; Chang et al. 2011; Ho et al. 2012). In addition, depending on
the studies conducted in the literature, since it is expected that the DEMATEL method will yield
effective results in supplier selection (Chang et al. 2011), this method was used in the study.
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey
6

310
Some studies describe the operation of the DEMATEL method in four steps (Fontela & Gabus; 1976)
and some studies describe it in seven steps in order to be more detailed (Ecer, 2020). For a better
understanding of the model, the method will be explained in seven steps. The relevant steps are listed
below (Ecer, 2020):
Step 1. Pairwise Comparison of Criteria: In the first step, the scale in Table 3 determines to what extent
the criteria have affected each other. Comparisons are made with values between 0 and 4. At this point,
the comparison is made by experts on the subject (Ecer, 2020).

Table 3: DEMATEL Method Comparison Values

Değer Tanım
0 Ineffective
1 Low Impact
2 Moderate Impact
3 High Impact
4 Very High Impact

Step 2. Creating the Direct Relationship Matrix: After the comparison of the criteria within themselves,
a direct relationship matrix is obtained based on this comparison. In this matrix, the values of the
elements in the first diagonal are always 0. Because any criterion has no effect on itself (Ecer, 2020).
𝑥11 𝑥12 … 𝑥1𝑛
𝑥21 𝑥22 … 𝑥2𝑛
. . . .
𝑥= . . . .
. . . .
{𝑥𝑛1 𝑥𝑛2 … 𝑥𝑛𝑛 }
Step 3. Obtaining the Standard Direct Relationship Matrix: The sums of all rows and columns of the
created direct relationship matrix are obtained and the highest sum is selected as a result. Afterwards,
all elements in the direct relationship matrix are proportioned to this maximum value and normalization
process is carried out (Ecer, 2020).
𝑥
(Equation 1) 𝑋̃ = 𝑚𝑎𝑘𝑠(𝑚𝑎𝑘𝑠 ∑𝑛 𝑛
𝑗=1 𝑥𝑖𝑗 ,∑𝑖=1 𝑥𝑖𝑗 )

Step 4. Finding the Total Relationship Matrix: In this step, firstly X ̃ is subtracted from the unit matrix
and the obtained matrix is inversed, and finally X ̃ is multiplied by the obtained result (Ecer, 2020).
(Equation 2) 𝑇 = 𝑋̃. (1 − 𝑋̃)−1
Step 5. Determination of D, R, D+R and D-R Values: While the R value is obtained by adding the
columns in the total relationship matrix, the D value is obtained as a result of adding the rows of the
same matrix. At this point, R means that a criterion is affected by a different criterion, while D means
that a criterion affects a different criterion, on the contrary. However, while D+R expresses the degree
of importance of any criterion, D-R determines the relations of the criteria. If a criterion affects other
criteria (D-R) it must have a positive result and the effect value is indicated as “1”, if a criterion is
affected by other criteria (D-R) it must have a negative result and the effect value is indicated as “0”
at this point (Ecer, 2020).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
7

311
Step 6. Determination of Threshold Value: In order to make the right decisions and perform a large
number of analyzes, a threshold value suitable for the analysis is needed. Threshold value can be
formed by finding the opinions of experts, the opinions of researchers or the average of the values in
the total relationship matrix (Ecer, 2020).
Step 7. Building Criterion Weights: In this last step of the method, the weights of the criteria can be
determined with Equation 3 and Equation 4 below, and in this way, the order of importance of the
criteria can be constituted (Ecer, 2020).
(Equation 3) 𝑤𝑖𝑎 = √(𝐷𝑖 + 𝑅𝑖 )2 + (𝐷𝑖 − 𝑅𝑖 )2
𝑊𝑖𝑎
(Equation 4) 𝑊𝑖 = ∑𝑛
𝑖=1 𝑊𝑖𝑎

3.2. Method Application

This research is carried out in the focus of the raw material supplier selection problem of the enterprises
operating in the marble sector in Afyonkarahisar. In this context, an evaluation is made by considering
the opinions of 5 experts in the purchasing department of marble enterprises. The criteria related to the
supplier selection problem are determined primarily as a result of literature review and then interviews
with experts in the marble industry. The criteria determined for supplier selection are as follows;
Quality, Price, Flexibility, Delivery Speed, Service, Performance Record, Financial Condition,
Geographical Location, Reliability and Communication criteria. The main purpose of this research is
to determine the weights of the criteria in supplier selection and to determine which of the supplier
selection criteria are influencing and which are affected. For this purpose, the weight values of all
criteria are calculated with the DEMATEL method. In addition, the importance level of all criteria and
their impact on other criteria and their effects from other criteria are determined by the DEMATEL
method. In line with these findings, it will be determined which supplier selection criteria should be
focused more in the supply of raw materials for the marble sector.

Table 4: Supplier Selection Criteria

K1 Quality
K2 Price
K3 Flexibility
K4 Delivery Speed
K5 Service
K6 Performance Record
K7 Financial Condition
K8 Geographical Location
K9 Reliability
K10 Communication

Step 1: Pairwise comparison of the criteria was carried out by giving a score between 0 and 4 by
experts, according to the degree of influence of the criteria with each other in Table 4.
Step 2: The direct relationship matrix was created in line with the expert opinions in the purchasing
department of the companies operating in the marble sector. The arithmetic average of the values
assigned to the criteria by the experts who made the evaluation in accordance with the DEMATEL
method. The scores resulting from the arithmetic mean of these values are shown in Table 5.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
8

312
 Table 5: Direct Relationship Matrix
K1 K2 K3 K4 K5 K6 K7 K8 K9 K10 Total
K1 0 3 3 4 4 4 3 3 3 3 30
K2 3 0 2 2 3 2 3 2 2 3 22
K3 2 3 0 2 2 1 2 3 2 2 19
K4 2 2 3 0 2 2 3 3 2 3 22
K5 1 2 2 2 0 3 2 3 2 2 19
K6 3 2 3 3 3 0 2 2 1 2 21
K7 3 1 2 3 2 2 0 1 2 2 18
K8 2 3 2 2 2 2 2 0 2 2 19
K9 2 3 3 3 3 3 3 4 0 3 27
K10 1 2 1 2 2 2 2 2 2 0 16
Total 19 21 21 23 23 21 22 23 18 22

Step 3: In this step, the normalization of the standard direct relationship matrix is performed. In
accordance with the DEMATEL method, the sums of all columns and rows must be calculated
separately in order to find the X ̃ coefficient. After the sums are obtained, the maximum reference sum
is determined and all elements in the direct relationship matrix are proportioned to the determined
maximum reference sum, and the normalization process is realized. In summary, the analysis is carried
out with Equation 1 given while explaining the method. The maximum reference value in this study
was determined as 30. After the relevant transactions were carried out, the direct relationship matrix
in Table 6 is created.

Table 6: Normalized Direct Relationship Matrix

K1 K2 K3 K4 K5 K6 K7 K8 K9 K10
K1 0,00 0,10 0,10 0,13 0,13 0,13 0,10 0,10 0,10 0,10
K2 0,10 0,00 0,07 0,07 0,10 0,07 0,10 0,07 0,07 0,10
K3 0,07 0,10 0,00 0,07 0,07 0,03 0,07 0,10 0,07 0,07
K4 0,07 0,07 0,10 0,00 0,07 0,07 0,10 0,10 0,07 0,10
K5 0,03 0,07 0,07 0,07 0,00 0,10 0,07 0,10 0,07 0,07
K6 0,10 0,07 0,10 0,10 0,10 0,00 0,07 0,07 0,03 0,07
K7 0,10 0,03 0,07 0,10 0,07 0,07 0,00 0,03 0,07 0,07
K8 0,07 0,10 0,07 0,07 0,07 0,07 0,07 0,00 0,07 0,07
K9 0,07 0,10 0,10 0,10 0,10 0,10 0,10 0,13 0,00 0,10
K10 0,03 0,07 0,03 0,07 0,07 0,07 0,07 0,07 0,07 0,00

Step 4: In this step, the total relationship matrix is obtained. The solution is provided with Equation 2
given while explaining the method of the total relationship matrix. The total relationship matrix formed
as a result of the analysis made with Equation 2 is given in Table 7.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
9

313
 Table 7: Total Relationship Matrix

K1 K2 K3 K4 K5 K6 K7 K8 K9 K10
0,2152 0,3259 0,3276 0,3749 0,3746 0,3548 0,3372 0,3477 0,2950 0,3372
K1
0,2524 0,1759 0,2385 0,2557 0,2838 0,2409 0,2764 0,2541 0,2173 0,2764
K2
0,2037 0,2470 0,1534 0,2291 0,2298 0,1874 0,2247 0,2594 0,1978 0,2247
K3
0,2218 0,2369 0,2644 0,1888 0,2501 0,2349 0,2730 0,2799 0,2144 0,2730
K4
0,1732 0,2146 0,2145 0,2266 0,1639 0,2418 0,2206 0,2561 0,1932 0,2206
K5
0,2458 0,2323 0,2616 0,2757 0,2754 0,1695 0,2398 0,2475 0,1817 0,2398
K6
0,2277 0,1829 0,2140 0,2562 0,2257 0,2137 0,1575 0,1979 0,1935 0,2200
K7
0,2050 0,2465 0,2172 0,2305 0,2312 0,2171 0,2252 0,1681 0,1973 0,2252
K8
0,2561 0,3028 0,3018 0,3193 0,3190 0,3007 0,3116 0,3488 0,1833 0,3116
K9

0,1528 0,1919 0,1635 0,2037 0,2034 0,1927 0,1990 0,2037 0,1751 0,1365
K10

Step 5: In this step, the sum of the column (R) and row (D) values in the total relationship matrix is
calculated. At the same time, the importance level (D+R) of the criteria and the relations of the criteria
(D-R) are determined. At this point, R means that a criterion is affected by a different criterion, while
D means that a criterion affects a different criterion, on the contrary. If a criterion affects other criteria
(D-R) it must have a positive result and the effect value is indicated as “1”, if a criterion is affected by
other criteria (D-R) it must have a negative result and the effect value is indicated as “0” at this point.
As seen in Table 8, while K1, K2, K6 and K9 are the criteria affecting other criteria, K3, K4, K5, K7,
K8 and K10 are the criteria affected by other criteria.

Table 8: Determination of D, R, D+R, D-R and Impact Group

D R D-R D+R Impact Group

K1 (Quality) 3,290026 2,153528 1,136497 5,443554 1
K2 (Price) 2,471439 2,356768 0,114671 4,828207 1
K3 (Flexibility) 2,15704 2,356591 -0,19955 4,513632 0
K4 (Delivery Speed)
2,437233 2,560436 -0,1232 4,99767 0
K5 (Service) 2,125069 2,557055 -0,43199 4,682124 0
K6 (Performance
Record) 2,369188 2,35362 0,015568 4,722807 1

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
10

314
 K7 (Financial
Condition) 2,089123 2,46497 -0,37585 4,554093 0
K8 (Geographical
Location)
2,163469 2,563115 -0,39965 4,726584 0
K9 (Reliability) 2,954933 2,048734 0,9062 5,003667 1
K10
(Communication) 1,822266 2,46497 -0,6427 4,287236 0

Step 6: In this step, the threshold value is determined. In this study, the threshold value was calculated
by taking the average of the values in the total relationship matrix. As a result of the relevant
calculation, the threshold value was determined as “0.2387”.
Step 7: In this last step, the weights of the criteria are calculated. The weight calculations for each
criterion are carried out with Equation 3 and Equation 4, which were given before the DEMATEL
method was explained. Finally, the priority order is determined according to the weight values of the
criteria. The criterion with the highest weight value is the criterion with the highest level of importance.
Table 9 shows the weights and priority orders of the criteria obtained as a result of the study.

Table 9: Values of Criterion Weights and Order of Priority

Criteria Weight (𝑾) Priority of Criteria
Ranking
K1 (Quality) 0,115694 1
K9 (Reliability) 0,105794 2
K4 (Delivery Speed) 0,104007 3
K2 (Price) 0,100478 4
K8 (Geographical Location) 0,098687 5
K6 (Performance Record) 0,098258 6
K5 (Service) 0,097824 7
K7 (Financial Condition) 0,095069 8
K3 (Flexibility) 0,093997 9
K10 (Communication) 0,090192 10
Total 1

According to the findings obtained with the DEMATEL method, the K1 “Quality” (0,1156) criterion
was the criterion with the highest weight value. It was observed that the second and third criteria with
the highest weight value were K9 “Reliability” (0.1057) and K4 “Delivery Speed” (0.1040) criteria.
The weight values and order of importance of the other criteria used in the study are as follows; K2
“Price” (0.1004) criteria, K8 “Geographical Location” (0.0986) criteria, K6 “Performance Record”
(0.0982) criteria, K5 “Service” (0.0978) criteria, K7 “Financial Condition” (0.0950) criteria, K3
(0.0939) “Flexibility” criteria and finally K10 “Communication” (0.0901) criteria.
4. FINDINGS AND DISCUSSION
In the study, it is seen that the quality criterion has the highest degree of weight in the selection of
suppliers in the marble sector (This finding overlaps with the studies of Rouyendegh and Erkan, 2012;
Davras and Karaatlı, 2014; Rouyendehg and Saputro, 2014; Kar and Pani, 2014; Özbek, 2016). This
result shows that the most considered criterion in the supplier selection of the marble industry is
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey
11

315
"quality". The main reason for this is marble companies work on job shop production, in other words
they produce on an customer order basis. Accordingly, enterprises supply raw materials according to
the desired quality in the customers' orders. In addition, reliability has the second highest degree of
weight in supplier selection. This is thought to be that enterprises want to work with suppliers that can
fulfill their promises.
On the other hand, it has been determined that the K10 “Communication” criterion, which has the
lowest weight value among the criteria, has less importance than the other nine criteria. It is thought
that the reason why the communication criterion has the lowest level of importance is due to the
practices such as some companies in the sector keeping long-term stocks and providing transportation
itself by choosing the product on site.
When the affecting and affected criteria are examined, it is seen that “Quality” (K1), “Price” (K2),
“Performance Record” (K6) and “Reliability” (K9) criteria are among the influencing criteria group.
Besides that, “Flexibility” (K3), “Delivery Speed” (K4), “Service” (K5), “Financial Condition” (K7),
“Geographical Location” (K8) and “Communication” (K10) criteria are among the affected criteria
group. The criteria that the company should focus on in supplier selection are in the affecting group.
In addition, F1 (one of the interviewed companies) stated that they have been working with almost
three years with existing stock, so the speed of supply is not very important to them. Another company,
F2, stated that in every raw material supply, the goods are seen on-site and they make the supply
themselves. For this reason, he stated that the speed of supply is not very important for them. It was
stated that the reason for choosing the goods by seeing the goods on site every time was the frequent
occurrence of workmanship differences even at the same point. Another firm, F3, stated that
compliance with procedures and ship exit is very important in the export process, so delivery speed is
a criterion that cannot be ignored. However, it is emphasized that the importance of geographical
location is moderate, since some types of marble are quarried in special regions.
As a result of the interviews, it is understood that the sector generally works on job shop production.
One of the companies F4 stated that customer satisfaction is the most important issue for them and that
achieving this satisfaction comes first. Besides that F4 stated that they supply raw materials from
various countries such as Ukraine and Greece, and that factors such as price and geographical location
remain in the background in the procurement of durable and visually pleasing stones. Another
company F5 emphasizes that they work on job shop production, that the raw material that needs to be
supplied in a certain type and quality has no alternative in some cases, and that all conditions must be
accepted from a certain point.
The results obtained from the study are similar to the studies in the literature (Rouyendegh and Erkan,
2012; Davras and Karaatlı, 2014; Rouyendehg and Saputro, 2014; Kar and Pani, 2014; Özbek, 2016).
5. CONCLUSION
In order to achieve their strategically important goals, companies must gain an advantage over their
competitors in highly competitive market conditions and at the same time must provide customer
satisfaction. For these reasons, especially in the manufacturing sector, the raw material that makes up
the final product should be obtained from the most optimal supplier in line with the interests of the
companies. In this context, the supplier selection problem is a very critical and important issue for
companies (Sureeyatanapas et al. 2018).
As in many industries, performance of the supply chain is also of vital importance in terms of
competitive advantage and sustainability in the marble sector and supplier selection is one of the main
factors determining this performance. Supplier selection can vary greatly from industry to industry and
from business to business. In addition, there are many different criteria to choose the most suitable
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey
12

316
supplier for the business, customer and market. MCDM methods are preferred because of allowing the
solution of problems in which a large number of contradictory and content integrity criteria have inside.
The interviews provide the opportunity to get to know the marble businesses better. In this context, it
has been observed that marble enterprises work on job-shop production (customer-oriented), keep
long-term stocks, and purchase the raw materials on site and transport them themselves. Hereby, it is
seen that the criteria they consider in the selection of suppliers are also affected by these operation
principles and methods. The criteria discussed in the study were formed by bringing them together in
the light of literature and interviews. The findings obtained from the study are summarized in Table
10.

Table 10: Values of Criterion Weights and Order of Priority

Criteria Impact Weight (𝑾) Priority Order
Group of Criteria
K1 (Quality) Affecting 0,115694 1
K9 (Reliability) Affecting 0,105794 2
K4 (Delivery Speed) Affected 0,104007 3
K2 (Price) Affecting 0,100478 4
K8 (Geographical Affected
0,098687 5
Location)
K6 (Performance Affecting
0,098258 6
Record)
K5 (Service) Affected 0,097824 7
K7 (Financial Condition) Affected 0,095069 8
K3 (Flexibility) Affected 0,093997 9
K10 (Communication) Affected 0,090192 10
Total 1

As a result of the analysis, it has been determined that "Quality" is the most important criterion in the
selection of raw material supplier in the marble sector. The main reason for this is thought to be due to
the demand-oriented service provided by the companies in the marble sector. Some companies have
stated that they can receive products of different quality at times, even from the suppliers they work
with, so they feel the need to choose products on-site of supplier. It seems that the importance of the
quality criterion causes some criteria (such as service, communication, delivery speed) to be less
important. Also, it is seen that the criterion of "quality" is widely used in the literature and is included
in studies as the most effective criterion (Rouyendegh and Erkan, 2012; Davras and Karaatlı, 2014;
Rouyendehg and Saputro, 2014; Kar and Pani, 2014; Özbek, 2016). However, it was determined that
"Reliability" was the second most important criterion. Companies in the marble sector always expect
trust from their raw material suppliers and find it risky to do business with a supplier that does not
provide trust. In the analysis, it was determined that the criterion with the lowest priority value among
the criteria was the "Communication" criterion. The main reason for this is to communicate with
suppliers when needed, rather than constant communication, due to the different demands of
customers.
The other result is the criteria that have an impact on other criteria in the selection of raw material
suppliers in the marble sector. These criteria are the criteria that are taken into consideration when
evaluating other criteria and affect the evaluation of the relevant criteria. In the study, it is seen that
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey
13

317
the criteria of "quality, reliability, price and performance record" are the criteria to be considered when
evaluating the "delivery speed, geographical location, service, financial condition, flexibility and
communication" criteria. For this reason, businesses that want to be a supplier in the marble industry
should pay more attention to these criteria. Improvements in these “affecteing criteria” will also
positively affect other criteria.
When the results are examined in general, it is seen that the issues that marble enterprises pay attention
to in the supplier selection process mostly focus on the product and the concern of trust in the supplier.
In future research, DEMATEL method can be applied in supplier selection in different sectors. It is
also suggested that comparative studies can be conducted with other MCDM methods.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
14

318
 REFERENCES
Aksakal E.ve Dağdeviren, M. (2010). ANP ve DEMATEL Yöntemleri İle Personel Seçimi Problemine Bütünleşik Bir
Yaklaşım. Journal of Gazi University Faculty of Engineering and Architecture, Volume 25, No 4, 905-910.
Akyüz, G. A. (2012). Bulanık VIKOR Yöntemi ile Tedarikçi Seçimi. Ataturk University Journal of Economics and
Administrative Sciences, 26(1), 197-215.
Bai, C., Kusi-Sarpong, S., Badri Ahmadi, H., & Sarkis, J. (2019). Social Sustainable Supplier Evaluation and Selection:
A Group Decision-Support Approach. International Journal of Production Research, 57(22), 7046-7067.
Bakir, S., Khan, S., Ahsan, K. & Rahman, S. (2018). Exploring the Critical Determinants of Environmentally Oriented
Public Procurement Using the DEMATEL Method. Journal of Environmental Management, 225, 325-335.
Başyazıcı, B. B. (2010). Stok Yönetimi ve Stratejik Satın Alma, Niğde University SSI, Department of Busıness.
Accounting-Finance Department.
Baykasoğlu, A., Kaplanoğlu, V., Durmuşoğlu, Z., Şahin, C. (2013). Integrating Fuzzy DEMATEL And Fuzzy
Hierarchical TOPSIS Methods For Truck Selection. Expert Systems With Applications, Volume 40, Issue 3, 15, 899-907.
Chang, B., Chang, C.W. & Wu, C.H. (2011). Fuzzy DEMATEL Method for Developing Supplier Selection Criteria.
Expert Systems with Applications, 38(3), 1850-1858.
Chen, C. T., Lin, C. T., & Huang, S. F. (2006). A fuzzy approach for supplier evaluation and selection in supply chain
management. International journal of production economics, 102(2), 289-301.
Christopher, M. (1998). Logistics & Supply Chain Management: Strategies for Reducing Cost and Improving Service,
Financial Times Professional Limited, London.
Christopher, M. (2016). Logistics and Supply Chain Management. Fifth Ed. FT Press; Pearson Education Limited.
Çakır, M. (2014). İscehisar İlçesinde Mermer Sanayisi ve Planlama Önerileri (Master's thesis, Afyon Kocatepe
University, Institute of Social Sciences).
Çetin, T. (2003). Türkiye Mermer Potansiyeli, Üretimi ve İhracatı. Gazi Journal of Gazi University Faculty of
Education, 23(3).
Çiçek, E., & Bay, M. (2007). Stratejik Küresel Tedarik Zinciri Yönetimi Ve Lojistik. Journal of Social Economic
Research, 7 (13), 91-117. Retrieved from https://dergipark.org.tr/tr/pub/susead/issue/28428/302795
Dağdeviren, M. ve Tamer, E. (2001). Tedarikçi Firma Seçiminde Analitik Hiyerarşi Prosesi ve 0-1 Hedef Programlama
Yöntemlerinin Kullanılması. Journal of Gazi University Faculty of Engineering and Architecture, 16(1), 41-52.
Dağdeviren, M. ve Erarslan, E. (2008). PROMETHEE Sıralama Yöntemi ile Tedarikçi Seçimi. Journal of Gazi
University Faculty of Engineering and Architecture, 23(1).
Davis, M. M., Aquiliano, N. J. & Chase, R. B. (1999). Fundamentals of Operations Management, Irwin McGraw-Hill
Inc.
Davras, G. M., Karaatlı, M. (2014). Otel İşletmelerinde Tedarikçi Seçimi Sürecinde AHP ve BAHP Yöntemlerinin
Uygulanması. Journal of Hacettepe University Faculty of Economics and Administrative Sciences, 32(1), 87-112.
DeBoer, L., Labro, E. & Morlacchi, P. (2001). A Review of Methods Supporting Supplier Selection. European Journal
of Purchasing & Supply Management, Volume 7, pp. 75-89.
Deepen, J.M. (2007). Logistics Outsourcing Relationships: Measurements, Antecedents, and Effects of Logistics
Outsourcing Performance, Physica-Verlag, Heidelberg.
Demir H. H. (2010). İmalat Sektöründe Bulanık TOPSIS Yöntemiyle Tedarikçi Seçimi. Istanbul University, Institute
of Social Sciences, Department of Business Administration, Department of Production, Published Master's Thesis.
Dickson, Gary. (1966). An Analysis Of Vendor Selection Systems And Decisions. Journal of Purchasing. 2. 28-41.
10.1111/j.1745-493X.1966.tb00818.x.
Dweiri, F., Kumar, S., Khan, S. A., & Jain, V. (2016). Designing an Integrated AHP Based Decision Support System
for Supplier Selection in Automotive Industry. Expert Systems with Applications, 62, 273-283.
Ecer F. (2020). Çok Kriterli Karar Verme (Geçmişten Günümüze Kapsamlı Bir Yaklaşım). 1. Baskı, S. 85-
93.Seçkin Publishing: Ankara.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
15

319
 Eleren, A. (2008). Tedarik Zinciri ve Lojistik Yönetimi. Ankara: Gazi Bookstore
Engelbrecht, C. (2004). Logistikoptimierung durch Outsourcing: Erfolgswirkungen und Erfolgsfaktoren. Wiesbaden.
Farley, G. A. (1997). Discovering Supply Chain Management: A Roundtable Discussion, APICS- The Performance
Advantage, Vol. 7 No.1, pp. 38-39.
Fawcett, S.E., Ellram, L.M., Ogden, J.A. (2007). Supply Chain Management from Vision to Implementation. Pearson
Education, New Jersey.
Fontela, E., & Gabus, A. (1976). The DEMATEL Observer. DEMATEL 1976 Report. Geneva, Switzerland: Battelle
Geneva Research Center.
Gabus, A., & Fontela, E. (1972). World Problems, An Invitation to Further Thought Within the Framework of
DEMATEL. Battelle Geneva Research Center, Geneva, Switzerland, 1-8.
Ghodsypour, S. & O´Brien, C. (1998). A Decision Support System for Supplier Selection Using An Integrated Analytic
Hierarchy Process and Linear Programming. Int. J. Production Economics, Volume 56-57, pp. 199-212.
Golmohammadi, D. & Mellat-Parast, M. (2012). Developing A Grey-Based Decisionmaking Model for Supplier
Selection. Int. Journal of Production Economics, Issue 137, pp. 191-200.
Gunasekaran, A., Patel, C., & McGaughey, R. E. (2004). A Framework for Supply Chain Performance Measurement.
International Journal of Production Economics, Vol. 87 No. 3, pp. 333-47.
Güçlü, A. (2013). Dış Kaynak Kullanımı (Outsourcing) ve Vergilendirilmesi. Ankara: Seçkin Publishing.
Güner, H. (2005). Bulanık AHP ve Bir İşletme İçin Tedarikçi Seçimi Problemine Uygulanması. (Unpublished Master
Thesis). Pamukkale University, Denizli.
Güner, H., & Mutlu, Ö. (2005). Bulanık AHP İle Tedarikçi Seçim Problemi ve Bir Uygulama. V. National Production
Research Symposium, Istanbul Commerce University, 25-27 November.
Ho, L.H., Feng, S. Y., Lee, Y. C., Yen T. M., (2012). Using Modified IPA To Evaluate Supplier’s Performance:
Multiple Regression Analysis and DEMATEL Approach. Expert Systems With Applications, 39, 7102–7109.
Horng, J. S., Liu, C. H., Chou, S.F., Tsai, C.Y. (2012). Creativity As A Critical Criterion For Future Restaurant Space
Design: Developing A Novel Model With DEMATEL Application. International Journal of Hospitality Management, 31
July 2012.
Hsu, C. W., Kuo, T. C., Chen, S. H., & Hu, A. H. (2013). Using DEMATEL to Develop A Carbon Management Model
of Supplier Selection in Green Supply Chain Management. Journal of Cleaner Production, 56, 164-172.
Hu, H. Y., Lee, Y.C., Yen, T.M., Tsai, C. H. (2009). Using BPNN And DEMATEL To Modify Importance–
Performance Analysis Model – A Study of The Computer Industry. Expert Systems With Applications, 36, 9969–9979.
Hu, H.Y., Chiu, S., Cheng, C.C., Yen, T.M. (2011). Applying The IPA And DEMATEL Models To Improve The
Order-Winner Criteria: A Case Study Of Taiwan’s Network Communication Equipment Manufacturing Industry.
Expert Systems with Applications, 38(8), 9674-9683.
IMIB (Istanbul Mineral and Metals Exporters' Association), (2021)
https://www.imib.org.tr/links/istatistikler/202012.html
Kar A. ve Pani A. (2014). Exploring the Importance of Different Supplier Selection Criteria. Management Research
Review Vol. 37 No. 1, 89-105.
Kannan, V. & Tan, K., (2002). Supplier Selection and Assessment: Their Impact on Business Performance. The Journal
of Supply Chain Management: A Global Review of Purchasing and Supply, Volume 38(3), pp. 472-489.
Kaufmann, L., Carter, C. & Buhrmann, C. (2010). Debiasing the Supplier Selection Decision: A Taxonomy and
Conceptualization. Int. Journal of Physical Distribution and Logistics Management, Volume 40(10), pp. 792-821.
Kaufmann, L., Kreft, S., Ehrgott, M. & Reimann, F. (2012). Rationality in Supplier Selection Decisions: The Effect
of the Buyer’s National Task Environment. Journal of Purchasing & Supply Management, Issue 18, pp. 76-91.
Koufteros, X., Vickery, S.K. & Dröge, C. (2012). The Effects of Strategic Supplier Selection on Buyer Competitive
Performance in Matched Domains--Does Supplier Integration Mediate Partnerships. Journal of Supply Chain Management,
Volume 48(2), pp. 93-115.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
16

320
 Leblebicioğlu, B. & Keskin, A. (2021). Evaluation of Supplier Selection Criteria with Fuzzy DEMATEL Method: An
Application on the Pharmacy Industry . Avrupa Bilim ve Teknoloji Dergisi, (23), 236-242. DOI: 10.31590/ejosat.864116
Li, C.W. ve Tzeng, G.H. (2009). Identification of A Threshold Value for The DEMATEL Method Using The
Maximum Mean De-Entropy Algorithm to Find Critical Services Provided By A Semiconductor Intellectual Property Mall.
Expert Systems with Applications, 36: 9891–9898.
Lin RJ. (2013). Using fuzzy DEMATEL to Evaluate the Green Supply Chain Management Practices. Journal of
Cleaner Production, 40, 32-39.
Nebol E., Uzel E., & Uslu T. (2015). Tedarik Zinciri ve Lojistik Yönetimi, İstanbul, Beta Publisher.
Özbek A. (2016). Tedarikçi Seçiminde Çok Kriterli Karar Verme Yöntemlerinin Kullanılması. Anadolu University
Journal of Social Sciences, Vol. 16, No. 1, 85 – 101.
Özçakar, N. ve Demir, H. (2011). Bulanik Topsis Yöntemiyle Tedarikçi Seçimi. Istanbul University, Faculty of
Business Administration, Institute of Business Economics, Journal of Management, 22(69), 25-44.
Punniyamoorthy, M., Thamaraiselvan, N. & Manikandan, L. (2013). Assessment of Supply Chain Risk: Scale
Development and Validation. Benchmarking: An International Journal, Vol. 20 No. 1, pp. 79-105.
https://doi.org/10.1108/14635771311299506
Rouyendegh D. ve Erkan T. (2012). Selecting The Best Supplier Using Analytic Hierarchy Process (AHP) Method.
African Journal of Business Management Vol.6 (4),1455-1462.
Rouyendegh B. D., ve Saputro T. E. (2014). Supplier Selection Using Integrated Fuzzy TOPSIS and MCGP Case
Study. 5th World Conference on Educational Sciences WCES 2013. Procedia-Social and Behavioral Sciences 116, 3957-
3970.
Saracoglu, Ilkay & Dağıstanlı, Hakan. (2017). Tedarikçi Seçiminde Bulanık Mantık-AHP ve VIKOR Yönteminin
Bağlantı Elemanları Firmasında Uygulanması. Journal of Yaşar University. 40-54. 10.19168/jyasar.377457.
Sen, S., Basligil, H., Sen, C. & Baracli, H. (2008). A Framework for Defining Both Qualitative and Quantitative
Supplier Selection Criteria Considering the Buyer–Supplier Integration Strategies. Int. Journal of Production Research,
Volume 46(7), pp. 1825-1845.
Shieh, J., Wu, H.H., Huang, K.K. (2010). A DEMATEL Method In Identifying Key Success Factors Of Hospital
Service Quality. Knowledge-Based Systems, Volume 23, Issue 3, April 2010, 277-282.
Si, S.L., You, X.Y., Liu, H.C., & Zhang, P. (2018). DEMATEL Technique: A Systematic Review of The State-of-the-
Art Literature on Methodologies and Applications. Mathematical Problems in Engineering, 2018, 1-33.v
Spina, G., Caniato, F., Luzzini, D. & Ronchi, S. (2013). Past, Present and Future Trends of Purchasing and Supply
Management: An Extensive Literature Review. Industrial Marketing Management, Issue 42, pp. 1202-1212.
Stadler H., & Kilger C. (2000). Supply Chain Management and Advanced Planning, Springer.
Storey, J., Emberson, C., Godsell, J. and Harrison, A. (2006). Supply Chain Management: Theory, Practice and Future
Challenges. International Journal of Operations & Production Management, Vol. 26 No. 7, pp. 754-774.
https://doi.org/10.1108/01443570610672220
Sumrit, D., Anuntavoranich, P. (2013). Using DEMATEL Method to Analyze the Casual Relations on Technological
Innovation Capability Evaluation Factors in Thai Technology-Based Firms”, International Transaction Journal of
Engineering, Management and Applied Sciences and Technologies, 4(2), 81-103.
Supçiller A. A. ve Deligöz K. (2018). Tedarikçi Seçimi Probleminin Çok Kriterli Karar Verme Yöntemleriyle Uzlaşık
Çözümü. International Journal of Economic and Administrative Studies. UİİİD-IJEAS, 355-368.
Sureeyatanapas P., Sriwattananusart K., Niyamosoth T., Sessomboon W., ve Arunyanart S. (2018). Supplier Selection
Towards Uncertaion and Unavailable Infırmation: An Extension of TOPSIS Method. Operations Research Perspective 5,
69-79.
Şahin, S. A. (2003). Satınalma Ve Risk Yönetiminin Tedarik Zinciri Yönetimindeki Yeri. Master Thesis, Istanbul
Technical University, Institute of Science and Technology, İstanbul.
Şimşek, A., Çatır, O., Ömürbek, N. (2015). TOPSIS ve MOORA Yöntemleri ile Tedarikçi Seçimi: Turizm Sektöründe
Bir Uygulama. Balıkesir University Journal of Social Sciences Institute, 18(33), 133-161.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
17

321
 Tan, K.C., Lyman, S.B., & Wisner, J. D. (2002). Supply Chain Management: A Strategic Perspective, International
Journal of Operations and Production Management, Vol: 22, No:6, pp.614-631.
Tanrıverdi, Y. (2010). Tedarik Zinciri ve Stok Yönetimi Üzerine Bir Uygulama. Master Thesis, Pamukkale University
Institute of Social Sciences, Denizli
Tanyaş, M. (2002). Lojistik ve Tedarik Zinciri Yönetimi, 3gen Dergisi, Sonbahar 2002:23-28.
The Council of Supply Chain Management Professionals (CSCMP)- Tedarik Zinciri Yönetim Profesyonelleri Konseyi
(TZYPK) (2009), https://cscmp.org/CSCMP/Educate/SCM_Definitions_and_Glossary_of_Terms.aspx Erişim Tarihi:
10.01.2021
Thornton, L., Autry, C., Gligor, D. & Brik, A. (2013). Does Socially Responsible Supplier Selection Pay Off for
Customer Firms? A Cross-Cultural Comparison. Journal of Supply Chain Management, Volume 49(3), pp. 66-89.
Tsai, S.B., Chien, M.F., Xue, Y., Li, L., Jiang, X., Chen, Q., & Wang, L. (2015). Using the fuzzy DEMATEL to
Determine Environmental Performance: A Case of Printed Circuit Board Industry In Taiwan. PloS one, 10(6), 1-18.
Tseng M.L., Lin Y. H. (2008). Application of Fuzzy DEMATEL to Develop A Cause and Effect Model of Municipal
Solid Waste. Environ Monit Assess, DOI 10.1007/s10661-008-0601-2.
Ungan, M. C. (2011). En İyi Tedarik Zinciri Uygulamaları ve Bir Saha Çalışması. Journal of the Faculty of Economics
and Administrative Sciences of Süleyman Demirel University, 16(2), 307-322.
Veni K. K., Rajesh R. ve Pugazhendhi S. (2012). International Conference on Modeling Optimization and Computing.
Development of Decision Making Model Using Integrated AHP and DEA for Vendor Selection. Procedia Engineering 38,
3700-3708.
Wang, R., Lin, Y. H., Tseng, M. L. (2011). Evaluation Of Customer Perceptions On Airline Service Quality In
Uncertainty. Procedia-Social and Behavioral Sciences, 25, 419-437.
Wetzstein, Anton & Hartmann, Evi & Benton, Wc & Hohenstein, Nils-Ole. (2016). A Systematic Assessment of
Supplier Selection Literature – State-of-the-Art and Future Scope. International Journal of Production Economics. 182.
http://dx.doi.org/10.1016/j.ijpe.2016.06.022
Wu, W.W., Lee Y.T. (2007). Developing Global Managers’ Competencies Using The Fuzzy DEMATEL Method.
Expert Systems With Applications, 32, 499–507
Yang J. L., Tzeng G.H. (2011). An Integrated MCDM Technique Combined With DEMATEL For A Novel Cluster-
Weighted With ANP Method. ,Expert Systems With Applications, Volume 38, Issue 3, March 2011, Pages 1417-1424.
Zopounidis, C. (2002). MCDA Methodologies for Classification and Sorting, European Journal of Operational
Research, 138, 227– 228
https://www.imib.org.tr/links/istatistikler/202012.htm

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
18

322
 A TIME-VARIANT CAPACITATED FACILITY LOCATION AND
NETWORK DESIGN PROBLEM FOR WASTEWATER TREATMENT
SYSTEMS

Seyda Alperen Pehlivan1, Duygun Fatih Demirel2, Sıdıka Melek Başak3

Abstract ⎯ With the increasing consumption and disposal of water resources, developing effective wastewater
treatment solutions continues to be vital to ensure environmental sustainability. It is essential to establish a
water treatment system or a water treatment network in any region, especially if there is no sufficient clean
water source. While establishing a water distribution network, details such as the distribution route and the
amount of water are crucial. This study is about the optimization of water recycling and sustainability where
there are not enough water resources and external supply is limited. Unlike the classical Facility Location
and Network Design models that addresses the problem on a static and time-invariant basis which are not
realistic and practical enough, a time-variant mixed-integer model is proposed that considers facility and link
capacities as well. We also proposed how the model can be extended with pressure and energy concerns and
formulations.

Keywords ⎯ Mixed-Integer Programming, Network Design, Optimization, Wastewater Treatment Network

INTRODUCTION
Water is vital for the survival of living organisms. From the smallest living organism to the largest living
being, it is water that sustains all biological life and all human activities. Moreover, water is essential for
natural systems, wildlife and environmental resilience and is placed at the center of sustainable development.
With the increasing consumption and disposal of water resources due to population growth and industrial
facilities, water scarcity is becoming an increasingly prominent and widespread problem worldwide. Besides,
water quality deteriorates rapidly. In the last decade, many research has been made about the threat of water
scarcity and very serious consequences are foreseen in the next 50 years (Boretti & Rosa, 2019; Hoekstra &
Wiedmann, 2014; Huynen et al., 2013; Kummu et al., 2016). Recently, World Economic Forum (2020)
determined the world’s top ten global risks and three of them were related to the water and environmental
sustainability.
Despite being a country covered with seas on three sides and plenty of rivers, water shortage is foreseen in
Turkey in the upcoming years. In particular, studies have been carried out recently to solve the water problem
in metropolises and large cities in the country (Bobat, 2019; A. M. Saatci, 2013; E. Saatci & Akpinar, 2007;
van Leeuwen & Sjerps, 2016). According to the water and climate change adaptation report published by
OECD, water quality and water quantity are among the primary concerns of Turkey (OECD, 2013).
With the increasing threat of water scarcity, developing effective wastewater treatment solutions continues to
be vital to ensuring environmental sustainability. The water distributed to consumers must meet certain quality
standards and quantities. If the water distributed to consumers is not pure enough for human consumption, the
content and components of this water may negatively affect human health. That is why water treatment is
essential.
According to 2018 data published by the Turkish Statistical Institute (TÜİK, 2018), 755 of 1399
municipalities in Turkey do not have any wastewater treatment plants. 21% of the total population live in
regions where there are no wastewater treatment plants. Moreover, 10% of the total population does not live in
municipalities that serve with sewerage network. Only 61% of the wastewater generated as a result of

1
Seyda Alperen Pehlivan, Yeditepe University, Faculty of Engineering, Department of Industrial Engineering, İstanbul, Turkey,
[email protected]
2
Duygun Fatih Demirel, İstanbul Kültür University, Faculty of Engineering, Department of Industrial Engineering, İstanbul, Tur key
[email protected]
3
Sıdıka Melek Başak, Yeditepe University, Faculty of Engineering, Department of Industrial Engineering, İstanbul, Turkey,
[email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

323
industrial and domestic water use is treated in water treatment plants. Besides all, water treatment plants in
Turkey have an average efficiency of 66.5%.
It is crucial to establish an efficient water treatment system or a water treatment network in any region,
especially if there is no sufficient clean water source. While establishing a water distribution network, details
such as the distribution route of the water and the amount of this water are crucial. This study is about the
optimization of water recycling and sustainability with limited resources in regions where there are not
enough water resources and external supply is limited. For this purpose, a dynamic capacitated facility
location and network design model has been proposed, which ensures the demand while improving the
efficiency of facilities under budget and time constraints.
In the remainder of this paper, studies about facility location and network design problems in the literature
examined first. Next, the mathematical formulation of the proposed problem is presented. The mathematical
model is applied to a designed case and computational results with randomly generated data are tested and
reported. Then, it is proposed how the model can be extended with pressure and energy calculations. Finally,
conclusions about this study are discussed.

LITERATURE REVIEW
In general, facility location problems are related to decisions about where facilities are best located and how to
allocate customer’s to facilities to meet the customers' demand for certain services or products. These
problems are used in many different services such as health, education, fire, production and storage, as well as
in supply chain methods such as water distribution networks. On the other hand, network design problems
deal with the making of network links and decisions to determine the flow of traffic on those links. All
decisions mentioned are made under the condition that the demands are met, taking into account the costs of
opening or establishing a link. The facility location-network design problem is a more complex problem
achieved by the integration of the two mentioned problems above.
The study of the facility location problem dates back to the early 20th century. Weber (1909) proposed the
first facility location formulation by taking into account the travel cost. Since then, studies in this field have
diversified and many papers have been published that proposed different mathematical formulations with
different objectives (Church & ReVelle, 1974; Elloumi et al., 2004; Hakimi, 1964; Klose & Drexl, 2005;
Kuehn & Hamburger, 1963). Recently, various review papers related to facility location problem have been
published (Farahani et al., 2012; Melo et al., 2009; Puerto et al., 2018; Şahin & Süral, 2007; Terouhid et al.,
2012).
Network design problems started to be studied more recently than the facility location problems. To the best
of our knowledge, Leblanc (1975), McGregor and Shen (1977) were the first ones to study this specific
problem. Magnanti and Wong (1984) proposed a solution technique for this problem, detailing the
decomposition techniques and algorithms comparatively. Yang and Bell (1998) discussed the solution
methods and presented a review. Farahani et al. (2013) proposed one of the most detailed review about
transportation network design.
The integration of these two models was studied by Melkote and Daskin (2001). They proposed an
uncapacitated facility location-network design problem. In this problem, there are nodes which are
representing each customer and candidate facility locations; while edges represent possible links between
nodes. Then, some heuristic methods have been developed and applied to solve this problem (Cocking et al.,
2009; Pearce & Forbes, 2018; Ravi & Sinha, 2006).
All of the above-mentioned papers addressed the facility location-network design problem on a static and
time-invariant basis. This situation is actually not very realistic and has little practical use. The demand and
the cost are expected to change over time. Therefore, it is more practical to handle this problem as a dynamic
or time-variant basis. Ghaderi and Jabalameli (2013) proposed a budget-constrained dynamic uncapacitated
facility location and network design problem, and applied for healthcare systems in Illam Province of Iran.
The main contribution of this study is to propose a novel dynamic and capacitated facility location and
network design problem in order to develop an effective wastewater treatment network. Unlike the most of the
studies in the literature, time varying behavior of demand and cost parameters is not ignored and both facility
and link capacities are included as a limitation. Besides, the capacities of the facilities are not fixed and
designed in a varying way with respect to the facility levels.

METHODOLOGY

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

324
The dynamic capacitated facility location and network design problem (DCFLNDP) is defined on a network
of nodes and a time horizon which involves set of time periods. The main assumptions used in the problem
can be given as: (1) each node is a demand point and a candidate location for hosting a facility, (2) facilities
are divided into levels with respect to their capacities and there can only be one facility in a node throughout
the time horizon, regardless of its level, (3) once construction is completed, facilities and links will be
operational until the end of the time horizon, (4) the links are capacitated like facilities and they can serve for
a limited amount of demand, (5) opening facilities and links are not instantaneous, construction times are
included, (6) the network is a customer-to-server system, as described by Melkote and Daskin (2001), (7)
there is a distance limit while constructing a link which means if the distance between two arbitrary nodes is
above this limit, a link between these nodes cannot be constructed, (8) links are directed and can only be
constructed in one direction.
DCFLNDP model determines the facilities and network links that are to be built at the candidate locations(that
satisfies the assumptions) throughout the time horizon. Moreover, it also decides on the fraction of demand of
the nodes served by a facility or a link. The notations used in this problem are given in Table 1.

Table 1. Notations used in DCFLNDP

Notation Description
Sets
N Set of nodes, i,j ∈N and set of demand points k ∈N
U Set of possible undirected links, (i,j) ∈ U :
L Set of facility levels, l ∈ L
T Set of time periods, t ∈ T
Parameters
Demand related to demand point k at time period t
Distance between node i and node j
Cost of constructing a facility with level l on node i at time period t
Cost of constructing a link (i,j) at time period t
Cost of travelling per demand unit on link (i,j) at time period t
Operating and maintenance cost of a facility with level l on node i at time period t
Operating and maintenance cost of a link (i,j) at time period t
Construction time for a facility with level l
Construction time for a link
Capacity of a facility with level l
CL Capacity of a link
Distance limit for constructing a link
Decision Variables
1 if a facility with level l is located on node i at time period t, 0 otherwise
1 if a facility with level l is operational on node i at time period t, 0 otherwise
1 if a link (i,j) is constructed at time period t, 0 otherwise
1 if a link (i,j) is operational at time period t, 0 otherwise
Fraction of the demand related to demand point k that flows on link (i,j) at time period t
Fraction of the demand related to demand point k served by a facility on node i at time
period t

Unlike the other classical facility location problems, in our context a commodity k is used to define specify
demand points in order to control flow traffic throughout the time horizon. Thus, the model decides not only
the facilities to be constructed or the links to be established, but also the facilities and the links where the
demand related to each demand point will be served each time period.
After defining the assumptions and notations, the MILP formulation of DCFLNDP may then be stated as
follows:

min. ∑∑∑ ∑∑∑ ∑ ∑ ∑ ∑ ∑ ∑∑ (1)

∈ ∈ ∈ ∈ ∈ ∈ ∈ ∈ ∈ ∈ ∈ ∈ ∈

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

325
s.t.; ∑ , ∈ ∈ ∈ (2)

, ∈ ∈ ∈ (3)

∑ , ∈ ∈ (4)

, ∈ ∈ (5)

∑ ∈ ∑ ∈ , ∈ ∈ (6)

∑ ∈ , ∈ ∈ (7)

∑ ∈ ∑∈ , ∈ ∈ (8)

∑ ∈ , ∈ ∈ (9)

∑∈ , ∈ ∈ (10)

, ∈ ∈ (11)

∑∈ , ∈ ∈ (12)

∈{ } ∈ (13)

The objective function (1) minimizes the sum of the construction cost and the operating cost of the facilities
and links as well as the travelling cost over a time horizon. Since the parameters associated with cost and
demand change over time, operating costs and constructing costs are unique for each time period.
Equation (2) ensures that a facility with level l on node i will be operational τl periods after the initial
investment decision is made. Facilities are not active and cannot serve any demand during their construction
period. Constraint (3) ensures that once a facility with level l on node i opened at time period t, it will remain
opened until the end of the time horizon as stated in Assumption 3.
Equation (4) ensures that a link (i,j) will be operational periods after the initial construction decision.
Constraint (5) ensures that once a link (i,j) is constructed at time period t, it will remain opened until the end
of the time horizon.
Equations (6)-(7) are the flow conservation equations that are split up in the Melkote and Daskin’s
formulation. Equation (6) ensures that each unit entering node i will either served by node i to ensure its
demand or transferred to other nodes at time period t. Equation (7) ensures that if there is a demand on node i,
it must be served by either from a facility on node i or by another facility from the other nodes in the network.
Constraints (8)-(9) are the capacity constraints. Constraint (8) ensures that the demand served from a facility
with level l on node i cannot exceed its capacity during each time period. It also states that a facility on node i
must be operational at time period t to serve any demand in that period. Constraint (9) ensures that in order to
use a link (i,j) to serve the demand of demand point k at time period t, the link must be operational in that
period and the total demand flows on that link cannot exceed its capacity.
Equation (10) ensures that each demand must be served for each demand point throughout the time horizon.
Constraint (11) precludes oppositely directed flows on a link which means flow on any constructed link will in
one direction only. Constraint (12) ensures that there can only be one facility in a node. Equation (13)
corresponds to the integrality and nonnegativity constraints.

EXPERIMENTAL RESULTS
In this section, computational results are presented for an experiment that is designed by using different
number of nodes ( ), levels, periods and for two different values of distance limit in order to evaluate the
effect of the proposed model. All test problems are solved with Python modeling library for optimization
DOcplex and CPLEX version of 20.1. Experiments are run under Windows with 2.8GHz processor and 16 GB
of RAM. The details of the experimental design with the specific number of different experiments are given in
Table 2.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

326
The fundamental point to note is that, indicates the total number of time periods in which the demand is
taken into account. As mentioned earlier, there is a construction period for both the facilities and the links.
Considering that the model ensures all the demand is met within the time horizon and there is no existing
network initially, an external time interval is assumed for the initial investments. The length of this time
interval is determined as the maximum construction time.

Table 2. Details of the Experimental Design

Experimental Design
E1 E2 E3 E4
E5 E6 E7 E8
E9 E10 E11 E12
E13 E14 E15 E16
E17 E18 E19 E20
E21 E22 E23 E24
E25 E26 E27 E28
E29 E30 E31 E32
E33 E34 E35 E36

Thus, in the remainder of the paper; t = 0* indicates the starting time of the time horizon (before activation)
and t = 1 indicates the starting time of the demand. The time difference between t = 0* and t = 1 is the
maximum construction time of any link or facility. The parameters used in the experimental design are
presented in Table 3.

Table 3. Parameters used in the Random Generation of Experimental Design

Parameter Value

0*

0*

0*

CL

The locations of the nodes are generated randomly over a 50 50 unit area in each test instance by using
Python library NumPy. As demonstrated in Table 3, the construction costs and operating costs associated with

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

327
the proposed model are generated using uniform random variables. Table 4 summarizes the results of the
thirty six test instances by presenting the objective value, the lower bound, the percentage of optimality gap,
CPU time of the specific instance and the average efficiency of the facilities. A time limit of 10000 seconds is
specified for the algorithm. If the time limit is reached, the algorithm stops working and presented the recent
results. Figure 1. illustrates the details of Experiment 19 as an example. In Fig.1-a), the locations of generated
nodes and their initial demands at t = 1 are visualized. In Fig.1-b), the state of the system at t = 20 (last
period) on the time horizon is summarized. The red dots represent the facilities that are constructed, while the
green lines represent the opened links within the time horizon.
Table 4. Computational Results of the Experimental Design
Experiment Objective Lower Gap(%) CPU Avg.
Value Bound Time(sec.) Efficiency(%)
E1 6277.15 6277.15 0.00 0.25 93.71
E2 5346.21 5346.21 0.00 0.49 91.45
E3 5585.84 5585.84 0.00 0.66 92.31
E4 6301.25 6301.25 0.00 1.72 93.66
E5 14473.10 14473.10 0.00 4.06 99.11
E6 12557.81 12557.81 0.00 1.89 99.06
E7 11868.58 11868.58 0.00 6.28 98.29
E8 12031.47 12031.47 0.00 5.38 93.14
E9 28493.48 28493.48 0.00 11.19 92.12
E10 30404.43 30404.43 0.00 46.97 96.78
E11 37544.39 37544.39 0.00 2.18 83.42
E12 18611.61 18611.61 0.00 19.69 98.25
E13 10284.80 10284.80 0.00 43.88 98.12
E14 13686.00 13686.00 0.00 34.07 96.53
E15 9976.98 9976.98 0.00 54.11 99.00
E16 12884.71 12884.71 0.00 89.27 97.85
E17 31675.61 31675.61 0.00 39.54 90.44
E18 28077.98 28077.98 0.00 1741.34 99.76
E19 16492.83 16492.83 0.00 117.66 94.49
E20 26992.94 26992.94 0.00 948.04 99.01
E21 61517.14 61517.14 0.00 4834.20 99.41
E22 45806.88 45806.88 0.00 1838.12 98.27
E23 39548.96 39548.96 0.00 1704.36 97.74
E24 44959.53 44734.73 0.50 10000 99.51
E25 13724.60 13724.60 0.00 202.11 99.35
E26 18975.98 18975.98 0.00 9183.86 99.31
E27 14062.76 14062.76 0.00 731.80 99.24
E28 15519.62 15519.62 0.00 1394.94 98.47
E29 40693.82 40199.68 1.21 10000 97.93
E30 39912.95 39424.48 1.22 10000 96.52
E31 29837.11 29837.11 0.00 7714.65 99.77
E32 34589.12 33803.19 2.23 10000 97.95
E33 72808.79 72808.79 0.00 8217.91 98.11
E34 50484.76 49874.09 1.20 10000 99.68
E35 65451.97 64030.47 2.17 10000 97.90
E36 68423.82 67056.78 1.99 10000 98.49

Table 5. presents the opened links and facilities and their construction period. Four facilities are constructed
and ten links are opened within the time horizon.

Table 5. Constructed Facilities and Opened Links in Experiment 19

Constructed Facilities Opened Links
Node Level Time Period Link Time Period
6 5 1 (1,6) 4
8 5 1 (1,8) 1
9 5 1 (2,6) 1
10 4 5 (2,9) 1

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

328
 (3,9) 1
(4,8) 1
(5,4) 1
(7,6) 1
(7,10) 4
(10,6) 1

a) Generated nodes and Initial Demands b) The Situation at t = 20

Fig. 1. The Visualization of Experiment 19

In Table 6., the decision variable which gives the fraction of the demand served, is presented in order to
better insight the working principle of the proposed model. To give an example, after the construction of a
facility on node 10, the demand of this node is met by a facility at this node. The fraction of serving the
demand may vary in each period. However, before the construction of a facility on node 10, the demand of
this node is met by the nearest facility which is on node 6.

Table 6. Decision Variable in Experiment 19

Demand
1 1 2 2 3 4 5 6 7 7 8 9 10 10
Point (k)
Source (i) 6 8 6 9 9 8 8 6 6 10 8 9 6 10
1 1.00 0.33 0.67 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
2 1.00 0.31 0.69 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
3 1.00 0.25 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
4 1.00 0.26 0.74 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
5 0.01 0.99 1.00 1.00 1.00 1.00 1.00 0.13 0.87 1.00 1.00 1.00
6 0.04 0.96 1.00 1.00 1.00 1.00 1.00 0.24 0.76 1.00 1.00 1.00
7 0.05 0.95 1.00 1.00 1.00 1.00 1.00 0.32 0.68 1.00 1.00 1.00
8 0.09 0.91 1.00 1.00 1.00 1.00 1.00 0.35 0.65 1.00 1.00 1.00
9 0.14 0.86 1.00 1.00 1.00 1.00 1.00 0.39 0.61 1.00 1.00 1.00
10 0.19 0.81 1.00 1.00 1.00 1.00 1.00 0.44 0.56 1.00 1.00 1.00
Time 11 0.20 0.80 1.00 1.00 1.00 1.00 1.00 0.45 0.55 1.00 1.00 1.00
12 0.24 0.76 0.95 0.05 1.00 1.00 1.00 1.00 0.50 0.50 1.00 1.00 1.00
13 0.25 0.75 0.90 0.10 1.00 1.00 1.00 1.00 0.50 0.50 1.00 1.00 1.00
14 0.28 0.72 0.44 0.56 1.00 1.00 1.00 1.00 0.55 0.45 1.00 1.00 1.00
15 0.30 0.70 0.72 0.28 1.00 1.00 1.00 1.00 0.53 0.47 1.00 1.00 1.00
16 0.31 0.69 0.52 0.48 1.00 1.00 1.00 1.00 0.60 0.40 1.00 1.00 1.00
17 0.35 0.65 0.53 0.47 1.00 1.00 1.00 1.00 0.61 0.39 1.00 1.00 1.00
18 0.37 0.63 0.58 0.42 1.00 1.00 1.00 1.00 0.64 0.36 1.00 1.00 1.00
19 0.40 0.60 0.64 0.36 1.00 1.00 1.00 1.00 0.68 0.32 1.00 1.00 1.00
20 0.41 0.59 0.69 0.31 1.00 1.00 1.00 1.00 0.68 0.32 1.00 1.00 1.00

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

329
 DISCUSSION AND FUTURE WORK
The proposed model in this study and the other studies in the literature are assuming symmetric travel costs
between nodes. However, in water network design or other transportation networks, altitude is a factor that
directly affects travel cost. Therefore, we intend to extend the mathematical model in a way that handles the
problem in three-dimensional space and travel cost can be asymmetric. Regarding this, Brand and Ostfeld
(2011) handled the energy and pressure calculations in a financial framework for a water network design and
their study is one of the main sources of inspiration for this study. We intend to use their formulations for
pumping pipeline construction cost, gravitational pipeline construction cost, pump construction cost, pump
energy cost and treatment plant construction cost. In addition, we consider the option of expanding the
capacity of the facilities opened within the time horizon.

CONCLUSION
This study presents a mathematical formulation for the capacitated facility location and network design
problem. Considering the problems in the world for water sustainability, it is crucial to develop efficient
network designs for wastewater treatment. The mathematical formulation proposed in this study ensures the
meeting the demand and increasing the efficiency of the facilities. As an application of the mathematical
model, an experimental design is presented and the results of this design are discussed. It is clear from the
results that, the efficiency of the network model is above 90% for almost every test instance and considering
the dynamic nature of the model, it appears to make it easier to set up a network.

REFERENCES
Bobat, A. (2019). Water Scarcity in Kocaeli/Turkey: Problems and Measures. The Journal of Scientific and
Engineering Research, 6, 39–52.
Boretti, A., & Rosa, L. (2019). Reassessing the projections of the World Water Development Report. Npj
Clean Water, 2(1), 1–6.
Brand, N., & Ostfeld, A. (2011). Optimal Design of Regional Wastewater Pipelines and Treatment Plant
Systems. Water Environment Research : A Research Publication of the Water Environment Federation, 83,
53–64.
Church, R., & ReVelle, C. (1974). The maximal covering location problem. Papers of the Regional Science
Association, 32(1), 101–118.
Cocking, C., Reinelt, G., & Pfetsch, M. E. (2009). Heuristics for Budget Facility Location-Network Design
Problems with Minisum Objective. In B. Fleischmann, K.-H. Borgwardt, R. Klein, & A. Tuma (Eds.),
Operations Research Proceedings 2008 (pp. 563–568). Berlin, Heidelberg: Springer.
Elloumi, S., Labbé, M., & Pochet, Y. (2004). A New Formulation and Resolution Method for the p-Center
Problem. INFORMS Journal on Computing, 16(1), 84–94.
Farahani, R. Z., Asgari, N., Heidari, N., Hosseininia, M., & Goh, M. (2012). Covering problems in facility
location: A review. Computers & Industrial Engineering, 62(1), 368–407.
Farahani, R. Z., Miandoabchi, E., Szeto, W. Y., & Rashidi, H. (2013). A review of urban transportation
network design problems. European Journal of Operational Research, 229(2), 281–302.
Ghaderi, A., & Jabalameli, M. S. (2013). Modeling the budget-constrained dynamic uncapacitated facility
location–network design problem and solving it via two efficient heuristics: A case study of health care.
Mathematical and Computer Modelling, 57(3), 382–400.
Hakimi, S. L. (1964). Optimum Locations of Switching Centers and the Absolute Centers and Medians of a
Graph. Operations Research, 12(3), 450–459.
Hoekstra, A. Y., & Wiedmann, T. O. (2014). Humanity’s unsustainable environmental footprint. Science,
344(6188), 1114–1117.
Huynen, M. M. T. E., Martens, P., & Akin, S.-M. (2013). Climate change: An amplifier of existing health
risks in developing countries. Environment, Development and Sustainability, 15(6), 1425–1442.
Klose, A., & Drexl, A. (2005). Facility location models for distribution system design. European Journal of
Operational Research, 162(1), 4–29.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

330
Kuehn, A. A., & Hamburger, M. J. (1963). A Heuristic Program for Locating Warehouses. Management
Science, 9(4), 643–666.
Kummu, M., Guillaume, J. H. A., de Moel, H., Eisner, S., Flörke, M., Porkka, M., … Ward, P. J. (2016). The
world’s road to water scarcity: Shortage and stress in the 20th century and pathways towards sustainability.
Scientific Reports, 6(1), 38495.
Leblanc, L. J. (1975). An Algorithm for the Discrete Network Design Problem. Transportation Science, 9(3),
183–199.
Magnanti, T. L., & Wong, R. T. (1984). Network Design and Transportation Planning: Models and
Algorithms. Transportation Science, 18(1), 1–55.
McGregor, P., & Shen, D. (1977). Network Design: An Algorithm for the Access Facility Location Problem.
IEEE Transactions on Communications, 25(1), 61–73.
Melkote, S., & Daskin, M. S. (2001). An integrated model of facility location and transportation network
design. Transportation Research Part A: Policy and Practice, 35(6), 515–538.
Melo, M. T., Nickel, S., & Saldanha-da-Gama, F. (2009). Facility location and supply chain management – A
review. European Journal of Operational Research, 196(2), 401–412.
OECD. (2013). Water and Climate Change Adaptation: Policies to Navigate Uncharted Waters. OECD.
Pearce, R. H., & Forbes, M. (2018). Disaggregated Benders decomposition and branch-and-cut for solving the
budget-constrained dynamic uncapacitated facility location and network design problem. European Journal of
Operational Research, 270(1), 78–88. https://doi.org/10.1016/j.ejor.2018.03.021
Puerto, J., Ricca, F., & Scozzari, A. (2018). Extensive facility location problems on networks: An updated
review. TOP, 26.
Ravi, R., & Sinha, A. (2006). Approximation Algorithms for Problems Combining Facility Location and
Network Design. Operations Research, 54(1), 73–81.
Saatci, A. M. (2013). Solving Water Problems of a Metropolis. Journal of Water Resource and Protection,
05(04), 7–10.
Saatci, E., & Akpinar, E. (2007). Assessing Poverty and Related Factors in Turkey. Croatian Medical
Journal, 48.(5), 628–635.
Şahin, G., & Süral, H. (2007). A review of hierarchical facility location models. Computers & Operations
Research, 34(8), 2310–2331.
Terouhid, S. A., Ries, R., & Fard, M. (2012). Towards Sustainable Facility Location – A Literature Review.
Journal of Sustainable Development, 5.
TÜİK - Veri Portalı. (n.d.). Retrieved August 5, 2021, from https://data.tuik.gov.tr/Kategori/GetKategori?
p=cevre-ve-enerji-103&dil=1
van Leeuwen, K., & Sjerps, R. (2016). Istanbul: The challenges of integrated water resources management in
Europa’s megacity. Environment, Development and Sustainability, 18(1), 1–17.
Weber, A. (1909). Theory of the Location of Industries. University of Chicago Press.
World Economic Forum. (2020). The Global Risks Report 2020. Retrieved from
https://www.weforum.org/reports/the-global-risks-report-2020/
Yang, H., & H. Bell, M. G. (1998). Models and algorithms for road network design: A review and some new
developments. Transport Reviews, 18(3), 257–278.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

331
LEAN TRANSPORTATION MANAGEMENT THROUGH AN APPLICATION
OF VEHICLE ROUTING PLANNING: A CASE STUDY

Nattapong Kongprasert 1*, Chichakorn Duriyarapee 2, Watcharain Larptrakool 3

Abstract ⎯ The digital transformation is changing business models, methods of production and distribution, and
the way firms compete. Every business needs to invest in information technology to acquire a competitive
advantage. The leading distributor of industrial tools and machines in Thailand is facing the challenge of
massive disruption and change. The rise of information technology has paved the new way for providing services
to meet customers’ needs or expectations. Therefore, the firm has considered investing in information
technology together with reducing costs or increase profits at the same time. This study proposed increasing
the efficiency of transportation management by using an application of vehicle routing planning. It focused on
the transportation planning process. Lean manufacturing is a philosophy and tool that was used to reduce costs
and eliminate activities. It was implemented along with the vehicle routing planning tool for supporting the
officers to assign routes for drivers to make goods deliveries with the shortest total distance and the least
delivery time. The results found that the number of vehicles decreased by 27.16%, the total delivery time was
decreased by 7.22%, and the transportation cost was reduced by 17.48%. They can improve the efficiency of
transportation management in a cost-effective and timely manner.

Keywords ⎯ Lean manufacturing, transportation management, transportation planning process, vehicle routing
planning

1. INTRODUCTION
Thailand is the second-largest economy in ASEAN after Indonesia. During 2019 and 2020, the country had
been significantly affected by the COVID-19 pandemic and the related lockdown, leading to low consumption
levels and disruptions in supply chains. All firms are facing business problems and challenges at the same time.
Small-size and low-tech firms have suffered due to reduced orders and a shortage of inputs, including raw
materials, owing to the disruption of the supply chain. They have to force to lay off employees or terminate the
business. The remaining companies have to struggle to survive and preserve their business. There are three ways
businesses from the World Economic Forum`s New Champions Community of forward-looking that are
interrelated and need focused strategies to ensure success. First, companies need to have the ability to quickly
change or adapt in response to the tumultuous and rapidly changing market. Second, companies have to sustain
productivity through digitized operations and focus on new opportunities spurred by digitalization for their work
and workforce. Third, companies have to focus on the importance of multistakeholder collaboration and the
need for systematic resilience (Menon, 2020). Then, this required lots of innovation, creativity and the
development of new business models. Entrepreneur needs to invest in information technology to acquire a
competitive advantage and preserve their business.

The automotive industry in Thailand is a key industry for Thai economy with strong infrastructures and a vast
network of small and large, local and foreign companies all along the car-production supply chain. It has been
significantly developed for over 50 years. Sumipol Corporation Limited is one of the leading distributors of
industrial tools and machines that participates in this research. This company assists the automotive
manufacturing company in productivity improvements, promote quality systems, and take part in driving the
Thailand Industry to 4.0 through automation and smart technology. Transportation is one of the major logistics
activities of this company that involves the physical delivery of goods from the company to the end customer.
Then, transportation costs can be one of the most significant logistical expenses. Moreover, this company is

1
Nattapong Kongprasert, Department of Industrial Engineering, Faculty of Engineering, Srinakharinwirot University, Nakhon Nayok, Thailand,
[email protected]
2
Chichakorn Duriyarapee, Department of Industrial Engineering, Faculty of Engineering, Srinakharinwirot University, Nakhon Nayok, Thailand,
[email protected]
3
Watcharain Larptrakool, Sumipol Corporation Limited, Bangkok, Thailand, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

332
facing the challenge of massive disruption and change. The rise of information technology has paved the new
way for providing services to meet customers’ needs or expectations. To minimize transportation costs and meet
customers’ needs or expectations, the company needs to adapt the business strategy to invest in information
technology together with reducing costs and acquiring a competitive advantage. Then, this study proposes the
increasing the efficiency of transportation management by using an application of vehicle routing planning. It
focuses on the transportation planning process. Lean manufacturing is a tool that is used to reduce costs and
eliminate activities. It is implemented along with the vehicle routing planning tool for supporting the officers to
assign routes for drivers to make goods deliveries with the least total distance and delivery time.

2. LITERATURE REVIEWS
Lean Manufacturing is a philosophy that is to minimize waste while productivity remains constant. It began and
was applied by Toyota during the 1950s which is famously known as Toyota Production System (TPS). The
goals of TPS are to eliminate waste, streamline processes, increase efficiency, improve productivity, respect
people, and please the customers. TPS directly influenced the creation of lean manufacturing. At first, Lean
manufacturing was presented as the generic version of TPS that focused on cost reduction and productivity
improvement for businesses to learn and implement into their own facilities (Womack et al, 1990). Over time,
Lean manufacturing is interpreted farther away from TPS and more its own unique thing. Today, Lean
manufacturing is not suited only for manufacturing. It can apply to every business and every process. It is the
strategy that seeks to deliver more value to customers as efficiently as possible (Womack & Jones, 2003). Lean
manufacturing focuses on the customer, workflow, accuracy, and relationships first while TPS focuses on
respect for employees and continuous improvement first. However, waste elimination and continuous
improvement are present in both strategies. The 5 Lean manufacturing principles are still at the core and guide
to implement Lean manufacturing starting by 1) specifying the value, 2) identifying the value stream, 3) making
the value flow, 4) configuring the pull system by the customer, and 5) pursuing towards perfection. Based on
the Toyota Production System, Lean manufacturing consists of a set of powerful “tools” and “techniques” that
assisted in the identification and steady elimination of waste (Muda) such as 5S, just-in-time (JIT), Jidoka,
PDCA, Kanban (Pull system), Andon, Gemba, Hejunka (Level production), Kaizen (Continuous improvement),
total productive maintenance (TPM), ECRS, single minute exchange of dies (SMED), value stream mapping
(VSM) and standardized work (Rohania & Zahraeea, 2015) (Liker & Hoseus, 2008)( Yadav et al, 2012). These
tools and techniques serve as the basis of TPS and Lean manufacturing that stabilizes processes, makes
production flow smoothly based on customer demand, identifies defects to improve quality, and ultimately
delivers what the customer needs. Many studies have applied the lean manufacturing to transportation
management. Bottani et al (2015) proposed an integrated approach to increase sustainability and efficiency of
logistics activities. The approach was conceived to design an integrated food district, starting from a set of
independent companies, and to quantify the economic savings that could be gained from the pooled management
of logistics and transportation activities. Villarreal et al (2016) studied a method of improvement based on lean
thinking, specifically in the forms of concepts derived from the Toyota’s seven production and business wastes
and tools such as VSM, has been applied to the road transport operations of a leading Mexican brewing
organization. Kurganov et al (2021) studied the evolution of lean thinking and optimize the value stream in the
transport provision of enterprises. From many studies, it illustrated that lean manufacturing creates value for
customers by maximizing productivity while minimizing wasted effort and expense.

Vehicle routing problem (VRP) concerns the service of transportation, distribution, and logistics company. It
involves finding optimal routes for a fleet of delivery cars and trucks, each with limited capacity, so as to
minimize the total transportation distance (Leelertkij et al, 2021). It is a set of routes for a fleet of vehicles based
at one or several depots must be determined for a number of geographically dispersed cities or customers. The
objective of the VRP is to deliver a set of customers with known demands on minimum-cost vehicle routes
originating and terminating at a depot. In real operations, route planning and scheduling are customarily
performed manually based on individual experience. VRP can basically be used to find a set of least-cost routes
serving a group of customers with a set of vehicles located at a central depot, under various types of objectives
and constraints (Allahyari et al, 2021). Many studies have applied the vehicle routing problem in many sectors.
Lin et al (2014) applied vehicle routing problem for selective waste collection contributes to a more efficient
reverse logistics system. Anghinolfi et al (2016) proposed a metaheuristic approach for the operational planning
of the daily logistic activities in a gas distribution network that based on vehicle routing with time window
model. Franca et al (2019) studied the vehicle routing software ORTEC Routing and Dispatch in order to
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

333
optimize the operation of a company that performs the collection of municipal solid waste from 1764
commercial establishments in Rio de Janeiro. Allahyari et al (2021) developed a rich variant of the vehicle
routing problem to solve the routing problem of security carriers for high-value shipment transportation. From
many studies, they shown that the vehicle routing problem is used to optimize routes for time, optimize routes
base on distance, and assign the drivers to the right stops.

Although many research papers have been published, there is no review focused on minimize transportation
costs and meet customers’ needs or expectations in Thai automotive industry by using an application of vehicle
routing planning. Therefore, the motive for this review is to present and increase the efficiency of transportation
management by using an application of vehicle routing planning.

3. MATERIALS AND METHOD

This study was to increase the efficiency of transportation management by using an application of vehicle
routing planning. It focused on the transportation planning process. The vehicle routing software TSQUARE
LOGISTICS was used to support operational decision-making and planning. Lean manufacturing was used to
reduce costs and eliminate activities in the transportation planning process. It was implemented along with the
vehicle routing planning tool (TSQUARE LOGISTICS) for supporting the officers to assign routes for drivers
to make goods deliveries with the least total distance and delivery time. This study was done in 4 phases. First,
it was to study the current transportation planning process. Second, it was to study the vehicle routing software
TSQUARE LOGISTICS. Third, it was to create the new transportation planning process. Forth, it was to
evaluate the performance of the vehicle routing software TSQUARE LOGISTICS.

First phase: Study the current transportation planning process

This phase was to study the current transportation planning process. The current transportation planning process
has 5 steps: 1) summarize and print the delivery order, 2) make the delivery plan, 3) Prepare the products/goods,
4) check the products/goods, and 5) deliver the product/goods as shown in Figure 1.

1. Summarize and 5. Deliver the

2. Make the delivery 3. Prepare the 4. Check the product/
print the delivery products/goods to
plan product/goods goods
order customers

Figure 1. Current transportation planning process.

1) Summarize and print the delivery order - Microsoft Business Solutions-Axapta is an Enterprise Resource
Planning (ERP) system that is used in the company. Salespersons make the sale orders and delivery orders all
day through Microsoft Business Solutions-Axapta. After salespersons do not add the sale orders and delivery
orders in the system, the transportation planning officers summarized and printed out the delivery order, packing
order and list of customers at 4.30 PM every day before the delivery date. It has about 200 list of customers per
day. The vehicle route can be assigned in 12 routes based on the customer addresses as shown in Figure 2.

Figure 2. The vehicle routes.

© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

334
2) Make the delivery plan - After summarize and print out the delivery order and list of customers, the expert
transportation planning officers make the delivery plan. They spend time about 1 hour and 30 minutes to make
the delivery plan per day for clustering the customers and assigning the vehicle routes for drivers by using their
experience. The company has 8 cars and 7 motorcycles for delivery of the products to customers. Motorcycles
are used for vehicle routes that deliver the products in Bangkok Metropolitan Region. Each vehicle needs to
deliver the products to many customers and come back to the company in working time to avoid the over time.
Thus, it is difficult for new transportation planning officers to make the delivery plan. They have to take more
time to learn and practice from the transportation planning officers that have more experience.

3) Prepare the products/goods - After making the delivery plan, the transportation planning officers prepare the
products or goods for each vehicle to be ready to deliver to customer tomorrow. They arrange the products into
different boxes and separate them by the vehicle routes for each truck.

4) Check the products/goods - Before going out to deliver the products, the drivers have to check and verify the
products in the box with a delivery note and invoice. Both documents detail the name and address of the
customer, list of the products and quantity of the goods,

5) Delivery the products/goods - The drivers have to carefully check the delivery time and the delivery point of
each customer to manage the driving time and avoid the traffic jam. Moreover, the recipients have to sign on
the document to confirm that they already received the products.

Second phase: Study the vehicle routing software TSQUARE LOGISTICS

This phase was to study the vehicle routing software TSQUARE LOGISTICS. The company has a policy to
improve the efficiency of transportation planning process by using the vehicle routing software TSQUARE
LOGISTICS. The constrains of this study based on the shortest total distance and the least delivery time. Figure
3 illustrate the framework of TSQUARE LOGISTICS. It was done in 4 steps: 1) prepare the delivery point data
and vehicle data, 2) upload the Excel file into the TSQUARE LOGISTICS and set the mode condition, 3)
generate the vehicle routes, 4) adjust and edit the vehicle routes, and 5) make the master delivery plan and
delivery report.

Delivery Point Data Master Delivery Plan

TSQUARE
LOGISTICS

Vehicle Data Delivery Report

Figure 3. The framework of TSQUARE LOGISTICS.

1) Prepare the delivery point data and the vehicle data – The delivery point data and the vehicle data have to
provide in Excel file format. The delivery point data is composed of delivery point ID, latitude and longitude of
delivery point, customer ID, service time, working time, weight, volume, type of vehicle, priority level, name
of customer, address of customer, and phone. The vehicle data is composed of vehicle ID, latitude and longitude
of company parking, working time, breaking time, speed limit, volume, weight, type of vehicle, priority level,
time to leave, reverse delivery, distance limit, and maximum drop per vehicle.

2) Upload the Excel file into the TSQUARE LOGISTICS and setting the mode condition – This step is to upload
the Excel file into the TSQUARE LOGISTICS and set the mode condition. Users can set the mode condition
that is consisting of allowing toll way, avoid traffic/shortest distance, loading efficiency, minimize
vehicle/utilize all vehicles, zone-based planning, and maximum drop per vehicle.

3) Generate the vehicle routes – This step is to generate the vehicle routes. The system proposes the best route
under the conditions that are selected from previous step. It displays the number of vehicles, vehicle type,

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

335
delivery time, volume, weight, number of delivery points, total distance, and the list of all delivery points of
each delivery vehicle as shown in Figure 4-6.

4) Adjust and edit the vehicle routes – Users can adjust and edit the ordering of delivery point or the vehicle
routes.

5) Make the master delivery plan and delivery report – This step is to make the master delivery plan and delivery
report in Excel format that can be used in the next process.

Figure 4. Generate the vehicle routes.

Figure 5. Information of each vehicle.

Figure 6. Information for each delivery point.

© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

336
Third phase: Create the new transportation planning process.
This phase was to create the new transportation planning process. The new transportation planning process is
an integration of the current transportation planning process and TSQUARE LOGISTICS process as shown in
Figure 7. It was done in 5 main steps: 1) summarize and export the delivery order in Excel file, 2) make the
master delivery plan and delivery report, 3) Prepare the products/goods, 4) check the products/goods, and 5)
deliver the product/goods.

1) Summarize and export the delivery order in Excel file - After salespersons do not add the sale orders and
delivery orders in the system, the transportation planning officers summarized and export the delivery order in
Excel file at 4.30 PM every day before the delivery date.

2) Make the master delivery plan and delivery report – This step has 5 sub steps that following the TSQUARE
LOGISTICS process

3) Prepare the products/goods, 4) check the products/goods, and 5) deliver the product/goods are the same as
the previous process.

2.1 Prepare the

delivery point data
and the vehicle data

2.2 Upload the Excel

file into TSQUARE
LOGISTICS

1. Summarize and 5. Deliver the

2.3 Generate the 3. Prepare the 4. Check the product/
export the delivery products/goods to
vehicle routes product/goods goods
order in Excel file customers

2.4 Adjust and edit

the vehicle routes

2.5 Make the master

delivery plan and
delivery report

Figure 7. New transportation planning process.

Forth phase: Evaluate the performance of the vehicle routing software TSQUARE LOGISTICS.
This phase was to evaluate the performance of the TSQUARE LOGISTICS. Two experiments were used to
evaluate the performance of the TSQUARE LOGISTICS. First, it was to compare the efficiency of the previous
process and new process. Second, it was to explore the break-even point of the TSQUARE LOGISTICS. The
results are shown and discussed in next section

3. RESULTS AND DISCUSSION

First: Compare the efficiency of the previous process and new process.
This experiment was to compare the efficiency of the previous process and the new process. The indicators were
the number of vehicles, total distance, total delivery time, and total transportation cost. This experiment was
done during March 1-31, 2021. The delivery points and the delivery orders were constant parameters. The data
from the previous transportation planning process is the information that occurs from actual activities and
situations. The data from the new transportation planning process is the information that was calculated from
the TSQUARE LOGISTICS. The results were summarized as shown in Table 1.

The number of cars and the number of motorcycles was decreased. They affected the total vehicles that was
decreased from 232 to 169. It has decreased by 27.16%. It shown that the TSQUARE LOGISTICS can cluster
and make the vehicle routes that are more effective than humans. The total distance was decreased from 31,696
km. to 26,236 km. It has decreased by 17.23%. The total planning time is an important and effective result. It

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

337
was decreased from 1,650 minutes to 755 minutes It has decreased by 54.24%. It illustrated that the TSQUARE
LOGISTICS can support the new transportation planning officers to make the delivery plan. The total service
time was increased from 18,659 minutes to 21,940 minutes. The service time of previous process is an actual
service time while the service time of new process is an estimate time. It can be error and effect on the total
service time. It has to adjust and update before uploading the data file to the TSQUARE LOGISTICS. The total
toll way cost and the total fuel cost were decreased. They affected the total transportation cost that can saving
29,952 Thai Baht. It shown that the TSQUARE LOGISTICS can generate and optimize the vehicle routes that
is more effective than humans.

Table 1. The comparative results of previous process and new process.

Second: Explore the break-even point.

Break-even point analysis (BEP) was used in this study to evaluate the margin of safety. The company has
a fixed cost for using the TSQUARE LOGISTICS. The company has to pay 15,000 Thai Baht per month.
During March 2021, it has 23 working days. The saving cost and the TSQUARE LOGISTICS cost were
used to calculate the break-even point. The break-even point result illustrated in Figure 8. It shown that the
saving cost and the TSQUARE LOGISTICS cost intersected at 11.52 days, it is break-even point. It can be
summarized that the break-even point for using the TSQUARE LOGISTICS is 12 days.

Figure 8. Break-even point result.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

338
 4. CONCLUSION
This study proposed increasing the efficiency of transportation management by using an application of vehicle
routing planning. The vehicle routing software TSQUARE LOGISTICS was implemented in the transportation
planning process. It was used for supporting the officers to assign routes for drivers to make goods deliveries
with the shortest total distance and the least delivery time. Two experiments were used to evaluate the
performance of the TSQUARE LOGISTICS. First, it was to compare the efficiency of the previous
transportation planning process and the new transportation planning process. Second, it was to explore the
break-even point of the TSQUARE LOGISTICS. The results shown that the number of vehicles decreased by
27.16%, the total delivery time was decreased by 7.22%, and the transportation cost reduced by 17.48%. The
break-even point for using the TSQUARE LOGISTICS is 12 days. It illustrated that TSQUARE LOGISTICS
can increase the efficiency of transportation management. It can cluster and make the vehicle routes that are
more effective than humans. The limitation of this study was the service time that was estimated from the
experience of the transportation planning officers. Thus, further research should focus on the activities of the
service time at the delivery point.

REFERENCES
Allahyari, S., Saeed Yaghoubi, S., & Van Woensel, T. (2021). A novel risk perspective on location-routing
planning: An application in cash transportation. Transportation Research. Part E: Logistics and Transportation
Review. 150
Anghinolfi, D., Paolucci, M., & Tonelli, F. (2016) A vehicle routing problem with time windows approach for
planning service operations in gas distribution network of a metropolitan area. IFAC-PapersOnLine. 49(12),
1365–1370
Bottani, E., Rizzi, A., & Vignali, G. (2015). Improving logistics efficiency of industrial districts: a framework
and case study in the food sector. International Journal of Logistics Research and Applications. 18(5), 402-423
Franca, L.S., Ribeiroa, G.M., & Chavesb, G.L.D. (2019). The planning of selective collection in a real-life
vehicle routing problem: A case in Rio de Janeiro. Sustainable Cities and Society. 47
Kurganov,V., Sai, V., Gryaznov, M., & Dorofeev, A. (2021). The emergence and development of lean thinking
in transport services. Transportation Research Procedia. 54, 309–319
Leelertkij, T., Parthanadee, P., & Buddhakulsomsiri, J. (2021). Vehicle routing problem with transshipment:
mathematical model and algorithm. Hindawi Journal of Advanced Transportation.
Liker, J.K. & Hoseus, M. (2008). Toyota Culture: the heart and soul of the Toyota Way. New York: McGraw-
Hill
Lin, C., Choy, K. L., Ho, G. T., Chung, S. H., & Lam, H. Y. (2014). Survey of green vehicle routing problem:
Past and future trends. Expert Systems with Applications, 41(4), 1118–1138.
Menon, N. (2020, Jul 23). Here are 3 ways businesses can survive and thrive through COVID-19 and beyond.
Retrieved from https://www.weforum.org
Rohania, J.M., & Zahraeea, S.M. (2015). Production line analysis via value stream mapping: a lean
manufacturing process of color industry. Procedia Manufacturing, 2, 6–10
Villarreal, B., Garza-Reyes, J.A., Kumar, V., & Lim, M.K. (2016). Improving road transport operations through
lean thinking: A case study. International Journal of Logistics Research and Applications: A Leading Journal
of Supply Chain Management. 20(2), 163-180
Womack, J.P., & Jones, D.T. (2003). Lean Thinking, Simon & Schuster. New York: Free Press
Womack, J.P., Jones, D.T., & Roos, D. (1990). The machine that changed the World: The triumph of lean
production, New York: Rawson Macmillan
Yadav, R., Shastri, A., & Rathore, M. (2012). Increasing productivity by reducing manufacturing lead time
through value stream mapping. International Journal of Mechanical and Industrial Engineering, 1(3), 31-35

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

339
 AN INTEGRATED SUPPLIER SELECTION MODEL CONSIDERING RISK
CRITERIA IN THE TIRE INDUSTRY
Elif Sena Kutlu1, Merve Er2

Abstract ⎯ Tire Industry has a highly competitive and dynamic structure with changing customer demands
and high number of alternative brands in the market. Tire dealers consider many criteria for choosing the
best suppliers, e.g., geographical location, brand awareness, quality, price, road handling, having roadside
assistance. It is critical to reduce supply related risks to survive in the global automotive tire market.
Therefore, this study proposed an integrated approach to select the most suitable supplier considering
different risk factors as well as traditional supplier assessment criteria in the tire industry. In the first phase,
suppliers were grouped based on their risk profile. The weights of the risk criteria were calculated by using
Fuzzy Analytic Hierarchy Process. Then, k-means clustering method was used to categorize suppliers based
on pre-determined risk criteria. In the second phase, the weights of the traditional criteria were assessed by
using AHP and TOPSIS method was used to select the best supplier within the lowest-risky supplier cluster by
using traditional supplier assessment criteria. The proposed approach integrated a MCDM method with the
cluster analysis of data mining field to ensure reduction of supply related risks. The proposed integrated
approach will increase the performance of the supplier selection process and mitigate the supplier-related
risk factors at the beginning of the procurement phase.

Keywords ⎯ Clustering, Fuzzy AHP, Supplier selection, Risk, Tire industry, TOPSIS

INTRODUCTION

In today’s global, dynamic, and competitive business environment, supply chain management has
become more difficult than ever. A successful supply chain management is vital for the performance of
companies (Alkan et al., 2017). Especially, selection of a supplier must be carefully accomplished to survive in
this complex and uncertain business environment. Supplier selection has a crucial role for the success of an
organization. Purchasing constitutes an important part of total production cost. Generally, a company spends
almost 60% of its total sales on purchasing raw materials, assemblies, components, parts, etc. (Hamdi et al.,
2015). Therefore, nowadays, making decisions quickly and effectively for supplier selection has been one of
the most important goals of companies.
Supplier evaluation and selection is a complex, risky, and hard multi-criteria decision making (MCDM)
problem that provides a systematical way to assess and select suppliers (Kuo et al., 2015). Supplier selection is
a very complex issue for companies, and there are many uncertainties about choosing the best supplier from the
alternatives offered. Companies select the criteria based on their requirements. Most common criteria in
supplier selection are cost, quality, and delivery performance (Rezaei et al., 2016; Tavana et al., 2017). In
recent years, companies started to understand the importance of considering risks of suppliers (Er Kara and
Oktay Fırat, 2018). Managing risks is a daily issue for supply chain and logistics management. The ability to
respond to and mitigate these risk events enables the company to be ahead of its competitors. Some of the
supplier related risk factors can be listed as follows: late delivery, variable cost, operational risk, quality, bad
performance history, uncertain lead time, and flexibility risk (Er Kara and Oktay Fırat, 2018).
In the literature, various approaches have been proposed to solve the supplier evaluation and selection
model (Mardani et al., 2015; Renganath et al., 2016). Especially, MCDM methods are commonly used to solve
this problem; AHP, ANP, DEA, TOPSIS, ELECTRE, DEMATEL, PROMETHEE, VIKOR, etc. (Mardani et
al., 2015; Chang et al., 2011). However, current business environment is changing with increasing
globalization, growing complexity of supply chains and higher competition and dynamism in the markets.
Therefore, supplier selection process should also be improved based on the requirements and developments in
today’s business conditions. Literature is still dominated with the studies considering traditional criteria such as
1
Elif Sena Kutlu, Marmara University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Turkey, [email protected]
2
Merve Er, Asst. Prof. Dr., Marmara University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]

340
cost, quality, performance, delivery, past performance attitude, facility, flexibility, and service (Amini and
Asoodar, 2016). Therefore, this study proposed a two-phased approach for the supplier evaluation and selection
process. In the first phase, suppliers were grouped based on their risk profile. K-means clustering algorithm was
used to categorize suppliers based on pre-determined risk criteria. Weighted scores were used as an input for
the K-means algorithm. Fuzzy AHP was used to identify the importance level of the risks. Clustering is a new
approach to classify suppliers into smaller and manageable groups. By clustering suppliers, a more realistic
comparison and a faster operation can be made. Also, the characteristics of each small group of suppliers can be
summarized easily (Parmar et al., 2010). In the second phase, TOPSIS method was used to select a supplier
within the lowest-risky supplier cluster by considering traditional criteria. AHP was also used to identify the
importance level of these traditional criteria.
Nowadays, considering supply side risks is becoming significant. Recently, the awareness on the
importance of supplier related disruptions gained increased popularity. However, this area is in its infancy
stage. There is a need for comprehensive approaches that consider various risk dimensions in the supplier
selection process. Therefore, this study incorporated supplier related risks into this strategical decision-making
stage. Additionally, there are less number of papers that utilize from the cluster analysis in the supplier
selection step. Hence, the study will make a significant contribution to the supply chain literature with the
proposed mixed approach.

SUPPLIER SELECTION PROBLEM AND THE TIRE INDUSTRY

Supplier evaluation and selection is usually a very complex and difficult multi-criteria decision-making
problem including many different criteria. Choosing the right supplier decreases the costs of the company,
increases supply chain performance, decrease supply related risks, and increase competitiveness of the
company in the market (Er Kara and Oktay Fırat, 2018; Supçiller and Çapraz, 2011). Various methods have
been employed for the supplier evaluation and selection problem in the literature, e.g., Multiple Criteria
Decision Making (MCDM) methods, mathematical programming models, metaheuristics, fuzzy logic
(Schramm et al., 2020; Moghaddam, 2015; Nazari-Shirkouhi et al., 2013). Many different criteria are used for
evaluating and selection suppliers. Most common criteria are cost, quality, and delivery performance (Rezaei et
al., 2016; Tavana et al., 2017). Literature is still dominated with the studies considering traditional criteria such
as cost, quality, performance, delivery, past performance attitude, facility, flexibility, and service (Haq and
Kannan, 2006; Amini and Asoodar, 2016).
Problems or disruptions in the supply level affects the performance of the whole supply chain network.
Hence, risk factors should be considered while evaluating and selecting the suppliers. The literature is
dominated with the studies considering more traditional approaches that consider more generic criteria such as
cost, quality, delivery speed, sustainability, etc. There is a growing trend on the consideration of the supplier-
related risks in the supplier evaluation process, however, this area is new (Er Kara and Oktay Fırat, 2018). In
recent years, supplier related risks gained increased importance especially due to recent famous supplier related
disruptions (Er Kara and Oktay Fırat, 2015). Hantsch and Huchzermeier (2013) determined supply chain
management risks as procurement risk, production risk, sales risk, financial market risk, risk specific to the
production network, political/legal risk, and other risks.
Supplier evaluation and selection criteria are selected based on the characteristics of the industry and
needs of the company. With the increased globalization, competition and diversity in the markets, this problem
is getting more complicated and difficult. Tire industry has one of the most competitive markets with changing
customer demands and the scope of offered products from various companies. Tire dealers consider many
criteria while choosing the best suppliers, such as geographical location of the dealer, their supply status, origin
of the manufacturer (domestic or foreign production), end-user’s perception, brand awareness, being a premium
brand, product quality, labeling criteria (fuel use, braking distance, road holding, sound level), price, road
handling, having a roadside assistance campaign. There are many alternative brands in the tire industry that
may provide these criteria. It is critical to reduce supply related risks to survive in this wide and competitive
market.
Original Equipment and Replacement tires are two segments in consumer tire industry. Tires fitted with
vehicles on the assembly line are called original equipment. Vehicle manufacturers buy tires from the
companies that conform to their specification limits. Customers come to dealers when they want to change their
tires. This market is called the replacement market. Recent years have seen an increasing number of low-end

341
Chinese brands which have introduced many low-cost tires into the global market. Strong growth in the global
market in recent years, and continued expected strong growth in the coming years, serves to make the market
look more attractive to potential new entrants. Consolidation is possible, and it could create opportunities for
new premium brands. Tire manufacturers have two main groups of buyers: vehicle manufacturers (such as
General Motors, Volkswagen, and Ford) and tire dealers. The tires and rubber market are difficult to enter due
to high fixed costs. Players in the market require big facilities with good infrastructure to be easily accessible
for supply and distribution; these kinds of areas require high rent, increasing fixed costs. In addition, they
would require high levels of power supply such as electricity and water, increasing utilities costs. Lastly, to
keep the machinery and the facilities up to date, they would require monthly maintenance from experienced
personnel, increasing fixed costs in the long run. In the world of consumer tires, tires are divided into Segment
A and Segment B. The Segment A tires are 17 rims and over. Segment B tires are 16 rims and under tires.
Summer, 4 seasons and winter tires are also available. Passenger, SUV/4x4, Light Truck are the product groups
in consumer tires. Sales channels are regional distributors, retailers, car dealers and car fleets. Tire life may
vary depending on usage and conditions, but if used under specified conditions and does not receive any
impact, it is approximately 4-8 years or 50.000-80.000 km. Also, when the tread depth is reduced to the legal
limit of 1.6 mm, it is time to change the tire. (Altın et al., 2013) The tire is important for safety and comfort and
at the same time expensive equipment. Appropriate tire selection: it has always been a challenging process for
car owners to decide.

METHODOLOGY

A two-stage approach is developed to solve the supplier evaluation and selection problem in the tire
industry. First stage includes the elimination of risky suppliers based on the results of cluster analysis that used
supplier risk profile data. Hence, a manageable set of suppliers with a lower risk-level is created. K-means
algorithm clusters observations into k groups where k is an input parameter. It assigns observations into
clusters based on their distance to the mean of the cluster. The means of the clusters are computed in each
iteration (Er Kara, 2017) Second stage includes the selection of the best supplier by using some traditional
supplier evaluation criteria by using the TOPSIS method. A technique for establishing order performance by
similarity to the ideal solution (TOPSIS) was first developed for solving MCDM problems. The classical
TOPSIS method is based on the idea that the best alternative should have the shortest distance from the
positive ideal solution and the greatest distance from the negative one. (Dymova et al.,2013)
First, a set of risk criteria was identified based on literature review. Selected risk factors are
communication risk, reliability risk, financial risk, logistics risk and price fluctuations in the market. Detailed
definitions are shown in Table 1.

Table 1. Definitions of the risk criteria for the first stage of the model

Risk Criteria Definition of risk criteria

Communication A good communication is important to understand customer requirements and solve problems.
Risk This type of risk includes poor, negative, difficult ineffective and unclear communication.
Trustfulness, confidence, and brand reputation are important for healthy long term business
Reliability Risk
relationship. This risk type includes non-compliance to promises and dishonesty
Financial operations and transactions are very important for business performance and
Financial Risk sustainability of the company. Financial risk includes poor financial management, economic
uncertainty, financial uncertainty, late payment and lack of different payment options
Logistics performance is very important for business continuity. Logistics risk includes late
Logistics Risk delivery, transportation failures, long distances, technical problems with transportation vehicle,
location of supplier, existence of multi modal transportation requirements
Quality, technology, and innovation of tires directly influence customer satisfaction. Quality risk
Quality Risk
includes quality defects, violation of quality specifications and limited product portfolio
Price fluctuations Price fluctuation directly impacts the overall profitability. It includes sensitivity of price to change
in the market in the market

342
 Sixty-four tire suppliers were evaluated based on the criteria in Table 1 by three experts. These experts
are tire dealers. Supplier selection was examined from tire dealers’ side. Eleven of the suppliers are from
Turkey and others are foreign companies. According the marketing research, customers give high importance to
tire’s production country. Far East produced are not preferred much. Additionally, risk criteria weights were
identified by applying fuzzy AHP method. Weighted risk data were used as an input for the k-means clustering
algorithm.
Supplier groups with different risk levels were identified as a result of the cluster analysis. Then, the
supplier cluster with the lowest-risk profile was selected for the second stage of the proposed model. AHP was
used to identify the weights of the traditional criteria. The selected traditional criteria were being a reliable
brand, being a brand used before, being a recommended brand, high quality, being enduring, high profitability,
can be found easily/high availability, having different tires suitable for different weather conditions, being a
foreign brand, being affordable, being worth for the money paid, being a domestic brand, having good price-
performance balanced products, investing in advertising, providing value added services such as road support,
making campaign-promotions regularly and having environment friendly products. TOPSIS method was used
to select the best suppliers among the identified supplier set.

RESULTS

First, three tire dealers from the industry evaluated supplier risk criteria based on a fuzzy scale that is
given in Table 2.

Table 2. Triangular fuzzy scale table

Scale Linguistic Fuzzy Triangular Scale

1 Equally important (1,1,1)
3 Weakly important (2,3,4)
5 Fairly important (4,5,6)
7 Strongly important (6,7,8)
9 Absolutely important (9,9,9)
2 (1,2,3)
4 The intermittent values (3,4,5)
6 (5,6,7)
8 (7,8,9)

The weights of the risk criteria that are identified by using the fuzzy AHP method are represented in
Table 3.
Table 3. Results and weights of risk criteria

Risk criteria Weights

Communication Risk 0.05
Reliability Risk 0.23
Financial Risk 0.14
Logistics Risk 0.06
Quality Risk 0.20
Price fluctuations in the market 0.33

64 suppliers were also evaluated by the three experts from the industry. These experts have different
expertise and knowledge in the area. Therefore, the following weights are assigned to these decision makers;
0.49, 0.39 and 0.12. Then, k-means clustering algorithm was used to analyze the weighted risk scores. Scree
plot was drawn to identify the number of clusters. The scree plot in Figure 1 represents the eigenvalues plotted
against the number of factors. A break occurs in two factors. Based on the eigenvalue criteria and the scree plot,
two factors seem reasonable for supplier risk data. Therefore, two clusters were identified based on k-means
algorithm by using SPSS software.

343
 Figure 1. Scree plot

The first supplier cluster includes 14 suppliers, and the second cluster includes 50 suppliers. Table 4
gives the suppliers in each cluster.

Table 4. Suppliers in each cluster

Cluster 1 Cluster 2
S1, S2, S3, S4, S5, S8, S10, S11, S12, S14, S15, S16,
S17, S18, S19, S21, S22, S23, S24, S27, S28, S29,
S6, S7, S9, S13, S20, S25, S26, S31, S39, S40, S41, S30, S32, S33, S34, S35, S36, S37, S38, S42, S43,
S50, S54, S55 S44, S45, S46, S47, S48, S49, S51, S52, S53, S56,
S57, S58, S59, S60, S61, S62, S63, S64

Cluster centers were assessed to label these clusters. Figure 2 shows the spider diagram of the cluster
centers based on different risk factors. As it can be seen from the figure, Cluster 2 has the highest risk levels,
and hence labeled as “risky supplier group”. Cluster 1 was labeled as “lower-risky supplier group”.

Figure 2. Spider diagram of the cluster centers based on different risk factors
Cluster 2 was eliminated due to the high-risk profile of the suppliers. Hence, cluster 1 was selected for
further analysis. Traditional supplier criteria were used in the second stage. The selected 17 traditional criteria
are as follows: being a reliable brand, being a brand used before, being a recommended brand, high quality,
being enduring, high profitability, can be found easily/high availability, having different tires suitable for

344
different weather conditions, being a foreign brand, being affordable, being worth for the money paid, being a
domestic brand, having good price- performance balanced products, investing in advertising, providing value
added services such as road support, making campaign-promotions regularly and having environment friendly
products.
The importance of the selected criteria was assessed by the three decision makers to identify the
weights of each criteria by using AHP method. The scale was used for AHP method is given in Table 5.
Table 5. Scale used in Stage 2
Value (aij) Definition
1 i and j are equally important
3 i is weakly more important than j
5 i is strongly more important than j
7 i is very strongly more important than j
9 i is extremely more important than j
2, 4, 6, 8 Intermediate values (Used when a compromise is needed.)

The weights of these criteria are given in Table 6.

Table 6. Weights of the criteria used in Stage 2

Traditional Criteria Weights
being a reliable brand 0.05
being a brand used before 0.04
being a recommended brand 0.05
high quality 0.05
being enduring 0.05
high profitability 0.04
can be found easily/high availability 0.11
having different tires suitable for different weather conditions 0.06
being a foreign brand 0.03
being affordable 0.12
being worth for the money paid 0.10
being a domestic brand 0.03
having good price- performance balanced products 0.08
investing in advertising 0.04
providing value added services such as road support 0.07
making campaign-promotions regularly 0.04
having environment friendly products 0.04

TOPSIS method is applied as a compromise ranking method to decide the best supplier. As a result of TOPSIS
method, ranking of suppliers are given in the Table 7. As a result, S6 was chosen as the best supplier.

345
 Table 7. Results of TOPSIS

Suppliers Ranking
S6 1
S7 8
S9 7
S13 9
S20 4
S25 2
S26 10
S31 6
S39 3
S40 5
S41 12
S50 11
S54 14
S55 13

CONCLUSION

Today, myriad of companies work with different suppliers from different regions. Therefore, both the
scope and complexity of the supplier evaluation and selection problem is higher than ever. Especially, the
industries in which there are many alternative suppliers struggle in selecting the most appropriate supplier
among a large set of suppliers. Therefore, this study proposed a two-stage approach for the supplier evaluation
and selection problem in this risky, volatile and dynamic business environment.
The proposed model was applied on a data set gathered from the tire industry. First, a set of risk criteria
was identified based on the literature review. 64 suppliers were assessed based on the identified risk criteria.
Suppliers were grouped into two clusters based on their risk profile by using the well-known K-means
clustering algorithm. Then, the low-risk profile supplier group was examined for further analysis. These 14
suppliers with lower risk levels were evaluated based on 17 traditional supplier evaluation criteria. Weighted
supplier evaluation data was used to identify the best supplier by using the TOPSIS method.
The proposed integrated approach will increase the performance of the supplier selection process and
mitigate the supplier-related risk factors at the beginning of the procurement phase. Additionally, elimination of
the risky suppliers will decrease the load of supplier selection phase and reduce the time required to assess
suppliers.

346
 REFERENCES

Alkan, A., Kasımoğlu, H. Ç., Çelik, C., & Aladağ, Z. (2017). AHP ve PROMETHEE yöntemleri ile lastik üreticisi bir
firma için tedarikçi seçimi. Sakarya Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 261-269.
Altın, M., Koca, A., Solmaz, H., & Yılmaz, E. (2013). Investgation of the Tyre Waste from Automobiles in Turkey.
Journal of Polytechnic, 51-56.
Amini, S., & Asoodar, M. A. (2016). Selecting the most appropriate tractor using Analytic Hierarchy Process–An Iranian
case study. Information processing in agriculture, 3(4), 223-234.
Chang, B., Chang, C. W., & Wu, C. H. (2011). Fuzzy DEMATEL method for developing supplier selection criteria.
Expert systems with Applications, 38(3), 1850-1858.
Dymova, L., Sevastjanov, P., & Tikhonenko, A. (2013). An approach to generalization of fuzzy TOPSIS
method. Information Sciences, 238, 149-162.
Er Kara, M., & Oktay Fırat, S. Ü. (2018). Supplier risk assessment based on best-worst method and K-means clustering:
a case study. Sustainability, 10(4), 1066.
Er Kara, M., Oktay Fırat, S.Ü. (2015). Tedarik Zinciri Risk Yönetiminin Gelişmesini Tetikleyen Risk Olayları Üzerine
bir İnceleme. IV. Ulusal Lojistik ve Tedarik Zinciri Kongresi, Gümüşhane, Türkiye, 21-23 Mayıs 2015, pp. 326-334.
Hamdi, F., Ghorbel, A., Masmoudi, F., Dupont, L. (2015) Optimization of a supply portfolio in the context of supply
chain risk management: literature review. Journal of Intelligent Manufacturing, DOI 10.1007/s10845-015-1128-3.
Hantsch, M., & Huchzermeier, A. (2013). Identifying, analyzing, and assessing risk in the strategic planning of a
production network: the practical view of a German car manufacturer. Journal of Management Control, 24, 125-158.
Haq, A. N., & Kannan, G. (2006). Fuzzy analytical hierarchy process for evaluating and selecting a vendor in a supply
chain model. The International Journal of Advanced Manufacturing Technology, 29(7-8), 826-835.
Kara, M. E. (2017). Data mining approach in supply chain risk management, (Doctoral dissertation, Marmara
University).
Kuo, T.C., Hsu, C.W., Li, J.Y. (2015) Developing a Green Supplier Selection Model by Using the DANP with VIKOR.
Sustainability, 7, 1661-1689.
Mardani, A., Jusoh, A., Nor, K., Khalifah, Z., Zakwan, N., & Valipour, A. (2015). Multiple criteria decision-making
techniques and their applications–a review of the literature from 2000 to 2014. Economic Research-Ekonomska
Istraživanja, 28(1), 516-571.
Moghaddam, K. S. (2015). Fuzzy multi-objective model for supplier selection and order allocation in reverse logistics
systems under supply and demand uncertainty. Expert Systems with Applications, 42(15-16), 6237-6254.
Nazari-Shirkouhi, S., Shakouri, H., Javadi, B., & Keramati, A. (2013). Supplier selection and order allocation problem
using a two-phase fuzzy multi-objective linear programming. Applied Mathematical Modelling, 37(22), 9308-9323.
Parmar, D., Wu, T., Callarman, T., Fowler, J., & Wolfe, P. (2010). A clustering algorithm for supplier base management.
International Journal of Production Research, 48(13), 3803-3821.
Renganath, K., & Suresh, M. (2016, December). Supplier selection using fuzzy MCDM techniques: A literature review.
In 2016 IEEE International Conference on Computational Intelligence and Computing Research (ICCIC) (pp. 1-6). IEEE.
Rezaei, J., Nispeling, T., Sarkis, J., Tavasszy, L. (2016) A supplier selection life cycle approach integrating traditional
and environmental criteria using the best worst method. Journal of Cleaner Production, 135, 577-588.
Schramm, V.B., Cabral, L.P.B., Schramm, F. (2020). Approaches for supporting sustainable supplier selection- A
literature review. Journal of Cleaner Production, 273.
Supçiller, A. A., & Çapraz, O. (2011). AHP-TOPSIS YÖNTEMİNE DAYALI TEDARİKÇİ SEÇİMİ UYGULAMASI.
Istanbul University Econometrics & Statistics e-Journal, 13(1).
Tavana, M., Yazdani, M., & Di Caprio, D. (2017). An application of an integrated ANP–QFD framework for sustainable
supplier selection. International Journal of Logistics Research and Applications, 20(3), 254-275.
Witten,I.H.;Frank,E.;Hall,M.A.;Pal,C.J. (2017) Data Mining: Practical Machine Learning Tools and Techniques, 4th ed.;
Morgan Kaufmann: Burlington, MA, USA, ISBN 978-0-12-804291-5.

347
 ANALYZING THE OHS RISKS EMERGED IN TRANSPORTATION OF
MEDICAL MATERIALS IN THE COVID-19 PANDEMIC
Aylin Adem1,Burcu Yılmaz Kaya2, Metin Dağdeviren3

Abstract ⎯ With the sudden outbreak of the COVID-19 pandemic about one and a half years ago, it is possible
to say that logistic activities about medical materials and equipment (masks, disinfectants, vaccines,
protective equipment, etc.,) have increased rapidly. In all logistics activities (including reverse logistics
related), there are some kinds of occupational health and safety-related risk factors (such as static working
position or static holding). However, these kinds of risk factors can differentiate in the transportation of
medical materials which are special for the COVID-19 pandemic. Within the scope of this study, it is aimed to
determine and analyze the occupational health and safety risk factors that can be emerged during all logistic
activities of pandemic medical materials under the fuzzy environment. In this analysis, the spherical fuzzy
extension of the Analytical Hierarchy Process technique, which is one of the most frequently utilized Multi-
Criteria Decision Making methods is employed. The reason for the utilization of the spherical fuzzy numbers
in the analyzing stage of this study is that they focus on the hesitancy degrees of decision-makers and they
allow assessments from a broader perspective in the decision-making process.

Keywords ⎯ COVID-19 pandemic, decision making, logistic activities, occupational health and safety,
spherical fuzzy numbers

INTRODUCTION

All kinds of logistics activities gain importance in cases such as natural disasters, endemics or pandemics that
affects the whole world such as COVID-19 (Mondal and Roy, 2021). In particular, providing health-related
materials to the regions where they are needed on time plays a critical role in the sustainability of health
services. This critical role has also been contributing to in making this subject a place in the literature under
the title of humanitarian aid logistics. However, COVID-19 has drawn a quite different structure, showing a
difference from previous natural disasters in terms of both the speed of emergence of the need for the
demanded products and the multiplicity of demand, as well as the way the disease spread (Kwon 2020). On
the other hand, the COVID-19 pandemic, which is currently on the agenda of Turkey and the world, has
increased the need for healthcare supplies, especially those that develop suddenly and originate from the
pandemic. This increase showed itself with the concentration in logistics activities. From another point of
view, while remote working models can be operated in almost all business lines, such a remote working model
could not be operated in logistics and especially in activities related to the transportation of pandemic-sourced
health material.

Since it is important to ensure the sustainability of the provided services in this area and to take the pandemic
under control, remote working models cannot be operated in these activities. Employees serving in this field
continued to work in the old order, even more intensively. Occupational Health and Safety (OHS) issues have
become much more important in an environment where there is a contagious disease. It would not be wrong to
say that the risk parameters brought by the epidemic disease are added to the existing OHS risks. On the other
hand, Zhang (2021) determined that the occupational group that got the most diseases after healthcare workers
with COVID-19 disease was those working on transportation and material moving. Thus, it is important to
investigate and analyze what these risks might be for logistics workers, who are among the occupational group
which is one of the most exposed to COVID- 19 related risks.

This study focuses on the determination and analysis of the possible OHS risks that may arise during the
forward and reverse logistics of health materials, which continues under the pandemic conditions affecting all
segments of the society. First of all, a deep literature survey was conducted to determine the possible OHS

1
Aylin Adem, Gazi University, Engineering Faculty, Department of Industrial Engineering, Ankara, Turkey, [email protected]
2
Burcu Yılmaz Kaya Gazi University, burcuyı[email protected]
3
Metin Dağdeviren, Gazi University, [email protected]

348
related risk factors and their sub-factors. In performing the analyze stage, the Analytical Hierarchy Process
(AHP) technique, which is one of the most frequently employed technique in the literature, which is one of the
multi-criteria decision-making techniques that establish the priority relationship between multiple factors was
utilized. The AHP technique was suggested by Saaty with a scale of 1-9 (Saaty 1980). This scale has been
criticized in the related literature for not adequately addressing the uncertainty in the decision making
problems and the data (Şenol 2019). The AHP technique was utilized by integrating with the developed
different fuzzy scales overtime to eliminate these criticisms (Dağdeviren and Yüksel, 2008; Adem and
Dağdeviren, 2016; Dağdeviren and Adem, 2018; Adem et al, 2018; Çolak et al, 2019; Şenol et al 2019; Adem
et al, 2020). In this study, the integration of spherical fuzzy sets which is a newly developed scale with the
AHP method was utilized to analyze the determined OHS risk factors. The reason for the utilization of the
spherical fuzzy numbers in the analyzing stage of this study is that they focus on the hesitancy degrees of
decision-makers and they allow assessments from a broader perspective in the decision-making process. This
study contributes to the literature in two ways: a) There is no previous study in the literature that addresses the
occupational health and safety factors that COVID-19 may create in logistics activities; b) There is no study in
the literature that analyzes this issue.

This paper is organized as follows: Section 2 gives the utilized fuzzy scale and the extension of the AHP
technique. Section 3 focuses on the OHS-related issues which emerge during the logistic activities of COVID-
19 related health materials. Section 4 gives an analysis of the OHS issues by utilizing the spherical fuzzy AHP
technique and presents the results of this analysis and a discussion of the results. Finally, Section 5 gives the
concluding remarks about the paper.

METHOD

By combining Pythagorean and Neutrosophic fuzzy sets, Kutlu Gündoğdu and Kahraman (2019) developed
and presented the Spherical fuzzy (SF) sets to the relevant literature. Despite the fact that it is a relatively new
set in the literature, spherical fuzzy sets have been used in plenty of research (advanced manufacturing system
selection problem- Mathew et al, (2020); petrol station location selection- Ayyıldız and Gumus (2020);
renewable energy location selection- Kutlu Gündogdu and Kahraman (2020); process mining technology
selection -Doğan (2021); distribution center site selection- Kieu et al (2021); supplier selection- Sharaf
(2021)). By focusing on the degrees of hesitation of decision-makers, these sets really help them to assess
decision-making difficulties from a broad viewpoint. The weights of main and sub-OHS factors of the logistic
activities were calculated using the spherical fuzzy extension of AHP (SF-AHP) in this paper. First of all,
theoretical information about spherical fuzzy numbers will be given, and then the calculation steps of the
method will be shown. The spherical fuzzy sets have membership, non-membership, and hesitancy degree
characteristics, just as Pythagorean fuzzy sets.

Let E1 and E2 be two universes. Let As ,and Bs of the universe of discourse E1 and E2 be as follows:
As  {x,( A ( x), vA ( x),  A ( x)) | x  E1} where
s s s

 A ( x) : E1  [0,1], vA ( x) : E1  [0,1],  A ( x) : E1  [0,1] and

s s s

0 2
As
( x)  v 2
As
( x)   2
As
( x)  1 x  E1
For each x,  A ( x) shows the membership function, vA ( x) represents the non-membership function,  A ( x)
s s s

shows the hesitancy degree.

Likewise, Bs  { y,(A ( y), vA ( y),  A ( y)) | y  E2 }

s s s

where Bs ( y) : E2  [0,1], vBs ( y) : E2  [0,1],  Bs ( y) : E2  [0,1] and

349
0   2 Bs ( y)  v 2 Bs ( y)   2 Bs ( y)  1 y  E2

For each y,  Bs ( y ) represents the membership function, vB ( y) shows the non-membership function,  Bs ( y )
s

shows the hesitancy degree (Kutlu Gündoğdu and Kahraman, 2020).

The following formulations show some of the arithmetical operations of the SF sets (Kutlu Gündoğdu and
Kahraman, (2019)).

As  Bs  {( 2A   B2   A2  B2 )1/2 , vA vBs ,((1   B2 ) 2A  (1   2A ) B2   A2  B2 )1/2 } (1)

s s s s s s s s s s s

As  Bs  { A  B ,(v2A  vB2  v2A v2B )1/2 ,((1  v2B ) 2A  (1  v2A ) B2   2A  B2 )1/2 } (2)
s s s s s s s s s s s s

k * As  {(1  (1   2A )k )1/ 2 , vk A ,((1   A2 )k  (1   A2   A2 )k )1/ 2} (3)

s s s s s

In this section, the steps of SF-AHP method were provided. These steps were adapted from Kutlu Gündoğdu

& Kahraman, (2020):

Step 1. Draw the borders of the decision problem (determine criteria, sub-criteria and if exist alternatives)

Step 2. Take pairwise comparisons of the elements of decision problem from experts.

Table 1. Linguistic scale for pairwise comparisons

Linguistic Expressions  , v,  Score Index (SI)
Absolutely more importance (AMI) (0.90.1,0.0) 9
Very high importance(VHI) (0.8,0.2,0.1) 7
High importance (HI) (0.7,0.3,0.2) 5
Slightly more importance(SMI) (0.6,0.4,0.3) 3
Equally importance(EI) (0.5,0.4,0.4) 1
Slightly low importance (SLI) (0.4,0.6,0.3) 1/3
Low importance(LI) (0.3,0.7,0.2) 1/5
Very low importance (VLI) (0.2,0.8,0.1) 1/7
Absolutely low importance (ALI) (0.1,0.9,0.0) 1/9

Table 1 shows the linguistic scale for pairwise comparisons. After taking experts evaluations, each linguistic
expression needed to be convert into score index in order to check the consistency ratio of decision matrices.

To calculate the score index of AMI, VHI, HI, SMI, the following formula is utilized.

SI  | 100* (  A   A )2  (vA   A ) 2  | (4)

 s s s s 

To calculate the score index of ALI, VLI, LI, SLI, the following formula is utilized.

1 1
 (5)
SI | 100* (  A   A ) 2  (vA   A ) 2  |
 s s s s 

350
Step 3. Calculate the consistency ratio of pairwise comparison matrices.
The consistency ratio calculation stages of the traditional AHP approach were implemented after determining
the score index of each element in the pairwise comparison matrices. As in the original AHP technique
(Dağdeviren et al, 2009), the acceptable consistency ratio limit is 10%.

Step 4. Compute the fuzzy weights decision making elements.

In order to calculate the fuzzy weights, the Spherical Weighted Arithmetic Mean (SWAM) is employed:

SWAM w ( AS 1 ,...., ASn )  w1 AS 1  w2 AS 2  ....  wn ASn

1/2 1/2
 n
 n
 n n
 (6)
 1   (1   2A ) wi  ,  v A wi ,  (1   2A ) wi   (1   2A   2A ) wi 
 i 1 si
 i 1
si
 i 1 si
i 1
si si


Where w=1/n.

Step 5. Defuzzify the fuzzy weights

For defuzzifing process, the following formula is utilized:

 A vA 
S ( wsj )  | 100* (3 A  s ) 2  ( s   A ) 2  | (7)
 s
2 2 s


The calculated crisp weights are normalized by utilizing (8):

S ( wsj )
wsj  n
(8)
 S (w )
j 1
s
j

Figure 1 shows the followed path of the paper which has mainly four sub-stages. In the first stage of the paper,
in order to determine OHS-related risk factors, deep literature survey is conducted. After determining OHS-
related risk factors, they are classified according to their attributes. After specifying the decision hierarchy, it
needed to be compute the local and global weights to figure out the priority relationship between the factors.
Finally, in the last stage of the proposed approach, the obtained results are interpreted.

Preliminary Stage Classification Stage of Prioritization stage Concluding Stage

•Conduct deep OHS related issues •Calculate the weights •Interprete the gained
literature review •Classify the of the main and sub- results
determined OHS OHS related factors
related issues by utilizing spherical
according to their fuzzy AHP technique
characteristics

Figure 1. The followed path

351
 OCCUPATIONAL HEALTH AND SAFETY ISSUES IN LOGISTIC ACTIVITIES

In this section, Occupational Health and Safety (OHS) factors that may arise during increased logistics
activities (forward or backward) with COVID-19 are analyzed. It is possible to deal with OHS factors under
two headings as occupational safety and occupational health. The factors discussed here are firstly evaluated
under two main headings regarding safety and health, and then, each of them is evaluated under its own
subheadings. Figure 2 shows how OHS risks are classified.

Figure 2. The classification of OHS related risk factors

Traffic accident, worker fall and material fall are determined as the sub-factors of the safety factors (Akan and
Dalbay, 2015). Obviously, one of the most critical risks that may arise in this process can be considered as
getting infected with COVID-19. Because while remote working models are operated in many business lines,
studies in this field have continued with more intensity. The fact that employees operating in this sector
continue to work full-time and actively increases the possibility of catching this respiratory-borne disease.

From another point of view, it has caused everyone in the logistics sector to work harder in order to keep up
with the suddenly developing and increasing demands. This intense work will cause an increase in static
sitting and a decrease in physical activity in drivers. These two conditions should be considered as serious risk
factors since it is known that they bring many discomforts in terms of occupational health and safety.

Working at a more intense pace than normal, while there is a pandemic, will also cause some psychological
problems in people. These disorders are listed under mental factors. Working hours are increased in order to
respond quickly to the increasing demand, which can lead to consequences such as intense stress and loss of
motivation.

352
Since people working in logistics activities do not take into account OHS training from time to time, the fact
that these people do not know how to use personal protective equipment poses a risk in terms of catching and
spreading the disease (Carlsten et al, 2021). On the other hand, it may not be possible to provide sufficient
social distance or open space during the transportation of materials and loading them into logistics vehicles.
The collection and disposal of waste materials (vaccine waste, masks, visors, safety glasses, and other
equipment) increases the risk of people becoming infected (Das et al, 2021).

Since it is known that COVID- 19 is transmitted between people through respiration, there are some studies
that suggest that the use of masks or even the use of visors in closed areas may cause negative effects on eye
health (there are concerns that the mask and visor block the visual field and cause disturbances in perception)
(Galleghar et al, 2021).

Another quite different OHS-related factor is the headache that can occur due to the heavy use of personal
protective equipment (Ong, 2021). The manifestation of these short-term pains like migraine attacks is one of
the health factors that should be considered due to COVID-19.

To be protected from COVID- 19, frequent hand washing or the use of alcohol-based disinfectants are
required. The use of disinfectants, frequent hand washing, and frequent use of alcohol-based cleaning products
such as cologne cause cracks and splits in people's hands, causing discomfort such as eczema (Burns et al,
2020). In short, experiencing skin damage can be counted as one of the occupational health and safety risks
caused by COVID- 19. It is possible to call these dermatological problems (Rachidi et al, 2021)

RESULTS AND DISCUSSION

In this part of the paper, it has been studied on the prioritization and analysis of the main and sub-factors. By
employing the linguistic scale which is provided in Table 1, the expert group was asked to compare the factors
at each level with the pairwise comparison approach. The consistency ratios of the compromised evaluation
matrices of the experts were checked and it was determined that all matrices were consistent.

For all evaluation matrices, consistency ratios were determined and are presented below the related table with
the CR acronym. Equation (6), the SWAM operator, was used in the calculation stages of spherical fuzzy
weights of factors and their sub-factors. Equation. (7) was used to deffuzzify the fuzzy weights after obtaining
spherical fuzzy weights. Finally, the acquired weights were normalized using (8). All evaluation matrices, CR
values, spherical fuzzy weights, crisp weights are given in appendices (Appendices 1). Table 2 shows the local
and global weights of OHS related risk factors in terms of their level (Level-1, Level-2, Level-3). Each level
was evaluated in itself and the interpretations were conducted according to this order.

According to the evaluation results of Level-1, the weights of the health factors were determined as 0.610;
while the weights of the safety factors were determined as 0.390. Accordingly, it is possible to say that health-
related factors are more important than safety-related factors in this process. Physical factors, mental factors,
traffic accident, worker fall, and material fall are the factors of Level-2. According to the global weights of
these factors, the most important factor was computed as physical factors (wg=0.340). After that, the mental
factors, which is one of the health factors were ranked (wg =0.271). Worker fall which is one of the safety
factors was ranked in the last order (wg =0.102) in Level-2.

353
 Table 2. Local and Global Weights of OHS Related Risk Factors in terms of Their Level

Level-1 Level-2 Level-3

Health
factors Local Global
(0,610) weights weights
Physical Local Global
factors 0,557 0,340 weights weights
Headache 0,147 0,050
Derm. problems 0,139 0,047
Eye-health problems 0,102 0,035
Covid-19 infection 0,197 0,067
Bad-posture 0,169 0,057
Low physical activity 0,117 0,040
Extreme exhaustion 0,129 0,044

Mental
factors 0,443 0,271
Loss of motivation 0,282 0,076
Stress 0,233 0,063
Distraction 0,183 0,049
Anxiety 0,302 0,082

Safety
factors
(0,390)
Traffic
accident 0,409 0,159
Worker fall 0,261 0,102
Material fall 0,331 0,129

In the last level of decision hierarchy (ie Level-3), in the evaluation of physical factors within themselves (ie
according to their local weights), being infected with COVID-19 has the highest value among the local
weights (wl=0.197). However, with the involvement of the mental factors, it can be said that the most critical
sub-factor among the global weights of the factors in Level-3 is anxiety (wg =0.082). The factor of being
infected with COVID-19 is followed by bad or poor posture among the physical risk factors. The first two
major mental factors are anxiety and a lack of motivation. Anxiety, motivation loss, and COVID-19 infection
were judged to be the first three most essential variables at the third level, based on their global weights. In
Level-3, issues with the eyes were ranked in the last order.

CONCLUSION

It is known that COVID-19 has a very intense effect on critical areas of our lives such as education,
purchasing and working models. In many respects, it is possible to say that it is almost impossible for us
to return to our old normal period (the period when we never encountered COVID-19). In order to
remove this disease and to prevent the spread of the disease, the logistics activities of health materials
related to the disease have also become quite frequent. While the option of working remotely is possible
in many other business lines, such an option does not exist for people working in this field. In order to
ensure the sustainability of the provided service, the health status of the workers working in this field
should also be kept under control.

This study aims to determine and prioritize the OHS factors that may arise in logistics activities.
Obviously, the presented OHS factors here are the OHS factors that emerged with COVID-19. A deep
literature surveys and brainstorming technique were utilized to identify and classify the OHS factors. In

354
the evaluation stage of the determined OHS factors and their sub-factors, the SF-AHP technique, which is
one of the accepted methods in the literature although it is a relatively newly developed technique that
takes into account the hesitancy degree of the decision-makers, was employed.

According to the results, it was determined that health-related factors are more critical than safety-related
factors in OHS factors expressed at 3 levels. In terms of health-related factors, it was concluded that
physical factors are more important than mental factors. Moreover, it was concluded that the most critical
factor among physical factors is being infected with COVID-19. Among the physical factors, bad or poor
posture is the factor that comes after the factor of being infected with COVID-19. Anxiety and loss of
motivation are the first two important mental factors. In terms of the global weights of the factors at the
3rd level, anxiety, loss of motivation and COVID-19 infection were determined as the first 3 most critical
factors. Eye health problems were ranked in the last order in Level-3. With this study, it is aimed to
conduct research and analysis on the changing OHS factors with the effect of COVID-19 in classical
logistics activities. According to the results of the analysis, a number of evaluations can be considered
regarding the measures to be taken.

In future studies, it may be possible to integrate different linguistic scales and different MCDM
techniques and compare the obtained results. In this study, the compromised decision matrices of the
expert group were used. However, group decision-making techniques can be utilized in future studies.
The weights of the evaluations conducted by the experts according to their level of expertise can be taken
differently.

Appendices1.
Evaluation Matrices, Related Spherical Fuzzy Weights and Crisp weights

Health Safety Spherical Crisp

factors factors Fuzzy Weights weights
Health factors EI HI (0.62,0.35,0.30) 0,610
Safety factors LI EI (0.42,0.53,0.33) 0,390

Physical Mental Spherical Crisp

factors factors Fuzzy Weights weights
Physical factors EI SMI (0.55,0.40,0.35) 0,557
Mental factors SLI EI (0.45,0.49,0.36) 0,443

Traffic Worker Material Spherical Fuzzy Crisp

accident fall fall Weights weights
Traffic accident EI HI SMI (0.61,0.36,0.30) 0,409
Worker fall LI EI SLI (0.41,0.55,0.32) 0,261
Material fall SLI SMI EI (0.51,0.46,0.34) 0,331
CR=0,033

Loss of Spherical Fuzzy Crisp

motivation Stress Distraction Anxiety Weights weights
Loss of motivation EI SMI HI SLI (0.57,0.41,0.30) 0,282
Stress SLI EI SMI SLI (0.48,0.49,0.33) 0,233
Distraction LI SLI EI LI (0.39,0.59,0.30) 0,183
Anxiety SMI SMI HI EI (0.61,0.37,0.30) 0,302
CR=0.074

355
 Eye- Covid-
Head Derm. health 19 Bad- Low Extreme Spherical Crisp
ache problems problems infection posture phy.act exhaustion Fuzzy Weights weights
Headache EI SMI SMI LI SLI SMI SMI (0.53,0.46,0.31) 0,147
Derm. prob. SLI EI SMI LI SLI SMI SMI (0.51,0.49,0.31) 0,139
Eye-health
problems SLI SLI EI VLI LI SLI SLI (0.38,0.60,0.29) 0,102
Covid-19
infection HI HI VHI EI SMI HI HI (0.69,0.31,0.23) 0,197
Bad-posture SMI SMI HI SLI EI HI SMI (0.60,0.39,0.29) 0,169
Low phy. Act. SLI SLI SMI LI LI EI SLI (0.43,0.56,0.30) 0,117
Extreme
exhaustion SLI SLI SMI LI SLI SMI EI (0.48,0.52,0.31) 0,129

CR=0.087

356
 REFERENCES

Adem, A., & Dağdeviren, M. (2016). A life insurance policy selection via hesitant fuzzy linguistic decision making
model. Procedia Computer Science, 102, 398-405.

Adem, A., Çakit, E., & Dağdeviren, M. (2020). Occupational health and safety risk assessment in the domain of Industry
4.0. SN Applied Sciences, 2(5), 1-6.

Adem, A., Çolak, A., & Dağdeviren, M. (2018). An integrated model using SWOT analysis and Hesitant fuzzy linguistic
term set for evaluation occupational safety risks in life cycle of wind turbine. Safety science, 106, 184-190.

Akan E., & Dalbay, N. (2015) Lojistik Sektöründe Kalite Ve İş Sağliği-Güvenliği Yönetim Sistemleri: Bir
Araştirma. Beykoz Akademi Dergisi, 3(1), 21-66.

Ayyildiz, E., & Gumus, A. T. (2020). A novel spherical fuzzy AHP-integrated spherical WASPAS methodology for
petrol station location selection problem: a real case study for İstanbul. Environmental Science and Pollution
Research, 27(29), 36109-36120.

Burns, E. S., Pathmarajah, P., & Muralidharan, V. (2021). Physical and psychological impacts of handwashing and
personal protective equipment usage in the COVID‐19 pandemic: A UK based cross‐sectional analysis of healthcare
workers. Dermatologic therapy, e14885.

Carlsten, C., Gulati, M., Hines, S., Rose, C., Scott, K., Tarlo, S. M., ... & de la Hoz, R. E. (2021). COVID‐19 as an
occupational disease. American Journal of Industrial Medicine, 64(4), 227-237.

Çolak, A., Adem, A., & Dağdeviren, M. (2019, July). Fuzzy Prioritization of Factors Affecting Employer Branding for
Employees. In International Conference on Intelligent and Fuzzy Systems (pp. 852-858). Springer, Cham.

Dagdeviren, M., & Adem, A. (2018, August). An Integrated MCDM Approach to the Problem of New Graduate Job
Selection Under Fuzzy Environment. In International Conference on Theory and Applications of Fuzzy Systems and Soft
Computing (pp. 951-957). Springer, Cham

Dağdeviren, M., & Yüksel, İ. (2008). Developing a fuzzy analytic hierarchy process (AHP) model for behavior-based
safety management. Information sciences, 178(6), 1717-1733.

Dağdeviren, M., Yavuz, S., & Kılınç, N. (2009). Weapon selection using the AHP and TOPSIS methods under fuzzy
environment. Expert systems with Applications, 36(4), 8143-8151.

Das, A. K., Islam, M. N., Billah, M. M., & Sarker, A. (2021). COVID-19 and municipal solid waste (MSW)
management: a review. Environmental Science and Pollution Research, 1-16.

Dogan, O. (2021). Process mining technology selection with spherical fuzzy AHP and sensitivity analysis. Expert
Systems with Applications, 178, 114999.

Gallagher, S., Clasing, J., Hall, E., Hammond, S., Howard, G., Mohrmann, T., & Vaccaro, J. T. (2021). Eye Health,
COVID-19, and the Occupational Health Professional: Round Table. Workplace Health & Safety, 69(8), 352-358.

Kieu, P. T., Nguyen, V. T., Nguyen, V. T., & Ho, T. P. (2021). A Spherical Fuzzy Analytic Hierarchy Process (SF-AHP)
and Combined Compromise Solution (CoCoSo) Algorithm in Distribution Center Location Selection: A Case Study in
Agricultural Supply Chain. Axioms, 10(2), 53.

Kutlu Gündoğdu, F., & Kahraman, C. (2019). Spherical fuzzy sets and spherical fuzzy TOPSIS method. Journal of
intelligent & fuzzy systems, 36(1), 337-352.

Kutlu Gündoğdu, F. & Kahraman, C. (2020). A novel spherical fuzzy analytic hierarchy process and its renewable
energy application. Soft Computing, 24(6), 4607-4621.

Kwon, O. K. (2020). How is the COVID-19 pandemic affecting global supply chains, logistics, and
transportation? Journal of International Logistics and Trade, 18(3), 107-111.

357
Mathew, M., Chakrabortty, R. K., & Ryan, M. J. (2020). A novel approach integrating AHP and TOPSIS under spherical
fuzzy sets for advanced manufacturing system selection. Engineering Applications of Artificial Intelligence, 96, 103988.

Mondal, A., & Roy, S. K. (2021). Multi-objective sustainable opened-and closed-loop supply chain under mixed
uncertainty during COVID-19 pandemic situation. Computers & Industrial Engineering, 159, 107453.

Ong, J. J., Chan, A. C., Bharatendu, C., Teoh, H. L., Chan, Y. C., & Sharma, V. K. (2021). Headache Related to PPE Use
during the COVID-19 Pandemic. Current Pain and Headache Reports, 25(8), 1-11.

Rachidi, H., Hamdaoui, S., Merimi, I., Bengourram, J., & Latrache, H. (2021). COVID-19: unbalanced management of
occupational risks—case of the analysis of the chemical risk related to the use of disinfectants in the dairy industry in
Morocco. Environmental Science and Pollution Research, 1-13.

Saaty, T. L. (1990). How to make a decision: the analytic hierarchy process. European journal of operational
research, 48(1), 9-26.

Şenol, M.B, Adem, A., & Dağdeviren M. (2019). A fuzzy MCDM approach to determine the most influential logistic
factors. Politeknik Dergisi, 22(3), 793-800.

Sharaf, I. M. (2021). Global Supplier Selection with Spherical Fuzzy Analytic Hierarchy Process. Decision Making with
Spherical Fuzzy Sets (pp. 323-348). Springer, Cham.

Zhang, M. (2021). Estimation of differential occupational risk of COVID‐19 by comparing risk factors with case data by
occupational group. American journal of industrial medicine, 64(1), 39-47.

358
 RESTRICTED DYNAMIC PROGRAMMING APPROACH FOR SOLVING
STOCHASTIC LOT-SIZING PROBLEMS
Dr. Mustafa Çimen1

Abstract ⎯ Despite the long history of the research on inventory optimization, the field is still attracting a
significant amount of academic interest. This is partially because that lot-sizing decisions relate with an
enormous number of worldwide logistics activities, and partially because the uncertain nature and complex
components of many problems in the field, which remain to be solved. Moreover, exact solution approaches
developed for many problems require long computation times, high-quality computing equipment and/or
commercial software. As a result, the needs of decision makers in practice for decision support tools that provide
close-to-optimal results using a feasible amount of computation time/effort motivate many researchers to
develop heuristic algorithms with lower computation requirements. With a similar motivation, in this research,
we provide an adaptation of Restricted Dynamic Programming, an existing heuristic approach known to reduce
the computation and memory requirements of the classical Dynamic Programming method, for solving single
product lot-sizing problems with stochastic demand. To the best of our knowledge, this is the first attempt to
adapt Restricted Dynamic Programming approach to stochastic lot-sizing problems. The added value of the
proposed algorithm is shown through a numerical comparison with the results obtained by the classical
Dynamic Programming method.

Keywords ⎯ decision support tool, lot-sizing, Restricted Dynamic Programming, stochastic demand

INTRODUCTION
Lot-sizing decisions in inventory management processes has been of academic interest for a long time. Yet,
only a fraction of uncertain and/or dynamic aspects and a large number of different assumptions of the lot-sizing
problem that are confronted in real-life cases could have been solved in the literature. One of the main issues
that challenges the researchers is the limited amounts of resources of the decision-makers, particularly in small
and medium sized enterprises (SMEs), in terms of time, money and computational equipment. In order for the
researches on lot-sizing decision support tools to have a larger impact on businesses, there is a need for
lightweight models that provide promising results with small computational requirements.
One of the main solution approaches that is widely addressed in lot-sizing literature is Dynamic
Programming (DP) approach. Brahimi et al (2006) state that DP is the most adequate approach for solving
polynomially solvable capacitated single-item lot sizing problems. Wagner and Whitin (1958) may be regarded
as the first to propose a DP based approach to optimize lot-sizing decisions. Since then, many studies employed
DP algorithms to solve various decisions in inventory management processes (e.g., Chen et al, 1994;
Jaruphongsa et al, 2004; Moqri et al, 2011; Çimen and Belbağ, 2016, etc.). However, despite its advantages
particularly in discrete and dynamic problems, classical DP approach carries the curses of dimensionality with
it, that refers to exponentially increasing computation time and memory requirements with respect to increases
in sizes of state, space and outcome spaces. The high requirements render the classical DP approach infeasible
for many decision makers in real-life problems.
In order to obtain feasible and promising decisions, several researchers turned towards DP heuristic
approaches instead of the classical DP approach. The studies by Absi et al (2013), Kang and Lee (2013) or
Çimen and Kirkbride (2017) are a few examples to such attempts. A more recent example is presented by Lin
et al (2021) on a multi-product, multi-supplier deterministic lot-sizing problem, for which a decomposition-
based heuristic is introduced. Dynamic programming recursions are formulated for constructing an aggregate
decision set based on results of pre-solved sub-problems. Another example worth to mention is published by
Visentin et al (2021), who generates and manages smaller sub-problems using a branch-and-bound approach,
and solves each sub problem with a DP recursion for solving a single product stochastic lot-sizing problem
under (R, s, S) control policy. The existence of such recent studies, in other words, the ongoing search for DP
1 Hacettepe University, Department of Business Administration, Management Science. Ankara, Turkey.

E-mail: [email protected].

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

359
based heuristic algorithms for stochastic -or even deterministic- lot-sizing problems, allude to the still
unsatisfied need for such decision support tools.
In this research, we will introduce a Restricted Dynamic Programming (RDP) based solution approach for
discrete, stochastic lot-sizing problems. RDP approach is based on the idea of keeping track of a fraction of the
states in each iteration, rather than all states. Despite the fact that more promising states/actions may be
neglected, this approach has potential to reduce computation time and memory requirements according to the
needs of the decision maker, and may still provide close-to-optimal decisions.
The RDP algorithm proposed in this study is based on the RDP approach introduced by Malandraki and
Dial (1996) for a time-dependent travelling salesman problem. Following Malandraki and Dial, their RDP
approach has been adopted to a number of routing-based problems in several studies (e.g., Gromicho et al, 2012;
Sun et al, 2015; Soysal and Çimen, 2017, etc.). RDP approaches has also been used in several other fields than
routing, such as minimum spanning tree (e.g., Gouveia et al, 2011) data broadcast scheduling (e.g., Wang and
Chen, 2014), runway scheduling (e.g., Malik and Jung, 2016) or handwriting recognition (e.g., Ikeda et al,
1981). The wide range of uses of the RDP show the potential benefits of the approach. Yet, we could not find
any attempts that employ RDP in a stochastic lot-sizing problem.
The aim of this research is to introduce an RDP based heuristic approach for stochastic lot-sizing problems.
To the best of our knowledge, this is the first attempt to employ an RDP approach in a stochastic inventory
problem. We test our approach on a capacitated single product, multi-period production-inventory problem
under production, holding and penalty costs and infinite planning horizon assumptions. The single dimensional
test problem is selected in order to be able to make a comparison to an exact solution algorithm (classical DP)
and show the strengths and weaknesses of the proposed approach, yet, we expect our research to shed light on
future studies that will employ RDP in multi-dimensional, complicated lot-sizing problems.
The rest of the paper is organized as follows: We will first formulate the addressed problem as a Markovian
Decision Process. Then, the proposed RDP based heuristic algorithm will be introduced, followed by several
numerical examples. Finally, we will conclude our findings and mention several future research directions.

THE PROBLEM AS A MARKOVIAN DECISION PROCESS

We address a single product inventory optimization problem under discrete stochastic demand and infinite
planning horizon assumptions, modelled as a Markovian Decision Process. The inventory system is assumed to
be a make-to-stock, periodic review system, that is, the inventory level is observed and corresponding
procurement decisions are made (once) at the beginning of in each period. There is a capacity limitation on the
inventory amount carried to the following period (inventory held at the end of each period, c), and any excess
inventory that exceeds this amount is assumed to be thrown away. Let denote the state (inventory level) of the
system at the beginning of a period. Then, the state space (I) comprises all possible beginning inventory levels
( = {0,1,2, … , }). At the beginning of each period, a procurement decision (capacitated by a maximum
amount, q) is made based on the observed state ( ∈ ; = {0,1,2, … , }). The procurement is assumed to be
immediate (within a negligible lead time).
The post-decision inventory ( = + ) can be used to serve the demand in that period. Demand (denoted
by D) is uncertain at the time of making the procurement decision, and is assumed to be Poisson with mean ̅
each period (any other discrete probability distribution may also be employed). Despite that there are no upper
bound for Poisson-distributed random variables, for the sake of feasibility we define an upper bound as
the minimum value that satisfies:

∑ ! ( = ) ≥ 0.999 (1)

Once the demand is realized, the products in inventory will be used to meet the demand. If the demand is lower
than the inventory level, then the excess inventory is carried to the following period. If the demand is higher
than the inventory level, any unmet demand is assumed to be lost (no backordering). Hence, the state transition
function is:

← max(min( + − , ) , 0) (2)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

360
 The aim in the problem is to find a lot-sizing plan that minimize the discounted long-term expected
inventory costs, which comprises production, holding and penalty costs. The immediate return/cost of a decision
/
(-. ) can be calculated as:
/
-. = 0 + ℎ max( + − , 0) + π max( − − , 0) (3)

where u, h and π are unit production, holding and penalty costs respectively. Yet, the demand is uncertain, which
leads to the expectation calculation of the immediate return:

/ 567( ,.8/)
3(-. ) = 0 + ℎ 4∑ ! ( = )( + − )9 + π 4∑ .8/ ( = ) ( − − )9 (4)

and the Bellman recursion:

/
:( ) ← min;3<-. = + ><∑ ! ( = ):(max(min( + − , ) , 0))=? (5)
/

where :( ) is the optimal cost-to-go (value function), and > is a discount rate. Solving this recursion iteratively
until :( ) converges (the classical DP approach) provides the optimal production plan that minimizes the
discounted long-term expected inventory costs (exact solution).

THE RESTRICTED DYNAMIC PROGRAMMING APPROACH

A classical DP algorithm iteration updates the value functions of the entire state space, through calculating
a minimization for each state over all feasible actions and an expectation for each action. This approach of value
function is called a full backup. Here, we will propose a sample backup, which reduce the computational
requirements of the process. The idea is to update a fraction of the value functions in each iteration. Various
approaches may be employed for selecting the subset of states for which the value functions will be updated in
each iteration, yet, in this research we will use a random sampling approach for the sake of simplicity and
applicability.
The proposed restricted dynamic programming algorithm composes three main phases. The first phase is
“data initialization”. Among other parameters (demand distribution, unit costs etc.) to be set, an important value,
the number of states for value function update (s) should be decided on in this phase. Lower sizes of s results in
shorter iterations, but the information to be learnt in each iteration will be more limited, since a limited number
of states may be repeatedly visited. Higher sizes of s may provide a better exploration at a cost of longer
iterations. The decision maker should decide on the value of s based on available computation resources (i.e.,
time and equipment) and magnitude of the need for accuracy of results.
The second and third phases are iteratively repeated. The second phase is composed of selecting @ states
(forming a sample states set, ′) to be visited, and adjusting related data. In this research, we select the states
based on a heuristical assumption that the most frequently visited states (therefore the most important states to
learn about in practice) are the ones around lower levels of inventory. Hence, the states are randomly chosen
using an arbitrary generation process:
B
′= < − ̅= +β∗θ (6)

where β is a binary random variable with equally likely outcomes {-1, 1} and θ is a random variable with
geometric distribution ( ( = F) = (1 − G) H G where G is calculated by:
IK
G = 1 − 0.001 (J 8 ) (7)

Third phase is composed of calculating approximate Bellman recursions using a backwards recursion. For
this purpose, we adopt the exact Bellman recursion provided above as:

/,
:L ( ) ← min 4∑ ;. N 8/O ∈PNN ( = | + − ∈ ′) 4-. + :( + − ))99 (8)
/

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

361
which is calculated for all states, where ′ is the set of selected sample states. This process is called as sample
backup, where using only @ successor states will reduce the computation time. The degree to which reduction
is needed is decided by the decision maker, at a probable cost of accuracy.

NUMERICAL EXAMPLES
Our aim in this section is to explore the applicability and expediency of the proposed approach. RDP is most
suitable and preferable when the problem cannot be solved by the classical DP. Hence, we compare the results
yielded by the RDP approach with those yielded by a single iteration of the DP approach, which are the very
first results that can be obtained by DP with minimum time and memory requirements.
The sample problem solved is a discounted, single product capacitated lot-sizing problem under linear
production, holding and penalty costs, infinite planning horizon and stochastic demand. The demand is assumed
to be Poisson with mean 10,000 units in each period. The production capacity is 10,500 units. Unit production,
holding and penalty costs are 1, 0.1 and 7 units respectively. The discount rate is set to 0.9. The aim is to
minimize the discounted long-term inventory costs.
The problem is first modeled and solved by the classical DP approach. The results obtained in 4225 seconds
show that the DP algorithm suggests a threshold policy, where the order-up-to level is set as close to 10,101
units as possible. Then, the RDP algorithm with @ = 100 (sample size) is employed for a single iteration. The
iteration is completed only in 83 seconds. The results obtained within less than 2% computation time of the DP
iteration generally follows a threshold policy where the order-up-to level is 9962 units. Figure 1 presents the
suggested order-up-to levels of the two algorithms.

11000
10000
9000
8000
7000
6000
5000
4000
3000
RDP
2000
DP
1000

Figure 1. The Comparison of Order-Up-To Levels for Different Beginning States Yielded by Single Iterations
of DP And RDP Approaches.

It can be observed from Figure 1 that the RDP approach generated similar decisions to the classical DP
approach, within less then 2% computation time. The effectiveness of the RDP algorithm is further tested under
different time limitations. Again, the RDP algorithm is beneficial when the classical DP approach is infeasible,
hence, the time limitations are defined as different percentages of the time that a single iteration of DP takes.
Then, the obtained decision sets are compared by a Monte Carlo simulation with 10,000 iterations. Figure 2
presents the results.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

362
 10500 Incurred cost by RDP 3,0% 60 3,0%
number of iterations
% difference from DP 2,5% 50 % difference from DP 2,5%
10400
2,0% 40 2,0%
10300
1,5% 30 1,5%
10200
1,0% 20 1,0%
10100 0,5% 10 0,5%
10000 0,0% 0 0,0%

Figure 2. Performance of the RDP Approach Under Various Time Limitations (As Percentages of the Time
That A Single Iteration of DP Takes.

Figure 2 shows that by the time that a single iteration of DP takes, RDP algorithm @ = 100 can complete
around 50 iterations. Furthermore, despite that the decisions found by a single iteration of RDP (0% time limit)
yields roughly 3% higher costs than that of DP, the gap decreases to 0.5% as the time limit increases higher than
10% (423 seconds).
Additionally, a sensitivity analysis is conducted to reveal the relation between the performance of the RDP
algorithm and the setting of the sample size (@). As stated above, the sample size setting is a choice between
higher number of iterations and higher quality information learnt in each iteration. Figure 3 presents the results
for the different settings of @, again obtained by a Monte Carlo simulation with 10,000 iterations. Note that the
time limit for RDP is set to 25% of the time that a single iteration of DP takes for this analysis.

10260 Incurred cost by RDP 0,8% 200 0,8%

number of iterations
10240 % difference from DP 0,7% % difference from DP
0,7%
0,6% 150 0,6%
10220
0,5% 0,5%
10200 0,4% 100 0,4%
0,3% 0,3%
10180
0,2% 50 0,2%
10160
0,1% 0,1%
10140 0,0% 0 0,0%
10 25 50 100 250 500 1000 10 25 50 100 250 500 1000

Figure 1. The performance of the RDP approach under different sample size settings.

It is apparent in Figure 3 that the best setting of @ for this problem setting is @ = 100, a medium setting
where the number of iterations and the quality of information obtained in each iteration is balanced. Lower
sample sizes allow up to 189 iterations, however, considering only few number of successor states prevent the
algorithm from having good results. On the contrary, high sample sizes reduce the number of iterations up to 2,
which eliminates the advantages of the RDP algorithm against the classical DP approach.

CONCLUSION
In this research, we present an adaptation of the RDP algorithm for optimizing lot-sizing decisions in
stochastic inventory problems. Even though our test bed is a classical lot-sizing problem, the proposed algorithm
is flexible to be adopted to other problem variants. The promising results obtained from the numerical study
encourages the use of the algorithm as a decision support tool for both academicians and practitioners.
A natural future research direction following this research is adopting the RDP algorithm to more complex
problems, such as multiple product, multiple facility settings. Incorporating various assumptions into the
algorithm, such as service level, deterioration or stochastic supply is also worth to investigate in future studies.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

363
 REFERENCES

Absi, N., Detienne, B., & Dauzère-Pérès, S. (2013). Heuristics for the multi-item capacitated lot-sizing problem with lost
sales. Computers & Operations Research, 40(1), 264-272.
Brahimi, N., Dauzere-Peres, S., Najid, N. M., & Nordli, A. (2006). Single item lot sizing problems. European Journal of
Operational Research, 168(1), 1-16.
Chen, H. D., Hearn, D. W., & Lee, C. Y. (1994). A new dynamic programming algorithm for the single item capacitated
dynamic lot size model. Journal of Global Optimization, 4(3), 285-300.
Çimen, M., & Belbağ, S. (2016). Süreç esnekliği varsayımı altında stokastik stok optimizasyonu probleminin dinamik
programlama ile çözümü. Gazi Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 17(3), 255-270.
Çimen, M., & Kirkbride, C. (2017). Approximate dynamic programming algorithms for multidimensional flexible
production-inventory problems. International Journal of Production Research, 55(7), 2034-2050.
Gouveia, L., Paias, A., & Sharma, D. (2011). Restricted dynamic programming based neighborhoods for the hop-
constrained minimum spanning tree problem. Journal of Heuristics, 17(1), 23-37.
Gromicho, J., van Hoorn, J. J., Kok, A. L., & Schutten, J. M. (2012). Restricted dynamic programming: a flexible
framework for solving realistic VRPs. Computers & Operations Research, 39(5), 902-909.
Jaruphongsa, W., Çetinkaya, S., & Lee, C. Y. (2004). A two-echelon inventory optimization model with demand time
window considerations. Journal of Global Optimization, 30(4), 347-366.
Ikeda, K., Yamamura, T., Mitamura, Y., Fujiwara, S., Tominaga, Y., & Kiyono, T. (1981). On-line recognition of hand-
written characters utilizing positional and stroke vector sequences. Pattern Recognition, 13(3), 191-206.
Kang, H. Y., & Lee, A. H. (2013). A stochastic lot-sizing model with multi-supplier and quantity discounts. International
Journal of Production Research, 51(1), 245-263.
Lin, M., Huh, W. T., & Wan, G. (2021). Multi-Period Lot-Sizing with Supplier Selection: Structural Results, Complexity
and Algorithms. Operations Research Letters, 49(4), 602-609.
Malandraki, C., & Dial, R. B. (1996). A restricted dynamic programming heuristic algorithm for the time dependent
traveling salesman problem. European Journal of Operational Research, 90(1), 45-55.
Malik, W., & Jung, Y. C. (2016). Exact and heuristic algorithms for runway scheduling. In 16th AIAA Aviation Technology,
Integration, and Operations Conference (p. 4072).
Moqri, M., Javadi, M., & Yazdian, A. (2011). Supplier selection and order lot sizing using dynamic programming.
International Journal of Industrial Engineering Computations, 2(2), 319-328.
Silver, E. A. (1973). A heuristic for selecting lot size quantities for the case of a deterministic time-varying demand rate
and discrete opportunities for replenishment. Production and Inventory Management, 2, 64-74.
Soysal, M., & Çimen, M. (2017). A simulation based restricted dynamic programming approach for the green time
dependent vehicle routing problem. Computers & Operations Research, 88, 297-305.
Sun, P., Dabia, S., Veelenturf, L. P., & Van Woensel, T. (2015). The Time-Dependent Pro_table Pickup and Delivery
Traveling Salesman Problem with Time Windows. Eindhoven University of Technology.
Visentin, A., Prestwich, S., Rossi, R., & Tarim, S. A. (2021). Computing optimal (R, s, S) policy parameters by a hybrid
of branch-and-bound and stochastic dynamic programming. European Journal of Operational Research, 294(1), 91-99.
Wagner, H. M., & Whitin, T. M. (1958). Dynamic version of the economic lot size model. Management science, 5(1), 89-
96.
Wang, S., & Chen, H. L. (2004). Restricted Dynamic Programming for Broadcast Scheduling. In International Conference
on Network Control and Engineering for QoS, Security and Mobility (pp. 139-150). Springer, Boston, MA.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

364
 THE ROLE AND IMPORTANCE OF OUTSOURCING IN LOGISTICS
ENTERPRISES

Şuayıp Özdemir1, Tuğrul Bayat2, Betül Kocatürk3

Abstract ⎯ In emerging global markets, with technological developments, increasing competition conditions
and increasing importance given to customer satisfaction, outsourcing of business functions, whether in the
core activity or not, is becoming a common management strategy for many companies and organizations.
Depending on the advantages of outsourcing, it is also preferred by logistics businesses. The aim of this study
is to reveal the perspectives of the logistics service companies on outsourcing and also to examine the
advantages they provide and the problems they encounter while determining the field of activity they prefer to
outsource. In this context, content analysis is applied to the data collected as a result of in-depth interviews
with ten logistics companies. As a result of the study, it is seen that all the logistics companies interviewed
prefer outsourcing in various fields such as transportation, storage, software, and labor. The reasons for
choosing outsourcing are primarily seen as cost reduction, customer satisfaction and coping with the
workload that occurs periodically. The study also shows that outsourcing continues to be of strategic
importance today.

Keywords ⎯ Logistics Activities, Logistics Service Providers, Outsourcing

1
Şuayıp Özdemir, Afyon Kocatepe University, Faculty of Economics and Administrative Sciences, Department of Business
Administration, Afyonkarahisar, Turkey, [email protected]
2
Tuğrul Bayat, Afyon Kocatepe University, Bolvadin Faculty of Applied Sciences, Department of Logistics Management,
Afyonkarahisar, Turkey, [email protected]
3
Betül Kocatürk, Afyon Kocatepe University, Department of Business Administration, 100/2000 YÖK Doctoral Scholarship Student,
Afyonkarahisar, Turkey, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
365
 INTRODUCTION
Nowadays, the global and competitive business world, where uncertainty prevails due to the rapid
development in transportation, management, communication and technology, businesses are trying to think of
ways to use their resources more effectively and efficiently in order to survive and develop (Sharifi & Zhang,
2001: 773; Jothimani & Sarmah, 2014; Aktaş, 2015). These ways include various applications such as
computer-aided manufacturing, just-in-time manufacturing, flexible production, agile manufacturing over
time. Outsourcing was included in these applications in the later twentieth century. Outsourcing helps
businesses to increase their competitiveness by reducing their costs, increasing their profitability and enabling
them to focus on their core competencies (Uygun et al., 2015) through the creation of value by an enterprise
that is an specialized in the field (Liou et al., 2011; Güçlü, 2013). Thus, businesses can focus on their best
areas and core competencies, and prefer outsourcing for other activities and processes (Maltz & Ellram, 1999;
Bulgurcu & Nakiboğlu, 2018).
Borders are disappearing due to the ever-expanding global markets, and the distances between production and
consumption points are increasing. Hereby, there is a greater need for logistics service providers that provide
flow of goods and service from production and consumption points. Therefore, logistics service providers
(3PLSP) play a crucial role in today's supply chain management. In order to reduce the burden of logistics
activities and to ensure customer satisfaction and overperform, businesses choose-prefer to outsource some or
all of the operations within the supply chain (Gri and Salker, 2018: 168).
The worldwide use and importance of logistics outsourcing due to globalization has been increasing
dramatically in the last decade. Also, it is expected that the sector will continue to grow in the future (Langley
et al, 2005; Deepen, 2007: 1). Especially, the globalization of operations strengthens worldwide dependence
of enterprises on logistics service providers (Zhu et al., 2002). Engelbrecht (2004) reveals that a company's
logistics performance has an impact on overall business performance and also states that outsourcing is an
important driver of logistics performance. Therefore, perfectly understanding the drivers of outsourcing
performance is critical information for managers in today's competitive business environment.
Outsourcing is still an emerging field of activity. In the literature, the number of studies on outsourcing,
especially in logistics activities of enterprises, is increasing day after day. In spite of that, no study has been
found on the outsourcing of logistics enterprises. The contribution of the study to the literature can be
summarized in four points:
i) Perspectives of logistics enterprises on outsourcing,
ii) Activities supplied by logistics enterprises within the scope of outsourcing,
iii) The reasons why logistics enterprises prefer outsourcing,
iv) The advantages and disadvantages of outsourcing to logistics enterprises are considered.
In this study, outsourcing, which is a strategic management practice to gain competitive advantage, is
addressed specific to logistics enterprises. In this context, the perspectives on outsourcing, the fields of
activity they prefer to outsource, the advantages they provide, and the problems they encounter are discussed
of ten companies operating in the leading logistics sector in Turkey. In addition, to what extent the logistics
enterprises outsource and which factors are effective in outsourcing decisions are examined.
The study consists of five section. After the introduction, the second section explains the concept of
outsourcing in terms of logistics. The third section includes the literature on outsourcing. The fourth section
provides information on the methodology and the last section contains conclusions and recommendations.

1. OUTSOURCING

One of the factors that fundamentally affects the structuring and functioning of today's organizations is the
understanding of gathering business activities around certain core competencies. Core competency refers to
the knowledge, skills and abilities that distinguish a business from other businesses, also plays a fundamental
role in realizing the vision of the business, and cannot be easily imitated by competitors (Koçel, 2020). Core
competency is the shaping and establishment of capabilities, talents, knowledge, experience, human resources
and intellectual capital, which are tangible and intangible in organizations. It is an opportunity that can be
established by the enterprise to enter the market or to close the knowledge gap. From the narrow point of

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
366
view, businesses can focus on their future strategies under favor of their basic capabilities and core
competencies (Sevim et al., 2008; Gilgeous and Parvee, 2011:217). In other words, a core competency is
defined as “The set of activities, skills or advantages that distinguishes the business from others and makes it
a valuable member of the team (organization or supply chain).” (Fawcett et al., 2014: 279).
In a dynamic and competitive international market, entrepreneurs are constantly looking for ways to
economize, improve the quality of goods and services, and increase the efficiency of their business in order to
maintain their business and compete successfully (Zitkiene and Dude, 2018: 342).Prahalad and Hamel (2000)
argue that to win in a global market, a company must be uniquely good at an activity that no other company
can do so well. They also state that companies that do not develop such a core competency will always find
themselves in the battle to protect their market shares and margins. Therefore, Prahalad (1990; 1993: 45)
emphasizes that while determining the basic skills, "Does the determined skill set/core competence contribute
significantly to what customers perceive as the added value of our organization?, Is the skill set difficult for
others to do the same or imitate (Is it difficult imitation for competitors hard to achieve?), Are we particularly
good at skill set, or are we willing to invest resources to excel?, Skill set, Is it broad enough to give us the
opportunity to enter many different markets or businesses? Is the core competency a meaningful resource to
differentiate itself from the competition? Does it provide the organization with a single distinguishable
identity? Can it cross borders for a single job? And can it cover new or existing business areas?” answers of
these questions can guide the way. It is defended that if the business can answer the questions in a meaningful
way, it should review the business processes in line with its basic capabilities and evaluate the outsourcing
option for the areas other than this (Prahalad, 1993 as cited in Sevim et al., 2008: 4).
In addition, Maltz and Ellram (1999) argued that businesses should ask these questions, such as “Which
purchasing-related activities should be outsourced because they can be performed more efficiently and
effectively outside the organization? What purchasing-related activities should not be outsourced as they can
be better performed internally? and Is outsourcing an opportunity or a threat to supply management? Are
there any factors pointing in one direction or the other?” and also he stated that they could decide whether to
use outsourcing in terms of purchasing by making use of the questions (Maltz and Ellram, 1999: 4).
Outsourcing is a management style at the strategic level that emerged in the 1980s and started to be seen
frequently in businesses after the 1990s. While it provides businesses with the advantage of focusing on their
core competencies, it gives the opportunity to improve processes by transferring some of their activities to
companies that are experts in their fields. The new economic order and the emerging developments brought
about various advancement imperatives for the businesses. In order to comply with the conditions of the day
and hold on to the market, businesses must be flexible and adapt quickly to innovations. Outsourcing is one of
the important management styles that enables businesses to be flexible, to carry out their business at an
advanced level in different areas, and to manage their core competencies in the best way (Güçlü, 2013: 16).
It is possible to say that outsourcing is not a concept that has emerged independently from other managerial
practices, although it offers businesses the opportunity to work more efficiently with less cost in the global
competitive environment. Practices such as "subcontracting" in the construction sector and "contract
manufacturing" in the manufacturing sector have played an important role in the development of outsourcing.
(Güçlü, 2013: 17).
UNCTAD (2017) stated that the share of services in GDP increased for all country categories between 1980
and 2015 at the meeting in Geneva. He also emphasized that these rates increased from 61% to 76% in
developed economies and from 42% to 55% in developing economies. These outcomes are an indicator of the
confidence in outsourcing and show that it will increase in the coming years.
Outsourcing is defined as “the transfer of responsibility for some or all of the activities previously performed
internally to an external third party business” (Ellram and Maltz 1997). Outsourcing is becoming more
popular tool for businesses to increase their productivity day after day (Maltz & Ellram, 1999; Bulgurcu &
Nakiboğlu, 2018). Outsourcing means that businesses buy the affairs or duties that are outside of their core
skills and competencies from other businesses that are specialized in their own field (Şahin and Berberoğlu,
2011: 34). Logistics process has been determined as an important activity that can be evaluated in outsourcing
in order to reduce costs and increase efficiency. Transferring all or part of the logistics function process to
(one) external business(s) can be defined as logistics outsourcing (Bulgurcu & Nakiboğlu, 2018).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
367
Logistics means all the actions that support all the necessary processes in the supply or production of anything
needed. On the other hand, logistics enterprises are businesses that operate in more than one logistics area in
order to support the production and consumption processes of everything needed in logistics (Keskin, 2015:
5). Recent conditions, attach great importance to customer service and customer satisfaction, come into
prominance of concepts such as just-in-time production, just-in-time distribution, increasing globalization and
competition, and the necessity of corporate collaborations push businesses to outsourcing (Şahin and
Berberoğlu, 2011: 34). Nowadays, it is seen that many businesses outsource some services with the aim of
focusing on their areas of expertise. Although logistics cooperation is becoming more important day after day,
it is seen that logistics has an important place among the activities that enterprises utilize from outsourcing
(Tüzemen, Yaprak, Güzel, 2017: 744). The increasing importance of logistics enterprises is effective in the
multidimensional approach of outsourcing, such as managerial (strategic), tactical, financial, operational, etc.,
specific to these enterprises (Munjal et al., 2018).
The emergence of concepts such as globalization, new economy and world markets point out the competitive
environment. In this environment, it is necessary for enterprises to use resources effectively and efficiently in
order to compete succesfully. At this point, businesses see outsourcing as a solution. Under favor of
outsourcing, businesses can focus on activities that provide competitive advantage and cooperate with
companies specialized in these areas in the management of resources outside their field of expertise, and
manage to carry out all their processes in accordance with the necessary quality standards (Tanyeri and Fırat,
2005: 268).
Zitkiene and Dude (2018) divide outsourcing into three main theoretical groups. These are as follows:
i – Theories for improving company indicators and reducing costs;
ii– Theories for a company's internal processes, managing resources, promoting innovation, and increasing
competitive advantage among companies operating in the same business;
iii – It consists of theories for developing interrelations with existing partners and seeking new ones,
expanding cooperation and making decisions regarding integration and merger with other companies.
Table 1 compares the differences between the classical management and outsourcing method.

Table 1. Differences Between Classical Management and Outsourcing Management

CLASSICAL MANAGEMENT OUTSOURCING MANAGEMENT
Positive results occur depending on the talent and Success depends on teamwork, not individual.
effort of the manager.
Information is only available during working hours. Uninterrupted access to information is possible.
Information may be inconsistent and inaccurate. Information is constantly checked.

Managers' decisions are awaited. Decisions are made in advance.

Managers have a variety of roles. Each task is performed by experts.

Reference: Gül, 2005: 162

It is possible to say that businesses gain a more professional perspective with outsourcing. Outsourcing has the
feature of taking a step forward of the planning and control processes, which are the basic components of
management. The fact that information is constantly accessible and under control, teamwork comes to the fore
and give the every job to experts makes outsourcing a forward carrier for businesses.
1.1. Reasons for Outsourcing
It is possible to divide the reasons that lead businesses to use outsourcing as external and internal factors.
While external factors consists of factors such as social, technological, economic, cultural changes,
globalization, rapid change, increased competition, internal factors consist of various factors such as of
focusing on core competencies, reducing costs, getting rid of the unwieldiness of the organizational structure

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
368
and increasing flexibility, increasing quality, process renewal, downsizing, following the latest technological
innovations, reducing risk, rank among successful businesses, out-of-control functions, resource transfer,
allocation of resources, reducing investment expenditures, utilizing from financial resources, become a
dominant player in the market, and resistance to change (Güçlü, 2013: 35-46).
The Outsourcing Institute lists the main drivers of outsourcing for businesses as follows (Şahin and
Berberoğlu, 2011: 37):
● Expediting the business process reengineering,
● Gaining world-class abilites,
● Cash flow provided by transfer of asset,
● Creating resources for other objectives,
● Activities that are difficult to manage or control,
● Developing the business objectives (focus on core competecency),
● Making capital funds available,
● Reducing operating costs,
● Reduce risks,
● Accessing inaccessible resources.
Sevim et al. (2008), on the other hand, states the reasons for benefiting from third party logistics activities for
the business as follows; reduce costs, focus on core competencies, increase flexibility, reduce capital
expenditures for logistics activities, to provide more appropriate reporting of logistics activities, to ensure the
use of information technologies, improve quality, connect with foreign markets and reduce the need for high-
level knowledge and skills.
It is common for many businesses to outsource their incidental operations to focus on core competencies.
Outsourcing of services continues to expand rapidly, involving more activities, especially due to the
development of information and communication technologies and the liberalization of international trade
(UNCTAD, 2004). In this context, it is seen that companies in the service sector tend to outsource unessential
services, such as building maintenance services in hospitals, IT services in banks. It is seen that outsourcing is
used for various reasons for both core and non-core activities in communication and transportation areas
(Dudė et al., 2021). In almost all definitions of outsourcing, it is mentioned that businesses are being in
tendency to outsourcing in order to focus on their core competencies. The fact remains that it is seen that
outsourcing is preferred in core activities in the logistics sector.
1.2. Advantages and Disadvantages of Outsourcing
In the 1980s, when the traditional management approach began to give place to the modern management
approach, the "Outsourcing" approach, which is one of the modern approaches, provides businesses with
advantages in many areas such as production, management and information technologies (Ömürgönülşen and
Sarıgül, 2021: 18).
Advantages of outsourcing to the businesss are as follows; reducing and controlling costs, providing
flexibility, focusing on core business, creating a variable cost structure, increasing knowledge and skills,
increasing quality, monitoring technology, increasing efficiency, contributing to risk management, adding
value to the business, creating an appropriate business structure, increasing competitiveness, ensuring balance
in income distribution can be listed. The disadvantages for the business are loss of flexibility, loss of control
over the supplier company, selection of the wrong supplier company, loss of the capabilities of the business,
negative effects on personnel, loss of cross-functional skills, problems arising from cultural differences, loss
of control over personnel, focus on short-term economic purposes, thinking that the agreement with the
supplier company will control itself, inability to provide information security and suppliers entering the
market (Güçlü, 2013: 54-60).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
369
From the supply chain point of view, for example by outsourcing logistics activities, the manufacturing
business not only reduces the logistics service burden on retailers, but also saves on capital investments in
various areas such as financial risk, investments in logistics assets, physical distribution systems, and
information networks. (Giri and Sarker, 2017: 176). The benefit of outsourcing will emerge when the
appropriate scales of these aspects are taken into account in the analysis of the system.
The decision to outsource (insource or refuse to insource) also depends on where the activities are carried out
(Buckley, 2011a). Multinational corporations in industrially developed economies outsource certain activities
to developing countries, such as outsource production to China and software services to India. The reason is
that they enable them to be more productive by utilizing advantage of the lower labor costs in these countries
(Saxenian, 2002).
Access to specialized services and knowledge-intensive activities through outsourcing on an international
scale allows the business to access information and resources linked to other businesses (Quinn, 1999). This
type of outsourcing often takes place outside the home country. Because the expertise that the business needs
to gain competitive advantage has spread all over the world (Papanastassiou & Pearce, 2009). Also, it can be
argued that outsourcing may have a negative impact on the business as it simultaneously leads to loss of
control, leakage and dissemination of information and resources to other businesses (Bettis, Bradley, &
Hamel, 1992; Stanko and Olleros, 2013). Besides that, outsourcing requires transaction costs from managing
the relationship with an external partner, possibly higher than the costs of organizing and controlling activities
internally (Geis, 2007; Parida, Wincent, & Oghazi, 2016). The negative externalities of outsourcing are likely
to be higher for large and technology-intensive enterprises in advanced economies. Businesses operating in
emerging or peripheral economies are less prone to such disadvantages as they generally do not have
specialized knowledge and resources (Munjal et al., 2018: 2).
Table 2 shows the advantages and disadvantages of outsourcing.

Table 2. Advantages and Disadvantages of Outsourcing

Advantages Disadvantages

1. Significant reduction in costs and 1. The need for a higher level and more complex
creating new revenue areas, communication,
2. An increase in the financial resources 2. Problems arising from the entry of new actors,
of the business, personalities and demands into the system,
3. A decrease in fixed 3. Decreased trust in the eyes of the employees
investments, and the union,
4. Higher quality and effectiveness, 4. Alienation of customers with the transfer of the function
to the service providers that the customer is used to,
5. More utilize form functional
experts, 5. Inconveniences that may arise from long-term
agreements for short-term savings,
6. Economies of scale,
6. Loss of service and personnel control over time,
7. Creating competitive advantage
by focusing on core 7. Conflict with the supplier business and causing new
competencies, problems.
8. Opportunity to test new demands in 8. The obligation to continuously monitor supplier
the market with outsourcing performance.
cheaper.

Reference: Pelit, 2007: 31.

Outsourcing has many advantages for businesses. Sharing risk and benefit in this collaboration drives
resource/service providers to a desire to provide better service. On the other hand, outsourcing have some
reservations. Therefore, the importance of the right choice and the right management about the resource

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
370
provider is undeniable. Because, even if the risks are shared, in case of any problem, the business is the
drawee of the customer.

2. LITERATURE REVIEW
When examing the studies on outsourcing, it is seen that the effect of outsourcing in logistics activities on the
key performance indicators (operational, financial, profitability, efficiency, etc.) of the enterprises operating in
the manufacturing or service sector is tried to be measured.
Aktaş and Uluengin (2005) revealed the current status of logistics activities in Turkey. SPSS and factor
analysis, cluster analysis and ANOVA methods were used in the analysis of the data collected by the survey
method. Depending on the findings, it showed that the power of 3PL enterprises in Turkey was
underestimated. In addition to the fact that 95% of the companies outsourcing logistics activities in Turkey
had foreign capital, Turkish businessmen thought that they should do their work themselves and were not
aware of the benefits of outsourcing logistics activities. When choosing a 3PL business, businesses chose the
carrier that has a good reputation and/or is easy to cooperate with. There was no significant relationship
between outsourcing decisions and the scale of the firm, however, they found that the level of outsourcing
differed significantly according to the presence of the logistics department. The results revealed that existing
3PL providers were actually playing the role of load carriers.
Gül (2005), collected data from a total of 26 companies from 3 sectors, namely white goods, automotive and
food, by means of a questionnaire in the manufacturing industry. Priorities of enterprises in outsourcing are; 1.
Improving organizational performance, 2. Increasing quality, 3. Reducing cost; besides that concerns about
outsourcing are: 1. Compromise on quality, 2. Focus on short-term economic goals, 3. Loss of talent and 4.
Loss of control. In addition, it has been revealed that the enterprises give importance to the criteria of quality,
talent, organizational structure, references, resources, price, cultural harmony in the selection of
subcontractors, respectively.
Sevim et al. (2008) examined the effect of outsourcing in logistics activities on service quality in their study in
Turkey. The data collected through the questionnaire were analyzed with the SERVQUAL method. According
to the results, it is seen that the companies that provide logistics service (3rd party logistics companies) cannot
provide customer satisfaction in the dimensions of responsiveness, empathy, assurance, physical values and
reliability, which are the elements of service quality.
Aktaş et al., (2011) examined companies that use outsourcing in logistics services and other companies that
meet their logistics needs insourcing, in terms of logistics services in Turkey. The data was collected through
the survey. As a result of the study, it has been seen that the types of logistics services that are outsourced and
planned to be outsourced affect the problems encountered in the outsourcing of these services, the logistics
costs, the decision makers in the outsourcing of logistics activities and the information resources used in the
decision-making process. In addition, the most frequently outsourced logistics activities were domestic and
international road transport. This service is followed by customs clearance and distribution to the customer
warehouse. While domestic road transport stands out again, distribution to the end customer and customs
clearance are considered as areas where less outsourcing is planned.
Zöngur et al. (2016) conducted face-to-face interviews with businesses in their study on the effects, causes and
risks of outsourcing in accommodation businesses. While the reasons for using outsourcing are primarily cost
reduction and unwilling to deal with hard work, however, the reasons for not using outsourcing are seen as
unwilling to take risks and worrying about loss of customers and reputation.
Özkan and Elma (2017) examine the level of cooperation between logistics service recipients and logistics
service providers in their study in Turkey. Semi-structured questions were used in the study, and the answers
from 7 logistics service companies and 10 randomly selected companies were analyzed. According to the
results, it is seen that the positive effect expected from the use of outsourcing will be achieved by eliminating
some prejudices and adopting modern and effective methods. It is also stated that the size of the companies is
closely related to the outsourcing, and the reason for the reluctance to outsource in Turkey is also related to
the size.
Bulgurcu and Nakiboğlu (2018) try to determine the criteria used by 14 companies operating in the cement
industry in the selection of logistics service providers in Turkey. The data collected by the questionnaire

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
371
method were analyzed with the Chang's Fuzzy Analytical Hierarchy Process (FAHP) method. According to
the results, it has been determined that the service/operation quality is the most important factor, and the
service price, which is below this factor, is the most important sub-factor.
Owusu et al. (2018) evaluates the in-port service provider (subcontractor) performance of Takoradi, a
marketing company in China. The data collected by the survey method were analyzed with SPSS. As a result
of the study, it is stated that outsourcing companies reduce costs, provide efficiency, improve operational
standards, and that service providers should have sufficient vehicles (especially trucks).
Munjal et al. (2018) investigated the effect of outsourcing on the financial performance of 1710 companies in
India. GMM method was used in the study in which panel data of 2001-2013 period was used. The results of
the study show that the positive effects of technological knowledge and professional performance are
determined by the company size, business growth and resource abundance. They emphasized that outsourcing
contributes more to the performance of especially small-scale enterprises.
Akben and Fidan (2019) used a questionnaire in their field study on Iron and Steel enterprises operating in the
Iskenderun region. They analyzed the data with SPSS 23. In the study, the areas of outsourcing on the basis of
logistics companies, the factors that are effective in the use and selection of outsourcing, the way of being
aware of the suppliers, why the outsourcing (3PL) is needed, and the problems arising from the suppliers are
emphasized. While the main reasons for outsourcing are cost reduction, competition, and the service
provider's quality physical tools and equipment, the reasons for unwilling outsourcing are seen as cost and
jobs that require expertise.
Dude et al. (2021) analyzed the use of outsourcing in subgroups of the service sector. Data collected with
semi-structured questions were analyzed with the Kruskal–Wallis non-parametric test. According to the
results, the use of outsourcing in the service sector was increasing; however, it is revealed that service
providers vary depending on the provision of services, the nature of the service company's activities and the
needs they meet, whether they are public or seeking profit.

3. METHODOLOGY
In this section, the aim of the study, the method and the findings are given.
3.1. Aim and Method
The aim of this research is to examine the outsourcing in the logistics sector. It is aimed to determine the
perspectives of logistics companies on outsourcing, their preference for outsourcing in their activities, which
activities are outsourced, the reasons for choosing outsourcing and the advantages they provide, the problems
they experience in outsourcing or the disadvantages they face with. It is seen that many enterprises continue
their logistics activities by making agreements with logistics companies and assigne the processes to them.
However, there are not enough studies on the outsourcing of logistics companies, which is a form of
management in developing their services.
Statistical databases (TUIK, DATABANK, WTO, Eurostat, STATOECD, UNCTAD, etc.) are lacking in
producing consistent and comparable data on outsourcing models. Also, it is emphasized that outsourcing is
complex and there is a lack of indicators related to outsourcing (Srivastava and Rajan, 2006), and the available
data is limited (Horgos, 2007) in the literature.
Qualitative research methods are used to reveal the meanings and interpretations of structures that are difficult
to quantify and do not begin with the formation of hypotheses. This type of research, such as the interview,
perfectly satisfies the assumptions of qualitative methodology and allows the researcher to examine research
participants' perspectives and collect wealth of data, discovering details and unique nuances (Gaižauskaitė &
Valavičienė, 2016; 392); Dude et al, 2021; 7 quoting from Žydžiunaite & Sabaliauskas, 2017). There are no
clear rules for determining the sample size of qualitative research (Rupšiene, 2007). In addition, the accuracy
of the decisions and evaluations of the collective evaluation models with equal weights of a small group of
experts do not lag behind the judgments and evaluations of a large group of experts (Libby & Blashfield,
1987). Augustinaitis et al., (2009) recommend that at least five and at most ten experts be included in the peer
review team in order to maintain the accuracy and reliability of the peer review.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
372
The population of the study consists of 10 leading outsourcing logistics service providers in Turkey. One of
the qualitative research techniques face to face interview method was used. As a result of in-depth interviews
with ten companies with semi-structured questions, content analysis was applied to the obtained data. The
MAXQDA 2020 program was used in the analysis. In this method, the most frequently mentioned and
emphasized words were determined as codes in the analysis in the answers to the questions asked to the
participants. With the obtained data and generated codes, the results are presented with various visual tools
such as code cloud analysis, code map, code matrix scanner analysis through the MAXQDA 2020.
3.2. Findings
Perspective and Preference of Businesses on Outsourcing
In this section, it was tried to understand the outsourcing perceptions of the enterprises and it was asked
whether they used outsourcing.
For this purpose, “How is your company's perspective on outsourcing and do you have any activities that you
prefer to outsource?” question was posed to the enterprises. It has been seen that only one company does not
look very favorably on outsourcing, but prefers it at certain rates, and all other companies prefer outsourcing
due to its various advantages and employ it in various activities.
Fields of Activity where Outsourcing is Preferred
After it was determined that all of the enterprises preferred outsourcing “In which fields of activity do you
prefer outsourcing?” question was posed. With this question, it is aimed to distinguish the exceptional
outsourcing uses that have diffuse throughout the logistics companies. It has been observed that all of the
logistics companies, whose main field of activity is transportation, use outsourcing in the vehicles where
freight transportation is carried out. One of the companies interviewed (B8) stated that they do not have any
self-owned goods and that they carry out transportation with sub-suppliers. All other businesses use a
significant amount of their own assets as well as a large amount of contracted vehicles. Other enterprise (B7)
stated that it has a vehicle supply team and they make the necessary connections and follow-up. (B10)
enterprise stated that it carries out the transportation within Istanbul within its own structure and rents a
vehicle for all domestic transportation. Also, (B4) stated that while it was in favor of continuing its activities
with its own vehicle and team, it rented a car when necessary.
Besides the vehicle, warehouse (B2, B3, B4, B7, B8), labor (B2, B3, B4, B6, B7, B9), software (B1, B3, B5,
B6, B7, B9, B10), staff education (B1, B3, B5) and financial audit (B2, B5, B9) are seen to be intensively
outsourced. Some companies (B3, B4) stated that they prefer to rent warehouses especially for short-term
projects. One of the companies (B2) stated that the warehouses are rented for a long time (like 10 years).
Firms outsourcing the software seem to have ongoing connections with the service provider, and have long-
term relationships for updating and consulting.
In addition, It is seen that outsourcing is used in various subjects such as security (B5, B9), service
maintenance (B4, B5), small-scale cargo (B3, B7), cleaning (B1, B9), staff meal (B7), human resources (B9),
equipment (B4), sustainability and green logistics studies (B7), packaging (B3), and providing activities and
training to customers (B7). The percentages of the fields of activity that businesses prefer to outsource are
shown in Figure 1.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
373
 Figure 1. Percentage of Outsourced Activities

The code matrix scanner shows and visualizes how often each code is used in each document reviewed. The
last row and column shows how much the codes are used in all documents in total. While the vehicle is the
most used code, the codes such as packaging, equipment, HR, activity and training presentation to customers,
staff meal, sustainability and green logistics are the least observed codes. The size of the squares seen at the
intersection of each document and the code reflects the frequency of use of codes in documents.

Figure 2. Code Matrix Scanner Analysis of Outsourced Activities

Reasons for Outsourcing and Advantages

In this section, the main drivers that push companies to outsource and the advantages that companies get from
outsourcing are asked to the enterprises. It is seen that most of the companies (B1, B2, B3, B4, B5, B7, B9,
B10) prefer to outsourcing in order to reduce costs and downsizing due to the high costs of acquiring own
property and working with a large number of personnel. It is seen that options such as using the vehicles of
transportation cooperatives on the way back play a role in reducing costs.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
374
It is seen that businesses can bring the capacity to the required level by providing subcontracted workers and
external resources, and in this way, they are able to take precautions against the risks that may arise and
continue their business in cases where the work that emerges out of the expected and the capacity is not
sufficient. Some companies (B1, B3, B7) emphasized that they use outsourcing to provide timely service in
order to ensure customer satisfaction and provide better service, especially with the intensity that occurs
periodically. Another enteprise (B6) stated that there is no business circulation that will keep the vehicles,
warehouses, and labor within the enterprise, therefore outsourcing is important in coping with workload and
providing continuous good service.
While some of the interviewed enterprises (B3, B5, B8) directly stated the advantage of gaining flexibility
from outsourcing, others did not. However, outsourcing is common in the sector in order to respond to the
changing business volume. From this point of view, it is possible to say that the advantage of flexibility is an
important factor.
Another advantage of outsourcing stands out as gaining speed (B3, B4, B6, B7, B9). Business volume changes
(B4, B5, B6, B7, B9) lead to situations where capacity is not enough, outsourcing emerges as an important
savior and enables rapid response and adaptation to change. Coping with workload (B1, B3, B4, B5, B6, B7,
B9) is seen as an intense response. Lack of resources (B1, B4, B5, B6, B9) is another common response. It can
be said that the resources of the enterprises are insufficient, especially in cases where there is a workload, and
therefore the enterprises turn to external resources.
Increasing quality (B3, B6, B7, B9, B10), profitability (B2), convenience (B9), downsizing (B3), competitive
advantage (B1, B7), productivity (B2) are seen as other answers.
It is also broadly seen that process improvement (B2, B3, B5, B6, B7, B9) is possible with outsourcing.
MAXQDA has the ability to visually present the intensity of codes assigned to documents with code cloud
analysis. In the code cloud analysis seen in Figure 3, the reasons for the outsourcing of the enterprises and the
advantages they provide were mentioned in the interviews with a frequency proportional to the size of the
words and were assigned to the documents with the same frequency.

Figure 3. Code Cloud Analysis Reasons for Tending to Outsource and the Advantages Provided

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
375
Code map can be used in observing the relationship between codes and viewing them as a whole. It can be
said that the codes located in close proximity to each other are in an intense relationship. A code map was
created in order to see the relationship between the intensely given answers. Coping with workload, changes
in work volume, lack of resources and increasing speed are observed as codes that are intensely related. In this
context, it is possible to say that businesses are faced with workload and lack of resources in cases where
business volume changes occur. In such cases, it is seen that outsourcing is intensly used to increase speed.
Similarly, the importance of outsourcing in coping with workload is seen to be important in obtaining all
returns such as increasing quality, customer satisfaction, providing better service, and process improvement. It
is seen that all the other mentioned codes are in relation with each other, and advantages related to each other
are obtained with the use of outsourcing.

Figure 4. Code Map of Reasons for Tending to Outsource and the Advantages Provided

Drawbacks of Outsourcing
In this section, the interviewed enterprises were asked about the problems they encounter with outsourcing
and the possible problems they may encounter. One of the enterprises (B8) emphasized the difficulty of
establishing control over service providers and stated that they do not prefer to make long-term agreements
with service providers. He said that in this way, they can stop working together immediately in case of any
negativity. He stated that this situation played a role in the increase in costs and outsourcing could be costly
for them. However, during the interview, the enterprise (B8) emphasized the principle of not acquiring any
own property and working entirely with external resources in order to gain mobility and flexibility, and stated

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
376
that it compensated for this situation by providing services at higher prices under favor of the prestige and
customer satisfaction it achieved.
Some other firms (B4, B6) also stated that outsourcing can be dangerous to the point of losing control over
service providers. One of the enterprises (B6) expressed the situation as follows: “Think of it this way. Can
you use the car you rented from outside or your own car more actively? Your own car is yours, you can use it
as you wish. Unfortunately, you cannot do this to an outside vehicle. The driver is different, the vehicle is
different. After all, because we are using someone else's car, there may be problems in directing or
dominating them.” Problems in management and the risk of loss of control are seen as a common
disadvantages in outsourcing.
Other companies (B2, B4, B6) stated that outsourcing could cause problems in staff management. Businesses
stated that they cannot be effective enough on the personnel who do not work within their own organization.
It can be seen that the main concerns of enterprises regarding outsourcing are losing control and the
emergence of troublesome situations in business processes.

DISCUSSION AND CONCLUSION

In the global market, the transformation continues depending on the development of information and
communication, technological progress, increasing risk and flexibility, focusing on the specialization of
companies and globalization (Zhu et al., 2001). Therefore, companies reallocate resources by directing them
to core business resources in order to improve the quality of their business and gain an advantage over their
competitors. In addition to the traditional ways of organizing work, new operational strategies are being
developed to separate functions and in some cases transfer them to service providers specialized in this field
(Zıtkiene and Dude, 2018: 346-347).
After the 1990s, it is seen that businesses tend to use outsourcing as a management style and transfer some of
their activities to different companies that are experts in their fields. It can be said that the main reasons for
this change are globalization, developing technology and increasing competition. There are various reasons
for this orientation, such as reducing costs, focusing on core competencies, and increasing quality. However, it
can become a threat to businesses if it is not managed properly or the wrong enterprise is selected. The fact
that it can offer a wide range of benefits to the business and become a threat makes outsourcing a strategic
decision issue.
Within the scope of the research, outsourcing of logistics companies has been examined and it has been seen
that they prefer outsourcing in various core and non-core activities and provide various advantages. There is a
perception that outsourcing is an indispensable element in the sector. It has been observed that outsourcing is
preferred in various subjects such as warehouse, labor, equipment, software, packaging, financial audit,
security, HR, staff dinner and training, maintenance, sustainability and green logistics, especially the vehicles
of transportation. The findings are in coninsides with the ressults of Dudė et al., (2021), they stated that “When
we look at the service sector, it is seen that companies tend to outsource non-core activities. In
communication and transportation sector, it is seen that outsourcing is used for various reasons for both core
and non-core activities”.
Owusu (2018) stated that having a sufficient number of vehicles (especially trucks) is an important factor in
the logistics service provider preferences of enterprises. Naturally, businesses want to be sure that they will
not have problems with the vehicle when transferring the transportation to another company. It is seen that all
of the logistics companies interviewed receive vehicle support from subcontractors in addition to their own
resources. It is seen that the main reason lying here is to fully respond to customer requests and to minimize
the risk of come up against to customers.
Considering the reasons that push companies to use outsourcing and the advantages they provide are as
follows; cost reduction (Owusu, 2018; Akben & Fidan, 2019; Zöngür, 2016, Gül, 2005), customer
satisfaction, coping with workload, process improvement (Gül, 2005, Owusu, 2018), business volume
changes, increasing speed and quality (Gül, 2005). It is seen that the logistics enterprises with certain financial
strength incur their own cost of ownership to a certain extent, and after a certain point they prefer to direct
their financial power to different areas.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
377
The significant impact of the operational and financial advantage obtained from outsourcing on customer
satisfaction has also been frequently mentioned. In addition, it is seen that the enterprises in the sector do not
have a continuous and regular business circulation and they are faced with a changing business volume. And
also It is thought that outsourcing is an important driver of this change, since it is essential to respond and
adapt quickly, to maintain quality and customer satisfaction.
Bulgurcu and Nakiboğlu (2018) revealed that the most important criterion in choosing a logistics service
provider is service/operation quality. In this study, it is seen that one of the most important reasons for
logistics enterprises to turn to outsourcing is to increase speed and quality. When the findings are examined,
the reasons for the outsourcing of the enterprises completely coincide with the customer demands. It is seen
that outsourcing plays an important role in increasing the service quality of logistics enterprises. On the other
hand, Özkan and Elma (2017) stated that the positive effect expected from outsourcing will occur with the
elimination of prejudices about outsourcing and the adoption of modern and effective methods. In the logistics
sector, it is seen that there are no negative judgments about the use of outsourcing, on the contrary, its
importance has been grasped. In addition, Özkan and Elma (2017) stated that large-scale companies need
more outsourcing than small-scale companies. It is thought that the large scale of all the companies
interviewed may be the reason for the positive perspective on outsourcing.
Management and controlling problems (Gül, 2005, Zöngür et al., 2016), human resources management
problems and financial problems (Akben & Fidan, 2019; Gül, 2005) are seen as problems encountered or
likely to be encountered with outsourcing.
In service businesses, it is very important to offer quality, high performance and extended service to customers
with cost advantage. Businesses should optimize their services in order to gain competitive advantage. In this
context, outsourcing in the logistics sector stands out as an important management style. It is observed that
enterprises in the sector actively use outsourcing in an effort to make the essential services they offer more
advantageous in operational and financial terms. On the other hand, the transformation of logistics requires
logistics companies to develop their advanced integrated logistics support skills. It is possible to talk about the
necessity and importance of more active use of outsourcing in providing more advanced services and
developing these skills.
In future research, MAXQDA can be applied in other sectors’ data to see the most and the less factor affecting
the sector, and the findings can be compared with different sectors.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
378
 REFERENCES
Akben, İ. & Fidan, O. (2019). Demir Çelik Sektöründe Lojistikte Dış Kaynak Kullanımı (3PL): İskenderun
Bölgesinde Bir İnceleme.
Aktaş, E., Agaran, B., Ulengin, F. & Önsel, Ş. (2011). The use of outsourcing logistics activities: The case of
Turkey. Transportation Research Part C-emerging Technologies, 19, 833-852. 10.1016/j.trc.2011.02.005.
Aktaş, K. (2015). Uluslararası İşletmelerde Stratejik Yönetim. İstanbul. Uluslararası Yönetim ve Sosyal
Araştırmalar Dergisi, 2(1).
Augustinaitis, A., Rudzkienė, V., Petrauskas, R., Dagytė, I., Martinaitytė, E., Leichteris, E., Malinauskienė, E.,
Višnevska, V. & Žilionienė, I. (2009). Lietuvos e. valdžios gairės: ateities įžvalgų tyrimas: kolektyvinė monografija.
Vilnius: Mykolas Romeris University Publishing Centre.
Bettis, R. A., Bradley, S. P., & Hamel, G. (1992). Outsourcing and industrial decline. The Executive, 6(1), 7–22.
Bhatnagar, R., Sohal, A. S., Millen, R. (1999). Third party logistics services: a Singapore perspective.
International Journal of Physical Distribution & Logistics Management. 29 (9), 569- 587.
Buckley, P. J. (2011a). Globalization and the global factory. Cheltenham: Edward Elgar.
Bulgurcu, B. K. & Nakiboglu, G. (2018). An extent analysis of 3PL provider selection criteria: A case on
Turkey cement sector. Cogent Business & Management. 5. 10.1080/23311975.2018.1469183.
Dudė, U., Žitkienė, R., Jurevičienė, D., Skvarciany, V., & Lapinskaite, I. (2021). Evaluation of Outsourcing
Development in the Service Sector. Economies, 9(2), 1-19.
Ellram, L. & Maltz, A. (1997). Outsourcing: Implications for SupplyManagement. Centre for Advanced
Purchasing Studies, US.
Fawcett, S.E., Ellram, L.M., Ogden, J.A., (2014). Supply Chain Management from Vision to Implementation
Fawcett Ellram Ogden First Edition, Pearson.
Gaižauskaitė, I., & Valavičienė, N. (2016). Socialinių tyrimų metodai: kokybinis interviu: vadovėlis [Social
Research Methods: Qualitative Interview: Textbook]. Valstybes Įmonė Registrų centras.
Geis, G. (2007). Business outsourcing and the agency cost problem. Notre Dame Law Review, 82(3), 955–1003.
Gilgeous, V., & Parveen, K. (2001). Core competency requirements for manufacturing effectiveness. Integrated
Manufacturing Systems, 12(3), 217–227. doi:10.1108/09576060110391183
Giri, B. C., & Sarker, B. R. (2017). Improving performance by coordinating a supply chain with third party
logistics outsourcing under production disruption. Computers & Industrial Engineering, 103, 168–177.
doi:10.1016/j.cie.2016.11.022
Güçlü, A. (2013). Dış Kaynak Kullanımı (Outsourcing) ve Vergilendirilmesi. Ankara: Seçkin Yayıncılık.
Gül, H. (2005). Dış kaynak kullanma nedenleri ve taşıdığı riskler: İmalat sanayiinde bir uygulama. Yönetim ve
Ekonomi Araştırmaları Dergisi, 3(4), 157-184.
Horgos, Daniel. (2007). International Outsourcing—Some Measurement Problems: An Empirical Analysis of
Outsourcing Activities in Germany. Hamburg: Department of Economics, Helmut Schmidt University.
Jothimani, D., & Sarmah, S. P. (2014). Supply chain performance measurement for third party logistics.
Benchmarking: An International Journal, 21(6), 944–963. doi:10.1108/BIJ-09-2012-0064
Keskin, H. (2015). Lojistik El Kitabı - Küresel Aktörlerin Lojistik Pratikleri. Ankara: Nobel Akademik
Yayıncılık.
Koçel, T. (2020), İşletme Yöneticiliği, Genişletilmiş 18. Baskı, Beta Yayınları, İstanbul.
Korucuk, S., & Erdal, H. (2018). İmalat İşletmelerinde Dış Kaynak Kullanımının İşletme Verimliliği Ve
Performansına Etkisi Üzerine Bir Araştırma, Stratejik Yönetim Araştırmaları Dergisi, 1(2), 1-20.
Langley, C. J., Dort, E. V., Ross, T., Topp, U. S., Ykes, S. R., Stata, R., & Dengel, T. (2005). Third-party
logistics–results and findings of the 10th annual study. Atlanta: Georgia Institute of Technology.
Libby, Robert, & Roger K. Blashfield. (1987). Performance of a composite as a function of the number of
judges. Organisational Behavior and Human Performance, 21, 121–29.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
379
 Liou, J. J. H., Wang, H. S., Hsu, C. C., & Yin, S. L. (2011). A hybrid model for selection of an outsourcing
provider. Applied Mathematical Modelling, 35, 5121–5133. doi:10.1016/j.apm.2011.04.020
Maltz, A., & Ellram, L. (1999). Outsourcing Supply Management. The Journal of Supply Chain Management,
35(2), 4–17. doi:10.1111/j.1745-493x.1999.tb00232.x
Munjal, S., Requejob, I. ve Kunduc, S.K. (2018). Offshore Outsourcing and Firm Performance: Moderating
Effects of Size, Growth and Slack Resources. Journal of Business Research, In Press.
Owusu, A., Nanjing H. & Duah, K. (2018). The Effect Of Outsourcing Intra-Port Haulage Operations on
Productivity: The Case of Cocoa Marketing Company Limited, Takoradi, European Journal of Research and Reflection
in Management Sciences, 6(1), 23-48.
Ömürgönülşen, M. ve Sarıgül, S. S. (2021). Tedarik Zinciri Yönetiminde Yöneticilerin Dış Kaynak Kullanımı
Konusunda Algılarının Değerlendirilmesi: Sağlık Sektöründe Bir Uygulama. Verimlilik Dergisi, 1, 17-34.
Özkan, A., & Elma, O. (2017). Demand of Outsourcing in Turkey: An Emprical Study. Revista Produção e
Desenvolvimento, 3, 78-88. 10.32358/rpd.2017.v3.233.
Papanastassiou, M., & Pearce, R. D. (2009). The strategic development of multinationals: Subsidiaries and
innovation. Basingstoke: Palgrave Macmillan.
Parida, V., Wincent, J., & Oghazi, P. (2016). Transaction costs theory and coordinated safeguards investment in
R&D offshoring. Journal of Business Research, 69(5), 1823–1828.
Prahalad, C.K.. (2000). The Core Competence of the Corporation. 10.1016/B978-0-7506-7223-8.50003-4.
Prahalad, C.K. (1993) The Role of Core Competencies in the Corporation, Research-Technology Management,
36(6), 40-47, DOI: 10.1080/08956308.1993.11670940
Prahalad, C. K., & Hamel, G. (1990, May–June). The core competence of the corporation. Harvard Business
Review, 79–90.
Pelit, E. (2007). İşletmelerde dış kaynak kullanımı ve etkinliği: otel işletmeleri açısından kavramsal bir
inceleme. Verimlilik Dergisi, 2007(3), 25-40.
Quinn, J. B. (1999). Strategic outsourcing: Leveraging knowledge capabilities. Sloan Management Review,
40(4), 9.
Rupšienė, L. (2007). Kokybinio tyrimo duomenų rinkimo metodologija (Metodology of qualitative collection of
data). Klaipėda: KU leidykla.
Saxenian, A. (2002). Transnational communities and the evolution of global production networks: The cases of
Taiwan, China and India. Industry and Innovation, 9(3), 183–202
Sevim, Ş., Akdemir, A., Vatansever, K. (2008). Lojistik Faliyetlerinde Dış Kaynak Kullanan İşletmelerin
Aldıkları Hizmetlerin Kaltesinin Değerlendirilmesine Yönelik Bir İnceleme. Süleyman Demirel Üniversitesi İktisadi ve
İdari Bilimler Fakültesi Dergisi, 13 (1), 1-27.
Sharifi, H., & Z. Zhang, (2001), Agile Manufacturing in Practice – Application of a Methodology, International
Journal of Operations& Production Management, 21(5/6), 772-794.
Srivastava, Dabhana, & Ramkishen Rajan. (2006). Global outsourcing of services: Issue and implication.
Harvard Asia Pacsific Review, 9, 39–40.
Stanko, M. A., & Olleros, X. (2013). Industry growth and the knowledge spillover regime: Does outsourcing
harm innovativeness but help profit? Journal of Business Research, 66(10), 2007–2016.
Şahin, A. G. & Berberoğlu, N. (2011). Lojistik Outsourcing Karar Süreci ve 3PL Firma Seçim Kriterleri. Online
Academic Journal of Information Technology, 2(5), 33-50.
Tanyeri, M. ve Fırat, A. (2005). Rekabet Değişkeni Olarak Dış Kaynak Kullanımı (Outsourcing). Dokuz Eylül
Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, 7(3), 268-279.
Tüzemen, A., Yaprak, B. ve Güzel, D. (2017). Firmaların 3PL (Üçüncü Parti Lojistik) Hizmet Sağlayıcılarını
Seçerken Kullandıkları Kriterler Üzerine Bir Çalışma: Erzurum İhracatçıları Örneği. Ataturk University Journal of
Economics & Administrative Sciences, 31(3), 775-792.
UNCTAD, (2004) World investment report:the shift toward services//United Nations Commission on Trade and
Development 2004 http://unctad.org/en/docs/wir2004_en.pdf

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
380
 UNCTAD (2017) The role of the services economy and trade in structural transformation and inclusive
development, Trade and Development Commission Multi-year Expert Meeting on Trade, Services and Development
Fifth session Geneva, 18–20 July 2017 Item 3 of the provisional agenda. https://unctad.org/system/files/official-
document/c1mem4d14_en.pdf
Uygun, O., Kacamak, H., & Kahraman, U. A. (2015). An integrated DEMATEL and Fuzzy ANP techniques for
evaluation and selection of outsourcing provider for a telecommunication company. Computers & Industrial
Engineering, 86, 137–146. doi:10.1016/j.cie.2014.09.014
Zhu, J., H. S. Lean and S. K. Ying (2002): The Third-Party Logistics Services and Globalization of
Manufacturing, in: International Planning Studies, 7 (1), 89-104.
Zhu Z., Hsu K., Liulie J. (2001). Outsourcing – a strategic move: the process and the ingredients for success,
Managing decision, 39/5, 373-378. https://doi.org/10.1108/EUM0000000005473
Zitkiene, R. & Dude, U. (2018). The impact of outsourcing implementation on service companies.
Entrepreneurship and Sustainability Issues. 6. 342-355. 10.9770/jesi.2018.6.1(21).
Zöngür, G., Yılmaz, K. G., Güngördü, A. (2016). Konaklama İşletmelerinde Dış Kaynak Kullanımı: Ankara
İlindeki Dört ve Beş Yıldızlı Otel İşletmelerinde Bir Uygulama. Gazi Üniversitesi İktisadi ve İdari Bilimler Fakültesi
Dergisi, 18(1), 217-235.
Žydžiünaite, V. & Sabaliauskas, S. (2017). Kokybiniai Tyrimai. Principai ir Metodai. Vilnius: Vaga,

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey
381
 MULTI-CRITERIA ASSESSMENT FRAMEWORK FOR FREIGHT
VILLAGES BASED ON OPERATIONAL EFFICIENCY CRITERIA
Burcu Yılmaz Kaya1, Aylin Adem2, Metin Dağdeviren 3

Abstract ⎯ Freight villages can be defined as sites specially established in an area equipped with certain
facilities and instruments with effective connections to all kinds of transportation modes, for the realization of
all logistics-related functions and activities more efficiently and effectively. Freight village concept and different
decision problems regarding to its applications has found place in the related scientific literature since 1960s,
yet, almost all of these studies concerned with optimization of intermodal transportation networks or location
selection, where, a significant gap considering efficiency analyses of freight villages could be addressed in this
point of view. In this study, a new set of operational efficiency criteria was defined based on the extent literature
review to be used in evaluation of business excellence performance potentials of freight villages. Eight
operational criteria were identified in this regards, where, the importance order of these criteria relative to
each other was also analyzed from three different perspectives with linear Best Worst Method (BWM).
Reliability of the results were denoted referring the computed CR values. Comparative analyses were performed
and results were interpreted in details to suggest an infrastructure for field practitioners and scientific
researchers to assess the operational efficiency potentials.

Keywords ⎯ Freight villages, Operational efficiency, Multi-criteria decision making, BWM

INTRODUCTION
The increase in the trade volume brought about by the developing technology and population growth has led to
innovations and developments in the logistics sector. In this direction, freight villages emerged and started to
be implemented first in the world and then in Turkey. Freight villages could be defined as sites established by
the public or private sector where all logistics-related activities were carried out, in order to reduce the negative
effects and to ensure that the logistics sector and companies can compete on both national and international
scales. Logistic villages are obliged to provide the best possible service to the transport and production sector.
Organizations that want to operate in these villages either buy or become tenants of buildings such as
warehouses, bulk cargo centers, storage areas, offices, car parks, and will be customers of freight village
organizations It is expected that all customers in a freight village will be able to easily access the activities they
will need for their operations and be equipped with the necessary facilities, e.g. fire department, customs, gas
stations, charging units, etc. In addition, if possible, these villages could also provide services such as required
qualified workforce and equipment for their customers (Aydın & Öğüt 2008).

In the related national and international literature, the most frequent analyzed problem type considering freight
villages was location selection, while, almost all the studies analyzed that problem in similar dimensions with
intersecting criteria sets including costs, demand and capacities, intermodal transportation convenience,
proximity to special zones, land parameters, and environmental impact criteria. This situation creates a
significant gap in the related literature in regards of efficiency analyses of freight villages, or, assessing the
performance potentials of business excellence of alternatives.

This study defines and analyzes a new set of operational efficiency criteria in the light of extent literature review,
and proposes a framework based on linear Best Worst Method (BWM) importance rankings computed regarding
perspectives of three different freight village parties for performance assessment and decision making problems
regarding both location selection and managerial excellence evaluation.

1
Burcu Yılmaz Kaya, Gazi University, Faculty of Engineering, Department of Industrial Engineering, Ankara, Turkey,
[email protected]
2
Aylin Adem, Gazi University, Faculty of Engineering, Department of Industrial Engineering, Ankara, Turkey, [email protected]
3
Metin Dağdeviren, Gazi University, Faculty of Engineering, Department of Industrial Engineering, Ankara, Turkey,
[email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

382
Remainder of this paper was organized as, ground arguments and research motivation of the study, and, the
literature review on freight villages were presented in Section 2, linear BWM method was presented in Section
3, Section 4 was devoted to real world application and revealed results were interpreted in Section 5, where,
Section 6 concludes the paper and presents possible future directions of current study.

GROUND ARGUMENTS
According to the literature research carried out for the studies between 1999-2021, most frequent analyzed
problem types could be grouped under two headings, freight village location selection (Taniguchi et al., 1999;
Wang & Lui, 2007; Ballis & Mavrotas, 2007; Li et al., 2010; Turskis & Zavadskas, 2010; Elgün, 2011; Can,
2012; Pekin & Macharis, 2013; Demiroğlu & Eleren, 2014; Tadić et al., 2014; Yıldırım et al., 2014; Aksoy &
Özyörük, 2015; Özceylan et al., 2016; Huang et al., 2019; Ma et al., 2020) and concept definition with general
information presentation about freight villages (Akandere, 2013; Tadić et al., 2014). Although they found a
limited number of applications, financing, impact on foreign trade, transportation activities, possible workforce
supply problems were also noteworthy. In addition, while it is seen that different solution methods have been
used, it has been observed that multi-criteria decision making methods were used most frequently (Ballis &
Mavrotas, 2007; Wang & Lui, 2007; Li et al., 2010; Turskis & Zavadskas, 2010; Elgün, 2011; Can, 2012;
Demiroğlu & Eleren, 2014; Tadić et al., 2014; Yıldırım et al., 2014; Özceylan et al., 2016; Huang et al., 2019;
Ma et al., 2020).

Results of the detailed literature review showed that the criteria considered in existing literature were often
focused on cost minimization i.e., transportation and operating costs or tax deductions ratios, or, represented
site selection constraints such as geographical parameters and location's suitability for intermodal transportation.
Additionally, it is seen that the studies examined freight villages with multi-criteria decision making techniques
generally used well known but bothersome methods such as AHP, ANP, DEMATEL or TOPSIS. Methods such
as AHP, ANP, TOPSIS, PROMETHEE, DEMATEL (Ballis & Mavrotas, 2007; Wang & Lui, 2007; Dağdeviren
& Yüksel, 2008; Dağdeviren & Eraslan, 2008; Li et al., 2010; Turskis & Zavadskas, 2010; Yılmaz &
Dağdeviren, 2010; Yılmaz & Dağdeviren, 2011; Can, 2012; Demiroğlu & Eleren, 2014; Tadić et al., 2014;
Yıldırım et al., 2014; Yılmaz Kaya & Dağdeviren, 2016; Özceylan et al., 2016; Tan et al., 2016; Huang et al.,
2019; Kılıç Delice & Can, 2020, Adem et al. 2022) give comprehensive and reliable results, hence, have had
the opportunity to be applied for many years in many different lanes, continue to maintain their popularity today,
on the other hand, there are new methods, such as BWM, which also provide ease of calculation by reducing
the number of pairwise comparisons and in this way avoid inconsistencies that might arise from assessment
biases of decision makers. Some studies from the current literature are summarized in terms of authors and years
of study, addressed problems types, and, employed criteria and solution methods are presented in Table 1.

Table 1. Literature Review Summary

Author & Year Criteria Set Solution Method Problem Type
Taniguchi et al. Transportation costs, Transportation time, GA, Queuing
Location selection
(1999) Intermodal transportation convenience, Demand theory, NLP
Wang & Lui Proximity to natural resourses, Area potential,
F-AHP, TOPSIS Location selection
(2007) Trasnportation costs, Economic/social impact
Traffic density, Transportation hubs, Direct loading,
Ballis & Mavrotas
Space employment ratios, Warehouse dimension, PROMETHEE Location selection
(2007)
Intermodal transportation convenience
Proximity to natural resourses, Geographical
Li et al. (2010) parameters, Investment cost, Expansion possibility, F-TOPSIS, GP Location selection
Information abilities
Turskis & Investment cost, Operation time, Expansion
AHP, F-ARAS Location selection
Zavadskas (2010) possibility, Proximity to market
Intermodal transportation convenience,
Elgün (2011) Geographical parameters, Infrastructure, Workforce Delphi method Location selection
convenience, Security
AHP,
Proximty to special zones, Investment cost,
Can (2012) ELECTRE, Location selection
Expansion possibility, Geographical parameters
TOPSIS

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

383
 Pekin & Macharis Transportation cost, Demand, Intermodal
GIS, LAMBIT Location selection
(2013) transportation convenience
Demiroğlu & Proximty to special zones, Location demographics, AHP,
Location selection
Eleren (2014) Demand, Site capacity, Geographical parameters PROMETHEE
Infrastructure, Intermodal transportation
convenience, Investment cost, Environmental F-DEMATEL, F-
Tadić et al. (2014) City logistic management
impact, Transportation cost, Security, Transportation ANP, F-VIKOR
hubs, Economic impact
Yıldırım et al. Expansion possibility, Investment cost, Proximity to AHP,
Location selection
(2014) special zones PROMETHEE
Aksoy & Özyörük Transportation cost, Demand, Transportation
0-1 LP Location selection
(2015) capacity, Investment cost, Site capacity
Proximity to special zones/natural resourses,
Özceylan et al. GIS, ANP,
Geographical parameters, Investment cost, Location selection
(2016) TOPSIS
Expansion possibility
Demand, Expansion possibility, Intermodal
Huang et al. (2019) Entropy, TOPSIS Facility planning
transportation convenience, National policies
Investment cost, Proximity to special zones,
Ma et al. (2020) CPFI, VIKOR Location selection
Economic impacts
Available total land of use, Alternative land use
Yılmaz Kaya et al. scenarios, Traffic generation/reduction potentials,
BWM Efficiency management
(2021) Transportation cost, Operating cost, Positive impact
on local business, Employment potential

Therefore, in this study, a new set of operational efficiency criteria defined on the basis of business management
excellence, which could be considered in freight village decision making problems was introduced. In addition,
the BWM method, which has not been used in any problem related to freight villages into our knowledge, was
used for the first time to find the importance of the criteria in the identified set in a way that expresses the views
of different parties of freight village concept.

BEST WORST METHOD (BWM)

BWM is a subjective pairwise comparison-based multi-criteria decision making method developed by Rezaei
(2015, 2016). BWM requires the information coming from two different comparison vectors: (i) preference of
the most important criterion over other, (ii) preference of all criteria over the least important criterion. The
method also provides the reliability information related to obtained values by computing the consistency ratio
of comparisons, the consistency could be improved with BWM while it reduces required pair-wise comparisons
in regards with other subjective weighting multi-criteria decision making methods like DEMATEL, AHP or
ANP.

The calculation steps of linear BWM were described herinafter (Rezaei, 2016; Kheybari et al., 2021; Petrudi et
al., 2021).

Step 1. The criteria set including n criteria {c1, c2, ..., cn} was defined.

Step 2. The criteria had the highest importance and the least importance were identified.

Step 3. The preference of criterion with the highest importance over all the other criteria was identified on a
scale of 1 to 9 (the value of “1” means that the importance levels of that criteria pair was to be equal, and, the
value of “9” means that the importance of the criterion with the highest importance is extremely significant to
the compared one). The comparison vector obtained in this regard was called “the best-to-others”, and denoted
as AB= (aB1, aB2, ..., aBj, ..., aBn). Here, aBj indicates the preference value of the criterion B over the criterion j.

Step 4. The preference of all criteria over the criterion with the least importance was identified on the same
scale of 1 to 9. The comparison vector obtained in this regard was called “the others-to-worst”, and denoted as
Aw= (a1W, a2W, ..., ajW, ..., anW). Here, ajW indicates the preference value of the criterion j over the criterion W.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

384
Step 5. The optimal weighting scores of all criteria (w1*, w2*, ..., wn*) and the index of judgement consistency
(ξ*) were calculated by solving the identified linear programming model denoted in (1) - (5) to minimize the
maximum absolute difference of {|wB/wj – aBj|, |wj/wW – ajW|}, for the criteria set including n criteria {c1, c2, ...,
cn}, where j=1, 2, ..., n.

min 𝜉 (1)
s.t.
𝑤𝐵
| − 𝑎𝐵𝑗 | ≤, 𝜉, 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑗 (2)
𝑤𝑗
𝑤𝑗
| − 𝑎𝑗𝑊 | ≤ 𝜉, 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑗 (3)
𝑤𝑊
∑𝑗 𝑤𝑗 = 1 (4)
𝑤𝑗 ≥ 0, 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑗 (5)

Here, the closer the optimal auxiliary variable ξ* values to zero will address higher consistency, and the
acceptability of the initial pair-wise comparison data could be calculated by CR = ξ* /CI, in means of the input-
based consistency ratio (CR) and consistency index (CI) threshold values proposed by Rezaei (2015).
Constraints derived with (2) represents “the comparison of best-to-others” vector, where, constraint derived
with (3) “the comparison of others-to-worst” vector, (4) ensures the sum of optimal weighting scores to be equal
to “1”, and, (5) is positivity constraint.

Interested readers could see the studies of Rezaei et al. (2017), Stević et al. (2017), Rezaei et al. (2018), Azizi
et al. (2019), Bahrami et al. (2019), Govindan et al. (2019), Liang et al. (2019), Liu et al. (2019), Grida et al.
(2020), Kaushik et al. (2020), Kolat et al. (2020), Özmen and Kızılkaya Aydoğan (2020), Sahebi et al. (2020),
Kumar et al. (2021) for further literature on BWM applications.

APPLICATION
A new set of operational efficiency criteria including eight criteria named “C1: Available total land of use”,
“C2: Alternative land use scenarios”, “C3: Traffic generation potential”, “C4: Traffic reduction potential”,
“C5: Transportation cost”, “C6: Operating cost”, “C7: Positive impact on local business”, and “C8:
Employment potential” was defined based on the literature review (Boile et al., 2011; Regmi & Hanaoka, 2013;
Tadić et al., 2014) and experts’ field experience. New criteria set was analyzed in three different perspectives
two as organizational parties and one as the social party with linear BWM; (i) freight village providers, the
organizations those are the real owners or managing companies of the freight village site, (ii) freight village
customers/tenants, the organizations which use freight village activities by renting or purchasing the buildings,
warehouses, equipment and services, and (iii) local residents, who live nearby and would be affected by
economic, social and environmental impacts of the freight village investment.

The linear programming models constituted considering referred eight criteria regarding the judgements of
decision makers representing above mentioned three different perspectives presented in Table 2.

Table 2. DM assessments for freight village operational efficiency criteria

Best/Worst C1 C2 C3 C4 C5 C6 C7 C8
DM1 C6 2 3 4 4 2 1 5 5
Providers DM2 C5 4 5 3 6 1 2 5 9
DM3 C6 4 8 3 3 2 1 7 7
DM1 C5 5 5 2 2 1 2 4 3
Best to others Tenants DM2 C5 6 6 3 3 1 2 5 5
DM3 C6 4 4 3 3 2 1 7 6
DM1 C8 6 4 3 3 5 6 2 1
Residents DM2 C3 6 3 1 2 8 9 2 2
DM3 C7 6 3 2 2 6 7 1 2

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

385
 DM1 C8 4 3 3 3 5 5 2 1
Providers DM2 C8 5 4 4 4 9 8 3 1
DM3 C2 2 1 3 3 7 9 3 3
DM1 C1 1 2 4 4 5 4 3 3
Others to
Tenants DM2 C2 2 1 3 3 8 6 4 4
worst
DM3 C7 3 2 5 5 6 7 1 2
DM1 C6 2 3 4 4 2 1 5 6
Residents DM2 C5 2 5 8 7 2 1 6 6
DM3 C5 2 4 5 5 2 1 7 6

The impact of importance of these three perspectives was also considered by calculating the optimal weighting
scores of different all decision makers related to each party. To explain the computation mechanism of BWM
method in more details, the linear programming model considering the judgements of DM1 regarding freight
village providers side was presented hereinafter (6).

min 𝜉
s.t.
𝑤 𝑤 𝑤 𝑤 𝑤 𝑤 𝑤
| 16 − 2| ≤ 𝜉 , | 16 − 3| ≤ 𝜉 , | 16 − 4| ≤ 𝜉 , | 16 − 4| ≤ 𝜉 , | 16 − 2| ≤ 𝜉 , | 16 − 5| ≤ 𝜉 , | 16 − 5| ≤
𝑤11 𝑤12 𝑤13 𝑤14 𝑤15 𝑤17 𝑤18
𝑤11 𝑤12 𝑤13 𝑤14 𝑤15 𝑤16 𝑤17
𝜉 ,| − 4| ≤ 𝜉 , | − 3| ≤ 𝜉 , | − 3| ≤ 𝜉 , | − 3| ≤ 𝜉 , | − 5| ≤ 𝜉 , | − 5| ≤ 𝜉 , | −
𝑤18 𝑤18 𝑤18 𝑤18 𝑤18 𝑤18 𝑤18
2| ≤ 𝜉 ,

𝑤11 + 𝑤12 + 𝑤13 + 𝑤14 + 𝑤15 + 𝑤16 + 𝑤17 + 𝑤18 = 1,

𝑤1𝑗 ≥ 0, 𝑗 = 1, … , 8 (6)

RESULTS
The importance levels related to each operational efficiency criterion was calculated by solving nine different
linear mathematical modelling systems separately, which were constituted in regards of each decision maker’s
assessments representing each freight village party’s standpoint (linear mathematical modelling system
constituted to analyze the assessments of DM1 regarding freight village providers party was presented in (6)).
Obtained weighting scores and computed consistency ratios regarding the reliability of expert assessments were
denoted in Table 3. As seen from the table, computed CR values being very close to zero proves that the
calculated weighting scores for the new criteria set are consistent. The comparative analysis of the calculated
final weighting scores were also illustrated for further analysis in Figure 1 and Figure 2.

Table 3. Assessment results

Providers Tenants Residents Final Weightings
Criteria
DM1 DM2 DM3 DM1 DM2 DM3 DM1 DM2 DM3 Providers Tenants Residents Overall
C1 0,167 0,097 0,094 0,042 0,063 0,090 0,060 0,054 0,053 0,115 0,062 0,056 0,078
C2 0,111 0,077 0,034 0,061 0,038 0,090 0,091 0,108 0,106 0,067 0,059 0,101 0,076
C3 0,083 0,129 0,125 0,153 0,127 0,120 0,121 0,280 0,159 0,111 0,133 0,175 0,140
C4 0,083 0,065 0,125 0,153 0,127 0,120 0,121 0,162 0,159 0,088 0,133 0,146 0,122
C5 0,167 0,330 0,188 0,258 0,304 0,180 0,072 0,041 0,053 0,218 0,242 0,054 0,171
C6 0,278 0,194 0,325 0,153 0,190 0,304 0,043 0,030 0,034 0,260 0,207 0,035 0,167
C7 0,067 0,077 0,054 0,077 0,076 0,035 0,181 0,162 0,277 0,065 0,059 0,201 0,108
C8 0,044 0,030 0,054 0,102 0,076 0,060 0,311 0,162 0,159 0,042 0,078 0,200 0,107
0,056 0,057 0,051 0,049 0,076 0,057 0,052 0,044 0,042

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

386
As it was represented in Figure 1 and Figure 2, importance levels of the criteria vary with respect to the
standpoint of the regarding party; “C6: Operating cost” was found to be the most important criterion for freight
village providers (0,260), where, for freight village customers/tenants and local residents, criteria “C5:
Transportation cost” and “C7: Positive impact on local business” were determined as the most important ones
(0,242; 0,201), respectively. Although the defined criteria have been chosen from the existing literature as the
most important criteria affecting operational efficiency in general, it has been seen that different parties give
different importance to these criteria when comprehensive structures such as freight villages were in question,
which have more than one side in the application domain. Hence, the criteria set defined in the study were
examined separately on a common basis both in terms of different perspectives and in order to indicate the
general ranking and priorities i.e., criteria importance, overall importance of each criterion was also calculated.
Table 3 also presents that the most important criterion was found to be “C5: Transportation cost” according to
the overall rankings, where all three parties were assumed as having equal dominance on the overall outcomes.
Additionally, overall weighting scores had the best fit for the case, as it was indicated in Figure 2.

Overall Residents Tenants Providers

C8
C7
C6
C5
C4
C3
C2
C1

0 0,05 0,1 0,15 0,2 0,25 0,3

Figure 1. Comparative analysis of criteria importance levels

Providers Tenants Residents Overall

C1
0,3
C8 C2
0,2

0,1

C7 0 C3

C6 C4

C5

Figure 2. Relationship among criteria weighting score output values

CONCLUSION
In this study a new assessment criteria set for operational efficiency analysis of freight villages was identified
in respect with the expert opinions and extent literature reviews. This criteria set was inspected in regards of
different perspectives of three participating parties, and BWM method was employed as a first in freight village
topic to the best of our knowledge. Three scenarios were explored, and according to the results “C5:

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

387
Transportation cost” was found to be the most important one in regards to overall weighting scores (0.171),
and followed by “C6: Operating cost” (0,167), “C3: Traffic generation potential” (0.140), and “C4: Traffic
reduction potential” (0.122). This study did not aim to handle the effects of different sides on the assessment
outcomes which could be seen as a limitation, so, all parties related to freight village applications were assumed
as having equal importance. In the era of sustainability and green applications; different dominance scenarios
could be considered for different affecting sides in future studies; for example, the effects on the ongoing life
could be treated as having more dominance on the overall criteria ranking, or, some different parties representing
environmental affects and resource consumption could be added to the decision making audience. In this case,
the criteria weightings, hence ranking, will likely change, and furthermore, beyond expressing the general
opinion in daily practices, it will also be possible to examine the extent to which these criteria can be taken into
account in the future to contribute sustainability.

Some other suggestions for future directions of this study could be mentioned as regarding a widened set of
criteria including sub-main criteria hierarchy to represent deepen insights related to freight village applications,
where, this study only regards developed single leveled criteria set. A subjective weighting MCDM method was
employed to analyze the importance structure of operational efficiency criteria in this study as another limiting
factor, where, objective MCDM methods such as CRITIC, Entropy, ARAS etc. could be considered to be used
as an alternative future application suggestion.

The criteria set identified and related importance weightings calculated here are suitable to be employed for
investment decisions, location selection, or, benchmarking problems related to the freight villages.

REFERENCES
Adem A., Yilmaz Kaya B., Dağdeviren M. (2022) Technology analysis for Logistics 4.0 applications: Criteria affecting
UAV performances. In: Kahraman C., Aydın S. (Eds) Intelligent and Fuzzy Techniques in Aviation 4.0. Studies in Systems,
Decision and Control, 372. Springer, Cham. doi.org/10.1007/978-3-030-75067-1_21
Akandere, G. (2013). Lojistik köylerin etkin yönetimi: Konya iline yönelik bir model önerisi. MSc Thesis, Selçuk
University, Institute of Social Sciences, Konya, Turkey.
Aksoy, O., & Özyörük, B. (2015). The importance of freight villages: An implementation in TCDD. Applied Mathematical
Modelling, 39(19), 6043–6049. doi:10.1016/j.apm.2015.01.034
Aydın, G. T., & Öğüt, K. S. (2008). Freight villages in Europe and Turkey (Avrupa ve Türkiye’de Lojistik Köyler).
Proceedings of International Railway Symposium 1471-1481, İstanbul, Turkey
Azizi, E., Javanshir, H., Jafari, D., & Ebrahimnejad, S. (2019). Presenting an integrated BWM-VIKOR-based approach for
selecting suppliers of raw materials in the supply chain with emphasis on agility and flexibility criteria (case study: Saipa
corporation). Scientia Iranica, 26(4E), 2601-2614. doi:10.24200/sci.2019.51101.2003
Bahrami, Y., Hassani, H., & Maghsoudi, A. (2019). BWM-ARAS: A new hybrid MCDM method for Cu prospectivity
mapping in the Abhar area, NW Iran. Spatial Statistics, 100382. doi:10.1016/j.spasta.2019.100
Ballis, A., & Mavrotas, G. (2007). Freight village design using the multi criteria method PROMETHEE. Operational
research an International Journal, 7, 2.
Boile, M., Theofanis, S., & Ozbay, K. (2011). University transportation research center: feasibility of freight villages in
the NYMTC region, Rutgers the State University of New Jersey.
Can, A. M. (2012). Selection the location of freight village in Samsun with multi-criteria decision making. MSc Thesis,
Erciyes University, Institute of Science, Kayseri, Turkey.
Dağdeviren, M., & Eraslan, E. (2008). Priority determination in strategic energy policies in Turkey using analytic network
process (ANP) with group decision making. International Journal of Energy Research, 32(11), 1047–1057.
doi:10.1002/er.1418
Dağdeviren, M., & Yüksel, İ. (2010). A fuzzy analytic network process (ANP) model for measurement of the sectoral
competititon level (SCL). Expert Systems with Applications, 37(2), 1005–1014. doi:10.1016/j.eswa.2009.05.07
Demiroğlu, Ş. & Eleren, A. (2015). Küresel lojistik köyleri ve Türkiye'de kurulmasi planlanan lojistik köy bölgelerinin
ÇKKV yöntemleriyle belirlenmesi. Dumlupınar Üniversitesi Sosyal Bilimler Dergisi, 4, 189-202.
Elgün, M. N. (2011). Ulusal ve uluslararası taşıma ve ticarette lojistik köylerin yapılanma esasları ve uygun kuruluş yeri
seçimi. Afyon Kocatepe Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 13 (2), 203-226
Govindan, K., Jha, P. C., Agarwal, V., & Darbari, J. D. (2019). Environmental management partner selection for reverse
supply chain collaboration: A sustainable approach. Journal of Environmental Management, 236, 784-797.
doi:10.1016/j.jenvman.2018.11.088
Grida, M., Mohamed, R., & Zaied, A. N. H. (2020). Evaluate the impact of COVID-19 prevention policies on supply chain
aspects under uncertainty. Transportation Research Interdisciplinary Perspectives, 8 doi:10.1016/j.trip.2020.100240

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

388
Huang, W., Shuai, B., Xu, Y., Zhang, S., & Mao, B. (2019). Railway Express Freight Train Service Sites Planning: A Two-
Stage Entropy-TOPSIS Approach. Transportmetrica A: Transport Science, 12, 5, 807-823.
Kaushik, V., Kumar, A., Gupta, H., & Dixit, G. (2020). A hybrid decision model for supplier selection in online fashion
retail (OFR). International Journal of Logistics Research and Applications, 1-25. doi:10.1080/13675567.2020.1791810
Kheybari, S., Javdanmehr, M., Rezaie, F. M., & Rezaei, J. (2021). Corn cultivation location selection for bioethanol
production: An application of BWM and extended PROMETHEE II. Energy, 228 doi:10.1016/j.energy.2021.120593
Kılıç Delice, E., & Can, G. F. (2020). A new approach for ergonomic risk assessment integrating KEMIRA, best–worst
and MCDM methods. Soft Computing. doi:10.1007/s00500-020-05143.
Kolat, D., Ajlan Kökçü, H., Kiranli, M., Özbiltekin, M., & Öztürkoğlu, Y. (2020). Measuring service quality in the logistic
sector by using Servqual and Best Worst Method doi:10.1007/978-3-030-31343-2_61
Kumar, A., Mangla, S. K., Kumar, P., & Song, M. (2021). Mitigate risks in perishable food supply chains: Learning from
COVID-19. Technological Forecasting and Social Change, 166 doi:10.1016/j.techfore.2021.120643
Liang, F., Brunelli, M., & Rezaei, J. (2019). Consistency Issues in the Best Worst Method: Measurements and Thresholds.
Omega, 102175. doi:10.1016/j.omega.2019.102175
Li, Y., Lui, X., & Chen, Y. (2010). Selection of logistics center location using axiomatic fuzzy set and TOPSIS
methodology in logistics management. Expert Systems with Applications, 38(6): 7901-7908.
Liu, A., Ji, X., Lu, H., & Liu, H. (2019). The selection of 3PRLs on self-service mobile recycling machine: Interval-valued
Pythagorean hesitant fuzzy best-worst multi-criteria group decision-making. Journal of Cleaner Production, 230, 734-750.
doi:10.1016/j.jclepro.2019.04.257
Ma, X., Akram, M., Zahid, K., & Alcantud, J. C. R. (2020). Group decision-making framework using complex Pythagorean
fuzzy information. Neural Computing and Applications. doi:10.1007/s00521-020-05100-5
Özceylan, E., Erbaş, M., Tolon, M., Kabak, M., & Durğut, T. (2016). Evaluation of freight villages: A GIS-based multi-
criteria decision analysis. Computers in Industry, 76, 38–52. doi:10.1016/j.compind.2015.12.003
Özmen, M., & Kızılkaya Aydoğan, E. (2020). Robust multi-criteria decision making methodology for real life logistics
center location problem. Artificial Intelligence Review, 53(1), 725-751. doi:10.1007/s10462-019-09763-y
Pekin, E., & Macharis, C. (2013). A GIS-based analysis of the potential for freight villages in Turkey. World Review of
Intermodal Transportation Research, 4(2/3), 157. doi:10.1504/writr.2013.058980
Petrudi, S. H. H., Ahmadi, H. B., Rehman, A., & Liou, J. J. H. (2021). Assessing suppliers considering social sustainability
innovation factors during COVID-19 disaster. Sustainable Production and Consumption, 27, 1869-1881.
doi:10.1016/j.spc.2021.04.026
Regmi, M. B., & Hanaoka, S. (2013). Location analysis of logistics centres in Laos. International Journal of Logistics
Research and Applications, 16(3), 227–242. doi:10.1080/13675567.2013.812194
Rezaei J. (2015). Best-worst multi-criteria decision-making method. Omega, 53: 49-57.
Rezaei J. (2016). Best-worst multi-criteria decision-making method: some properties and a linear model. Omega, 64: 126-
30.
Rezaei, J., Hemmes, A., & Tavasszy, L. (2017). Multi-criteria decision-making for complex bundling configurations in
surface transportation of air freight. Journal of Air Transport Management, 61, 95-105.
doi:10.1016/j.jairtraman.2016.02.006
Rezaei, J., van Roekel, W. S., & Tavasszy, L. (2018). Measuring the relative importance of the logistics performance index
indicators using best worst method. Transport Policy, 68, 158-169. doi:10.1016/j.tranpol.2018.05.007
Sahebi, I. G., Masoomi, B., & Ghorbani, S. (2020). Expert oriented approach for analyzing the blockchain adoption barriers
in humanitarian supply chain. Technology in Society, 63 doi:10.1016/j.techsoc.2020.101427
Stević, Ž., Pamučar, D., Zavadskas, E. K., Ćirović, G., & Prentkovskis, O. (2017). The selection of wagons for the internal
transport of a logistics company: A novel approach based on rough BWM and rough SAW methods. Symmetry, 9(11)
doi:10.3390/sym9110264
Tadić, S., Zečević, S., & Krstić, M. (2014). A novel hybrid MCDM model based on fuzzy DEMATEL, fuzzy ANP and
fuzzy VIKOR for city logistics concept selection. Expert Systems with Applications, 41(18), 8112–8128.
doi:10.1016/j.eswa.2014.07.02
Tan, M., Adem, A. & Dağdeviren, M. (2016). Using the analytic hierarchy process (AHP) for supply chain risk
management (SCRM): A case study in automotive sector. LM-SCM XIV International Logistics and Supply Chain
Congress, 191-202.
Taniguchi, E., Noritake, M., Yamada, T., Izumitani, T. 1999), Optimal size and location planning of public logistics
terminals, Transportation Research Part E: Logistics and Transportation Review,35, 3, 207-222.
Turskis, Z., & Zavadskas, E. K. (2010). A novel method for multiple criteria analysis: Grey Additive Ratio Assessment
(ARAS-G) method. Informatica, 21(4), 597-610. doi:10.15388/Informatica.2010.307
Yildirim, B., F., & Önder, E. (2014). Evaluating potential freight villages in Istanbul using multi criteria decision making
techniques. Journal of Logistics Management, 3(1): 1-10
Wang, S., Lui, P. (2007). The evaluation study on location selection of logistics based on fuzzy AHP and TOPSIS.
International Conference on Wireless Communications, Networking and Mobile Computing, 3779-3782, doi:
10.1109/WICOM.2007.935.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

389
Yılmaz, B., & Dağdeviren, M. (2010). Comparative analysis of PROMETHEE and fuzzy PROMETHEE methods in
equipment selection problem. Journal of the Faculty of Engineering and Architecture of Gazi University. 25, 4, 811-827,
doi: 10.17341/gummfd.80883.
Yılmaz, B., & Dağdeviren, M. (2011). A combined approach for equipment selection: F-PROMETHEE method and zero–
one goal programming. Expert Systems With Application. 38, 11641–11650, doi: 10.1016/j.eswa.2011.03.043.
Yılmaz Kaya, B., & Dağdeviren, M. (2016). selecting occupational safety equipment by MCDM approach considering
universal design principles. Human Factors and Ergonomics in Manufacturing & Service Industries. 26, 2, 224-242,
doi:10.1002/hfm.20625.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

390
 IMPROVEMENT OF THE PRODUCTION EFFICENCY WITH LEAN
MANUFACTURING TECNIQUES
Eray Türkmen1, Elif Tuyan2, Orhun Gökdemir3, Majed Alkhatib4, Kübra Nur Şahin5,*, İsmet Söylemez6

Abstract - In manufacturing, material flow is the one of the essential components of production processes. A smooth
material flow helps to reduce material transportation, storing, and handling costs. Also, it reduces waste, energy
consumption, and space usage. In that study, the problem of designing an in-house logistics system for supplying
materials to workstations is considered to assure efficient manufacturing processes for a soft drink company in
Turkey. We address the challenges of redesigning the layout of the production system and the timely delivery and
storage of material with the objective of improving resource-efficiency as well as reducing traffic related to delivery
of material to production line. Production layout and the corresponding material transports and storage systems
are reorganized with lean manufacturing techniques on the operational level. This study is especially focusing on
implementing lean manufacturing techniques at the soft drink company to increase productivity and reduce wastes.
In the study, root causes of the problem were determined using lean manufacturing techniques. The suggested
approach is tested on real data. The results are analyzed to illustrate the value of the proposed approach for the
design of shelves system of the logistics processes. This study is motivated by a real case study and addresses the
material supply and layout design problems at production lines.

Keywords: 5S, in-house logistic, kaizen, lean manufacturing, material handling

INTRODUCTION
Lean manufacturing is one type of the manufacturing process which is the most popular currently especially at the
in-house logistic. In lean manufacturing, the first aim is to reduce waste and increase value-adding activities while
decreasing non value-adding activities and wastes activities. To maximize the profit, the companies must do some
implementation to improve the production efficiency of the factories. An increase in the quality of the production
and the variation of the products leads to an enhancement of the income of the company. During these processes, if
a halting about technological devices and the production process occurs due to technical reasons, it causes some
losses in revenue of the company.

Thanks to lean techniques, such problems can be easily avoided. Even necessary precautions are taken before
problems arise. A more efficient process is achieved with a lean production system in which all processes are
included. By providing all employees with a perspective of continuous improvement, the problems in the processes
are minimized before they occur. Implementation of lean techniques has gained importance since it has achieved
an outstanding success in production processes. There is a huge difference between lean production and traditional
production systems. In traditional systems parts feeding and material handling activities are performed in
disorganized way. The disorganization causes wastes in production area.
1 Eray Türkmen, Abdullah Gül University, Faculty of Engineering, Department of Industrial Engineering, Kayseri, Turkey,
[email protected]
2 Elif Tuyan, Abdullah Gül University, Faculty of Engineering, Department of Industrial Engineering, Kayseri, Turkey, [email protected]
3 Orhun Gökdemir, Abdullah Gül University, Faculty of Engineering, Department of Industrial Engineering, Kayseri, Turkey,

[email protected]
4 Majed Alkhatib, Abdullah Gül University, Faculty of Engineering, Department of Industrial Engineering, Kayseri, Turkey,

[email protected]
5, * Kübra Nur Şahin, Abdullah Gül University, Faculty of Engineering, Department of Industrial Engineering, Kayseri, Turkey,

[email protected], corresponding author

6 İsmet Söylemez, Abdullah Gül University, Faculty of Engineering, Department of Industrial Engineering, Kayseri, Turkey,

[email protected]

391
In this study, it is aimed to eliminate unnecessary transports observed in production processes by using lean
techniques. First, the current situation analysis was made and the root causes of the problems were revealed. Then,
proposed solutions were made for the identified problem. Thanks to lean techniques, the process has been
standardized and waste has been eliminated. In-house logistics transportation is minimized. The problem of
designing an in-house logistics system for supplying materials to workstations is considered to assure efficient
manufacturing processes for a soft drink company in Turkey.

The reminder of this paper is organized as follows. After the section of the introduction, the literature review is
given and problem definition is described. The results obtained are included in the next section. Finally, conclusions
are presented in the last section.

LITERATURE REVIEW
The objective of this article is to understand the concept of lean manufacturing, its philosophy, various tools and
techniques, lean implementation benefits and barrier towards lean implementation. There are lots of studies about
lean manufacturing applications in the literature. Some of them are as follows.

The fishbone method clearly shows all the factors that cause a problem to occur and the relationship between the
causes. Yazdani and Tavakkoli (2012) state that they showed the reasons for the problem by using the fishbone
method, and after determining the effects of these reasons, they reached three main solutions. The importance of
determining the causes before solving the problem is significant to reach the correct solution. After learning the
main causes of the problem using the fishbone diagram, the right solutions to solve the problems in the ayran filling
area was proposed.

Bhattacharya et al. (2014) examine the Root Cause Analysis for the manufacturing industry. This analysis aims to
investigate the immediate problem to understand underlying causes. The Root Cause Analysis benefits the producer
in several ways, such as identifying permanent solutions, preventing the recurring of the problem, higher customer
satisfaction. Serrat (2017) explains the Five Whys Technique, which explores the cause-and-effect relationship by
simply asking questions. In this technique, the question "why" is asked to the problems and repeatedly to the results
five times.

According to Arslankaya and Atay (2015), inefficiency problems can be avoided by just implementing some
improvements and scheduling the processes. Authors presents some information about the application of
maintenance management and the lean manufacturing techniques in the plants to prevent the losses that occurred
by some unexpected break downs and also improve the quality of the work and increase the motivation level of the
workers at a plant which produces dairy products such as milk, butter, cheese, yogurt, ayran, etc. Problems are the
lack of communication when the employer encounters a breakdown during working and the layout of the shelf at
the production area. The solutions to these problems are to create a maintenance system that is computerized and to
classify the shelf system by using 5S implementation and lean manufacturing techniques. Jiménez et al. (2015)
provides 5S methodology implementation experiences in engineering labs. Thanks to the 5S method; learning,
control and maintenance of the resources and activities involved are performed in less time and with a considerable
reduction of cost.

Ostaev et al. (2019) aim to introduce advanced and necessary methods of milk production accounting and
management into commodity cooperative structure enterprises. To solve to problem, the company has created a
group of workers for the organization of the lean manufacturing.

Further research conducted by Janjić et. al. (2019) illustrates Kaizen approach to be useful to implement in different
industries. Moreover, several factors influence kaizen success, depending on the area of application. Authors
elaborate on the need of communication development to be involved in kaizen process. García et al. (2013) elaborate
on the significance of kaizen approach to be an effective approach in many working fields. The research focuses on
evaluating factors that affect kaizen efficiency by conducting a questionnaire. García et al. (2013) indicate that

392
implementing 5S is a useful strategy to for an effective kaizen implementation, emphasizing that integrating two
approaches at a time can help develop more productive workplace. In addition, standardizing operations and
clarifying guidelines and procedures are significant factors to positively influence kaizen process. Moreover, Arya
and Jain (2014) describe Kaizen to be a process that considers the system. In addition, workers and individuals are
also involved in the process as they can be the base for a successful kaizen.

Antony et al. (2020) were established the practical implications of lean manufacturing studies carried out worldwide
and identifies novel research streams. They considered more than 400 articles published between 2010 and 2019.
Value stream mapping(VSM) is the one types of the lean manufacturing techniques. VSM was applied in the
automotive manufacturing organization in India (Singh, and Singh 2020).

In order to tackle problems in a printer assembly company, the current value stream mapping was drawn and
according to the observations Kanban system applied in the printer assembly company by Lin et al. (2020).

Palange, A., & Dhatrak, P. (2021) present the review article which gave the detailed information about different
manufacturing sectors (automotive, electronics, plastic, textile, food, dairy, foundry, stampings, maintenance)
applied which types of the lean techniques such as 5S, Kanban, value stream mapping, SMED (Single-Minute
Exchange of Die), Kaizen. Psomas, E. (2021) prepared a systematic literature review. Author considered articles of
publishing houses which are Emerald, Science Direct/Elsevier, Taylor and Francis, Springer Link about the lean
manufacturing techniques. Also, author classified different ways such as geographic research areas, article types.
In the investigation of the connections between lean practices and financial performance measures are given by
Dieste et al. (2021). They focused on the financial performance metrics for the application of the lean manufacturing
techniques.

Morover, many studies have been done which use the lean manufacturing techniques. Gubta and Jain (2013), Jasti
and Kodali (2013), Hartini and Ciptomulyono (2015), Cherrafi et al. (2016), Rewers et al. (2016), Negrão et al.
(2017), Psomas and Antony (2018), Palange and Dhatrak (2021), Maware et al. (2021) have reported the studies in
the literature reviews. These given reviews may be examined for the detailed information about the lean techniques
and implementation areas.

PROBLEM DEFINITION
It is important to make observations during the process to determine and understand the problems in a production
stage. After a detailed observation of the production process in the soft drink company, several problems are
determined and examined in the filling procedure of the Ayran. Two filling machines runs by an operator, and the
operator fills the product during the day according to the orders. The first problem during the process is the
transportation of packaging materials from the warehouse to the filling area. The operator stops the machine to bring
packaging materials several times during the day. Therefore, a significant amount of time is loosed because of not
making preparation beforehand. Another one is disorganization/mess around the machines which causes loss of
time during the process (see in Figure 1). As a result of the observations, it was determined that there was a waste
of time for material handling systems. In the Figure 1, the current situation of the process is given.

This paper aims to find solutions to the problems which negatively impact the procedure of the product filling.
Firstly, the production area should be organized carefully. Secondly, the materials which will be needed during the
day should be stored around the machines right amount and the right time. Therefore, a layout plan should be
proposed for a suitable place to keep materials. Thirdly, the filling process should be standardized and the reasons
that end up with stopping the machine should be minimized. Furthermore, better conditions should be provided to
the operator and the problems of the machines that cause errors should be solved.

After the application of several methodologies to solve the problems in the filling process, the success of the
applications will be measured by several aspects such as standard time, the number of ayrans filled during the day,

393
total completion times of the orders, and the total amount of time loosed during the day or month. Time study will
be held in the current system and after application as well to compare the results.

Figure 1: Ayran Filling Machine

PROPOSED METHODOLOGY
Understanding of the current situation time study analyse has be done. Time study was utilized to measure the
improvement and efficiency of method implemented in the firm. For the ayran packaging process, the time study
was conducted in two separate ways. One of the studies was held on while the machine is constantly giving errors,
while the other one was held when the machine gives no error. In this way, the loss of time caused by the machine
errors can be calculated efficiently. Each study was divided into 7 elements which are the sub-processes of ayran
packaging. When those 7 processes are done, 3 packages of ayrans are replaced onto the palette. Since each palette
carries 100 packages of ayrans, it required 33 cycle to fulfil a palette. Standard time of the packaging procedure was
calculated. After the packaging process, the fulfilled palette is transferred to the dispatch room. Another time study
was conducted to measure its standard time.
The important issue for the ayran filling process is bringing package materials from the warehouse. The time was
calculated when operator went to the warehouse. The nonvalue-added time wasted for package transfer from
warehouse was calculated as 20.05 minutes daily. In total, 124 minute is observed as non-value-added time in a
day. However, in this study, we will suggest to prevent the only nonvalue-added time wasted for package transfer
from warehouse. For the wasted time about package transfer which makes 441 minutes in a month, and it
corresponds approximately 4.55% percent of ayran filling process (the factory works for 440 minutes excluding
breaks).
To clearly define root issues and problems, fishbone diagram used as it provides useful visualization of which will
enable to identify potential factors that cause or affect a certain process. Such tool is useful to determine the causes
of disorganization in the area around filling machine.

394
 Figure 2: Fishbone Diagram
Implementing such analysis is useful for finding root cause of the problem and tackle it. To find the root cause of
time waste in the two filling machines, 5Why analysis should be implemented. Potential 5Why to answer are:

Figure 3: 5Why Diagram

 Why there is waste of time in Ayran filling machines? Because the worker stops filling machines to
bring packaging materials.
 Why does the worker bring packages instead of storing it in close area? Because this will lead to more
disorganization.
 Why this will lead to disorganization? Because that will make packaging material temporarily stored
there on the ground which will be as obstacles in the workplace.
 Why are packaging material stored on the ground? Because there is no proper storage area close to
machines such as shelves.
 Why there is no proper storage area? Because the responsible employee does not believe such effect
influences productivity.

395
Answering those questions made the root problem clearer. Accordingly, it is obvious that considering this project
can be an indicator that the last cause is no longer correct as the purpose of this project is to resolve workplace issue.
Further issues will be addressed in the upcoming period.
A considerable procedure to focus on which can improve by applying lean manufacturing techniques such as 5S
and kaizen. According to figure included previously, it is obvious that workplace needs to be cleaned and organized,
for this circumstance, 5S technique should be implemented to solve those issues. Moreover, kaizen approach should
be applied to establish guidelines and standardize operations.
Initially, 5S method is developed to resolve issues regarding the workplace for Ayran filling machine, as follows:
1- Sort: eliminate unnecessary items from workplace.

Figure 4: Current and Proposed Layout

Current and proposed layout is given in the Figure 4. Also movements of the operator are given in the Figure 5.

Figure 5: Movement of the Operator Before and After Sorting the Area.
2- Set in order: arrange tools and items that is used in filling machine, such as cups and trays.
To arrange the necessary tools and items in workplace, shelve should be placed near the working area, as shown in
Figure 6. Shelve design is provided below based on the sizes of items and tools that will be stored on it.

396
 Figure 6: Visualisation of the Shelve to be Placed in Workplace.
3- Shine: Keep workplace and machine clean and neat.
4- Standardize: Build consistent guidelines and instructions operations used in filling machines with a
Standard Operation Procedure (SOP) study.
Table 1: Standard Operation Procedure in Filling Machine
Procedure Key Points

Preparation-Setup for the processes Check the transferred ayran, bring materials from
1 warehouse, set-up the machine.

Package the ayrans and put them onto palette and repeat Fill the plastic plate with ayrans, put 3 filled plates to
2 it until palette is full (33 times) the palette.

When the palette is full transfer the palette to the Make sure the palette is well balanced and replaced in
3 dispatch room and bring an empty one. a proper place of dispatch room.

5- Sustain: Repeat effort to improve workplace and continuously measure performance.

Before-After Kaizen
Before: The workspace of the soft drink company is examined and the Ayran filling process area is analysed. The
working area is unorganized and messy because the boxes are everywhere in the production area. This leads to
some problems while the employees work such as unnecessary movements. While Ayran is filled in the machine,
the worker stands still while performing the job, which requires more physical effort and might lead to less
productivity.
After: After analysing the problem, it is suggested that shelve system can be used for the boxes, tables and palette
can be placed in a specific area for both machines. Also, the shelves should be replaced close to the machine as
much as possible. The reason behind for that when the workers need a raw material, they can reach the materials
easily and quickly. Additionally, to remove the excessive physical effort spent by worker, a chair shall be placed
next to working table.

397
 Figure 7: Before-After Kaizen

CONCLUSION
It is essential to be prepared for all the processes beforehand and assure the readiness of operations. Otherwise,
production may face issues that lead to inefficiency and loss in time and money. Therefore, it is important to
standardize the production process to minimize time waste and maintain quality. This study is especially focusing
on implementing lean manufacturing techniques at the soft drink company to increase productivity and reduce
wastes. The current system needs immediate strategies to be implemented to resolve time waste and disorganizations
issues in the workplace. This study aims to analyse the filling machines at the firm by techniques such as fishbone
diagram and 5Why and solve problems that reduce related efficiency by lean tools such as 5S and Kaizen.
With lean techniques, integration and coordination of logistics and material handling operations activities increase
the efficiency of the system. In this paper, lean manufacturing approaches have been used for in-house logistics
activities. The in house logistics costs in the filling area is reduced with designed shelve system. With this shelve
system, the optimal quantity of each material and the appropriate frequencies have been determined so just in time
production approach is encouraged in the firm. The new shelves system is reduced disorganized in the production
area. As a result, the total in house transportation time waste that needs to be taken out from the filling process is
reduced 4.5%.
At the future works, machine failure causes can be found and improvements can be made. In this way, inefficiency
caused by machine failures can be eliminated. And also there are some limitations; firstly, optimum stock levels for
the required amount of empty boxes are not considered. In addition, if it is determined from which product to be
produced in which time interval by taking into account the daily schedule, a just-in-time production system can be
provided by preparing which type of product when and in the required time at the future studies.

398
 REFERENCES
Antony, J., Psomas, E., Garza-Reyes, J. A., & Hines, P. (2020). Practical implications and future research agenda of lean
manufacturing: a systematic literature review. Production Planning & Control, 32(11) 889-925.

Arslankaya, S., & Atay, H. (2015). Maintenance management and lean manufacturing practices in a firm which produces dairy
products. Procedia-Social and Behavioral Sciences, 207, 214-224.

Arya, A. K., & Jain, S. K. (2014). Impacts of Kaizen in a small-scale industry of India: a case study. International Journal of
Lean Six Sigma, 5(1), 22-44.

Bhattacharya, J., Pharmaceutics, M. P., & Hrm, M. B. A. (2014). Root Cause Analysis–A Practice to Understanding and Control
the Failure Management in Manufacturing Industry. International Journal of Business and Management Invention, 3(10), 12-
20.

Cherrafi, A., Elfezazi, S., Chiarini, A., Mokhlis, A., & Benhida, K. (2016). The integration of lean manufacturing, Six Sigma
and sustainability: A literature review and future research directions for developing a specific model. Journal of Cleaner
Production, 139, 828-846.

Dieste, M., Panizzolo, R., & Garza-Reyes, J. A. (2021). A systematic literature review regarding the influence of lean
manufacturing on firms' financial performance. Journal of Manufacturing Technology Management. 32(9),101-121.

García, J. L., Rivera, D. G., & Iniesta, A. A. (2013). Critical success factors for Kaizen implementation in manufacturing
industries in Mexico. The International Journal of Advanced Manufacturing Technology, 68(1), 537-545.

Gupta, S., & Jain, S. K. (2013). A literature review of lean manufacturing. International Journal of Management Science and
Engineering Management, 8(4), 241-249.

Hartini, S., & Ciptomulyono, U. (2015). The relationship between lean and sustainable manufacturing on performance:
literature review. Procedia Manufacturing, 4, 38-45.

Janjić, V., Todorović, M., & Jovanović, D. (2020). Key success factors and benefits of Kaizen implementation. Engineering
Management Journal, 32(2), 98-106.

Jasti, N. V. K., & Kodali, R. (2014). A literature review of empirical research methodology in lean manufacturing. International
Journal of Operations & Production Management, 34 (8), 1080-1122.

Jiménez, M., Romero, L., Domínguez, M., & del Mar Espinosa, M. (2015). 5S methodology implementation in the laboratories
of an industrial engineering university school. Safety science, 78, 163-172.

Lin, C. K. X., Helmi, S. A., Hisjam, M., & Ali, A. (2020, April). Implementation of lean manufacturing in an electronic
assembly company. In AIP Conference Proceedings (2217(1), 030198). AIP Publishing LLC.

Maware, C., Okwu, M. O., & Adetunji, O. (2021). A systematic literature review of lean manufacturing implementation in
manufacturing-based sectors of the developing and developed countries. International Journal of Lean Six Sigma,
https://doi.org/10.1108/IJLSS-12-2020-0223.

Negrão, L. L. L., Godinho Filho, M., & Marodin, G. (2017). Lean practices and their effect on performance: a literature
review. Production Planning & Control, 28(1), 33-56.

Ostaev, G. Y., Tihonova, A. V., Rylova, N. I., Tihonova, M. S., Malikova, A. M., Karimova, N. A., & Sokolova, I. N. (2019).
Lean production management: Accounting and cost control in processing enterprises of the consumer cooperation
system. Amazonia Investiga, 8(23), 830-841.

Palange, A., & Dhatrak, P. (2021). Lean manufacturing a vital tool to enhance productivity in manufacturing. Materials Today:
Proceedings, 46(1), 729-736.

399
Palange, A., & Dhatrak, P. (2021). Lean manufacturing a vital tool to enhance productivity in manufacturing. Materials Today:
Proceedings. 46(1), 729-736.

Psomas, E. (2021). Future research methodologies of lean manufacturing: a systematic literature review. International
Journal of Lean Six Sigma. https://doi.org/10.1108/IJLSS-06-2020-0082.

Psomas, E., & Antony, J. (2019). Research gaps in Lean manufacturing: a systematic literature review. International Journal
of Quality & Reliability Management, 36(5), 815-839.

Rewers, P., Trojanowska, J., & Chabowski, P. (2016). Tools and methods of Lean Manufacturing-a literature review.
In Proceedings of 7th International Technical Conference Technological Forum (28-30).

Serrat, O. (2017). The five whys technique. In Knowledge solutions (pp. 307-310). Springer, Singapore.

Singh, J., & Singh, H. (2020). Application of lean manufacturing in automotive manufacturing unit. International Journal of
Lean Six Sigma. 11(1), 171-210.

Yazdani, A. A., & Tavakkoli-Moghaddam, R. (2012). Integration of the fish bone diagram, brainstorming, and AHP method
for problem solving and decision making-a case study. The International Journal of Advanced Manufacturing
Technology, 63(5), 651-657.

400
SUPPLY CHAIN RISK MANAGEMENT: A LITERATURE REVIEW ON
DEFINITION, THEORY, AND THE ROLE OF SIMULATION AND
EXISTING OPTIMIZATION METHODS

Gizem Topcu1, Oğuzhan Çağıl 2, Yasutay Susoy 3, Banu Çalış Uslu 4

Abstract ⎯ Supply chain risk management (SCRM) aims to identify, evaluate, mitigate the effects, or seize
opportunities, if any, of unexpected events/situations/conditions (ESC). These ESCs may occur at any stage
and at any time, from the raw material to the final product, and often adversely affect the activities in the
process. The purpose of this study is to present a comprehensive review of the supply chain literature
addressing SCRM-related issues using optimization methods. First, simulation and optimization methods
were analyzed with the parameters taken into account in solving the problems. Then, critical challenges
were explored specifically for industries to identify future research directions. With the findings from the
research, scholars, supply chain experts, and managers can choose the most appropriate optimization
method for the problems they face, depending on the expected output, and identify new areas for their future
work.

Keywords ⎯ Optimization, Risk Management, Simulation, Supply Chain Risk Management

INTRODUCTION
The supply chain includes all processes composed of similar or different types of activities within from
the procurement of raw materials (RM) to up to the after-sales services of the final product (Munir et al.,
2020; Danacı et al., 2019). As a result of globalization, companies have had to shorten product life cycles,
ensure that the right amount of goods and services are supplied and delivered at the right time, at the most
cost-effective and highest quality. Moreover, they have also had to develop responsive and adaptable
systems to manage uncertainty in the market and offer high levels of customization.
These challenges, caused by a high uncertainty in the supply chain, have forced companies to establish
a robust SCRM system that will reduce their vulnerability to risks in their supply chain. However, these
risks can be faced in many ways, and if they are not managed properly, they can cause disruptions or poor
operational performance.
Generally accepted definition of risk is the sources of the crisis for an organization (Natarajarathinam
et al., 2009) or the activities likely to produce undesirable results (Giannakis and Papadopoulos, 2016).
However, considering the positive aspects of risk and possible gains will gain more importance day by day.
Research with many different theoretical perspectives is available in the literature to define and understand
the risks in the supply chain (Dubey et al., 2017; Ben-Daya et al., 2019; Baryannis, et al., 2019).
One of the most widely used methods for assessing current risks is the L matrix method (Özfirat et al.,
2017). Depending on this method, the magnitude of the risk is measured by multiplying the probability and
severity. According to the result from the calculation, risks that will cause high loss if they occur are major
(significant) risks.
Optimization methods are frequently used in SCRM to mitigate risks and suggest the best risk
management strategies (Oliveria et al., 2019; Hamdi et al., 2018; Baryannis, et al.,2019). Especially in the
optimization of complex and stochastic processes, simulation-based solutions can produce better results

1 Gizem Topcu, Marmara University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]
2 Oğuzhan Çağıl, Marmara University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]
3 Yasutay Susoy, Marmara University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]
4 Banu Çalış Uslu, Marmara University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Turkey,
[email protected]
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

401
than alternatives (Avcı& Selim, 2018; Oliveria et al., 2019). Simulation can only be used to monitor the
current state of the system without any optimization, as well as help to provide a well-defined decision
support system with scenario analysis that is used to optimize the bottlenecks in stochastic processes (Yin
& McKay, 2018; Calis, 2016; Şen et al., 2019).
There are plenty of studies in the literature about SCRM and optimization approaches. However, no
research has been found that examines optimization approaches with the parameters considered in problem-
solving and defines the methods used in each step of the risk management process separately. The research
objectives of the study are summarized below
RQ1: What are the current research methodologies regarding the SCRM,
RQ2: What is the role of simulation and optimization for SCRM, and what parameters should be
considered in optimizing risk management processes?
RQ3: What are the challenges, potential research gaps, and future directions within the scope of the
SCRM system?
In this study, firstly, SCRM is examined in terms of definitions and theories. Then, commonly used
parameters and methodologies used to solve the problem were investigated. Based on the findings, a general
SCRM framework was prepared. Next, challenges and future directions are discussed. Eventually,
conclusions are stated.

METHODOLOGY
Many search engines can be used to access scientific articles using many different publishers and
journals. The methodology that is used in this study is illustrated in Figure 1.
In this research, google scholar was used for initial access to the articles, then the following steps were
applied in the analysis of the relevant literature.
1. To define relevant articles; SCRM was investigated under four basic phases: identification,
assessment, management, and monitoring.
2. Key words are identified (Supply Chain, Supply Chain Risk Management, Supply Chain Risk
Management Methods, Supply Chain Risk Simulation, SCRM Simulation Approaches)
3. Total of 201 articles were obtained after the search. articles that did not have complementary
information in terms of content were excluded.
4. In order to reveal the current approaches, optimization parameters, and solution methods of the
problem, 129 articles were examined.
5. Based on the research findings, a relational framework is presented that addresses the SCRM stages
and applicable methodologies

Literature Definition Development of

survey to identify the a holistic framework
concept of SCRM of critical
characteristics and
methods
Figure 1. Framework of Methodology

LITERATURE REVIEW
Since supply chain risk management is vital for companies to increase their level of competition and
maintain their existence, plenty of studies, including different solution methodologies, have emerged (Šoško
et al., 2019). Nowadays, sustainability (social, environmental, economic, technical, and organizational
risks) is among the most widely used methodology (Zineb et al., 2017; Giannakis & Papadopoulos, 2016;
Abdel-Basset & Mohamed, 2020). These studies have drawn attention to the potential of companies to
experience potential risks at high impact levels in a supply chain management model developed without
considering sustainability factors. Sustainable supply chain risk management includes social,
environmental, economic, technical, and organizational risks at the micro and macro level. This
management approach, developed by organizations that produce products and services, involves a green
supply chain, ecological and economic performance management models, and natural and social effects of
risks (Moktadir et al., 2021; Rostamzadeh et al., 2018). In addition to sustainability, there are also different
methodologies applied to SCRM such as value-oriented (Trkman et al., 2016), driven by risk matrix (Qazi
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

402
et al., 2018), text-mining (Chu et al., 2020), Strategic Sustainability (Schulte et al., 2020) and the others
(Shojaei & Haeri, 2018; Tepe Küçükoğlu, 2020; Kırılmaz & Erol, 2016; Demir & Kayali, 2020; Buran &
Aağca, 2019).
Within the scope of this research, the studies in the literature were primarily classified under four
basic risk management steps, which are.
1. Step 1: SCR Identification,
2. Step 2: SCR Assessment,
3. Step 3: SCR Management, and
4. Step 4: SCR Monitoring
Afterward, the optimization methods and the relevant parameters are summarized in Table 1.
Step 1-Supply Chain Risk Identification
Risk identification includes identifying potential sources of risks and related uncertainties within the
SC network (Rajagopal et al., 2017). Risk identification is the prerequisite of the process of assessment and
reducing the effects of risks in the entire network from source to return that may arise in the achievement
of the organization's purpose and the favorable possibility that can be revealed (Wijaya,2021).
Many approaches can be used when defining risk. Within the scope of this research, it is
recommended to classify risks primarily with the relationship diagram shown in Figure 1. It is thought that
the application of this classification will ensure defining risks in a way that includes each process in the
entire SC network based on company objectives.
After this classification, they can be classified according to their effects on the process if the risk
occurs. Raspotnig and Opdahl (2013) defined the risks that may occur in computer-based systems under six
different headings below;
1. Harm 4. Failure
2. Hazard 5. Error
3. Threat 6. Fault

Another study is conducted by Beltrão et al. (2019). They have used eight different classifications to
prioritization of construction risks in Brazilian public enterprises. These risks class are;
1. Environmental 5. Social
2. Construction 6. Economic
3. Political 7. Political
4. Project 8. Management

Figure 2. Supply chain risk identification

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

403
 Simulation is a type of powerful methodology commonly used for the evaluation, monitoring and
modeling part of the supply chain management system. Simulation approaches are not used much in SCR
Identification because systems are tested rather than diagnosed through simulation.
Step 2-SCR Assessment
The risk identification phase is critical to specify which of the risks that may be encountered in the
SC process is assessable. The effectiveness of solutions that can be applied to the risks identified in the Risk
assessment is examined. Kern et al (2012) defined risk assessment as follows: “The objective of the risk
assessment is the detailed analysis of the identified supply chain risks. The risks are prioritised based on the
dimensions ‘probability of occurrence’ and ‘business impact’ ”. Here we can say that risk assessment is the
pioneer of risk management. Different assessment approaches have been evaluated and proposed on
different scales in the literature (Tran et al., 2018).
It is important to be able to evaluate multiple criteria in Risk assessment, so AHP, ANP, and their
varieties are most applied (Vishnu et al., 2019). A study by Sachin Kumar Mangla (2015) used the fuzzy
AHP method in the green supply chain. This method has been preferred because it can consider the human
factor in risk analysis and gives successful results in uncertainty (Mangla et al., 2015). In another study,
Zubayer (2019) identified supply chain risks in the ceramics industry in Bangladesh using the Fuzzy
TOPSIS method. This method integrates imprecise variables and uncertainty into the decision model
Another study is the MCDM method used in the automotive industry supply chain made by Mzougui et al.
(2020). In this article, he studied the use of a multi-criteria decision-making method when calculating the
number of risk priorities. Besides, Tanjung et al. (2019) used two main methods in the supply chain in the
wooden toys industry: the House of risk (HOR) method in potential risk calculations, and the ANP to
determine the best strategy. In another study, Randy K. evaluated flexibility using the Bayesian Network at
gas and oil supply chain. This method evaluates risk in the event of uncertainty and shows the relationships
between variables, thus explaining the extent to which the change in a variable will affect it (Hossain et al.,
2019).
Furthermore, studies have been conducted on different software, methodologies, approaches, and
optimizations. For example, Aqlan (2016) has quickly developed software that includes a probability-effect
matrix, probability theory, and fuzzy logic for risk assessment. Aqlan and Lam (2015) also used a fuzzy
inference system approach to measure the total risk value and Bow-Tie analysis to measure the probability
of risk factors.
Risk assessment also includes risk measurement and risk selection. For this reason, it should be SCR
compatible with the Performance Measurement System at this stage. The SCPMS is created by considering
factors such as business processes, actors, connections, performance indicators, standards and procedures.
In this way, it is decided which risks will be included in the simulation. Analyzes of risk impacts usually
apply financial aggregates (eg cost, profit) to assess and select risks, but other outcomes such as lost sales,
availability level can also be applied.
Step 3-SCR Management
SCRM aims to develop strategies for the implementation of the risks identified at the previous stages
in the supply chain and the solutions/approaches that have been evaluated. It aims to strengthen the chain
through cooperation with supply chain actors, to ensure continuity with profitability, and to compete. There
are many approaches and methods in the literature to achieve these goals of SCRM (Fan & Stevenson,
2018).
The methods that are claimed to be effective in SCRM in the articles reviewed in the literature are
Transaction Cost Analysis, Channel Coordination, Constraints Theory, Total Quality Management methods.
Transaction Cost Analysis is based on financial models, the method examines prices to determine
whether goods can be sold at a low price or a high price, and usually takes place over a while. In the studies
(Um & Kim, 2017), the activities of Korean manufacturing companies were examined and evaluated. The
study (Treiblmaier, 2018) presented a framework built on the studied economic theory. A study (Schmidt
& Wagner, 2019) was investigated to evaluate how blockchain can affect supply chain relationships,
particularly in terms of transaction costs and governance decisions.
Channel Coordination shows that channels should be looked at together, not alone, to achieve the
most beneficial outcome and reduce overall costs. The study (Amit Ranjan, 2019) evaluated pricing in a
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

404
dual-channel supply chain. He studied coordination. He did this work in the field of rapidly growing internet
and smartphone users. A study (Raj & Biswas, 2018) setting up a framework for the dedicated distribution
of sustainability responsibilities in a decentralized environment was reviewed. The study (Qian & Chan,
2020) examined the coordination of a chain that included a socially responsible manufacturer and a fair-
minded retailer. This can be said as a two-tier supply chain. The problems in this chain were focused on.
In the study (Kain & Verma, 2017), Material Logistics Management provides an overview of
Logistics / Logistics management in the supply chain and their current issues related to logistics in today's
business world and provides a conceptual methodology for the relevant subject.
Constraints Theory is the method where a constraint limits the production potential or the supply
potential when it comes to supply chains. In the study (Singh & Misra, 2017), the current application model
of the Constraints Theory, especially in the downstream supply chain, is tried to be analyzed. In a different
study (Puche, 2014), the widely used Beer Game scenario was applied and the points to be considered were
highlighted. In a study conducted for the widespread use of electric vehicles (Naor & Coman, 2021) the
subject was discussed. Production and market constraints are examined in this study. The study (Costas &
Pino, 2014) where TOC is important in management strategy through the Drum - Buffer - Rope (DBR)
methodology was addressed, and improvements were made. The study (Ceniga, 2014) discussing the
possibility of applying the TOC instrument in distribution systems was examined.
Total Quality Management explains what supply chain management is by briefly analyzing the
flow characteristics of different supply chain management and the various processes implemented in the
supply chain process to improve product quality through complete quality control. Study of various
approaches that can improve the supply chain management system, Total Quality Management, and supply
chain channel (Paswan, 2020). Studies that form a theoretical basis for the integration of quality
management (QM) and supply chain management (SCM) were discussed. The results of these studies
(Sampaio & Fernandes, 2017) were taken into account and their positive effects were evaluated. To analyze
the concept of supply chain quality management (SCQM), the determined applications were compared. The
results of the study (Kaur, 2019) defining the relationships between these practices were discussed. Links
have been established between applications.
A computer model is developed to implement the SC architecture and combine simulation processes.
Different approaches can be used to create probability distributions, parameters, and risk scenarios. An
appropriate statistical evaluation should be performed to test characteristics such as input data, consistency,
accuracy, robustness, and traceability. Simulation can be used for the following purposes: developing risk
scenarios according to risk intensity, designing scenarios according to the risk exposure levels of SC actors,
defining optimal experiments for the simulation model.
Step 4-SCR Monitoring/Control
Real-time monitoring of the supply chain network helps visibility into potential unexpected macro
and micro level ECS related to products, demands, or processes that may disrupt the flow of the network
and increases the operational efficiency of processes by increasing the ability to prevent potential risks
(Tsang et al., 2018; Ho et al., 2015).
There is a limited number of studies in the literature that includes supply chain risk monitoring. An
integrated abnormality diagnosis model has been proposed by Zhang et al. (2011). The model has been
tested employing simulation, and authors indicate that abnormal data types can be identified effectively by
the proposed model. Another research is an IoT-based risk monitoring system that has been developed by
Tsang et al. (2018). The authors have stated that the developed IoT-based system provides a solid basis for
identifying and evaluating product quality and occupational safety risks in the supply chain. Kim et al.
(2016) have developed an intelligent risk management framework. In order to ensure real-time risk
monitoring, radio frequency identification (RFID) technology has been used.
The simulation model will simulate optimal scenarios to control improvements and adjust the model.
SCPMS provides relevant information to assess whether the solutions produced are satisfactory. Otherwise,
there must be a return to the development of new solutions. Here, the risk ranking can be checked again,
standards and procedures can be revised, and risk sources can be reviewed.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

405
 Table 1. OR methods with a comparison table

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

406
 Based on our research, for the Uncertainty part, “Environmental Uncertainty” and “Demand
Uncertainty” were the most used parameters in more than half of the articles, meaning the conditions that
can be changed within a day, or quickly are very crucial in the Supply Chain Risk Management. For the
Supply Chain section, “Supply Network”, “Supply Chain Collaboration”, “Supplier Integration” parameters
have been used each more than one-third of all articles. This can be interpreted in a way that the relationships
and the order in the Supply Chain Risk Management procedure are vital.
For the last and additional part, Others, many single independent parameters could have made it to the
table, but the most significant ones can be said as “Disruption” in the first place with almost half of the
articles, followed by “Economical Fluctuations” and “Law Effect” with almost one fourth each of all the
articles read. This can be evaluated, such that the routine flow of the Supply Chain and its management to
be dealt with risks can be interrupted with such elements, that are not directly from the supply or demand
side, but overall, from the daily life itself. There were many different methods used in the articles such as
Systematic Literature Review (SLR), structural equation modeling (SEM), AHP and ANP methods
combined in a single article, survey-based qualitative case studies, etc.
Supply Chain Risk Management Methodology Framework
Based on the data obtained from Table 1 and the literature reviewed within the scope of this
research, the methodology framework shown in Figure 3 was prepared.

Figure 3. Supply Chain Risk Management Methodology Framework

Effect of Some of Optimization Methods on SCRM
The five most commonly used optimization methods in the literature and the benefits that can be obtained
in the SCRM management system are summarized below;
o Analytical Hierarchy Process: AHP is a structured method for organizing and analyzing complex
decisions. (Mathivathanan et al., 2017)
o Fuzzy TOPSIS: Fuzzy TOPSIS is a method in which the ratings of various alternatives against
various subjective criteria and the weights of all criteria are evaluated in linguistic terms represented
by fuzzy numbers. (Chu & Lin,2003)
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

407
o Bayesian Network: Bayesian model is a probabilistic graph model and expresses a set of random
variables that have conditional dependencies with each other as a directional non-return graph. (Ojha
et al., 2018)
o Multi-Criteria Decision Making: MCDM methods can deconstruct the problem of decision making
into discrete steps, compare the relative importance of criteria and select the optimal alternative using
rigorous mathematical models. These methods can clarify the interrelations between criteria and
minimise the subjectivity ofthe selection (Linkov & Moberg, 2012).
o HOR Method: The HOR method is a type of method used to calculate potential risks. (Tanjung et al.,
2019)
Effect of Simulation Methodology on SCRM
The benefits that can be obtained in the SCRM management system through the four simulation
methodologies most commonly used in the literature are summarized below;
o Discrete Event Simulation (DES): DES (Discrete Event Simulation) has been a foundation for
manufacturing and supply chain process simulations for many years. DES is useful for events
consisting of complex queuing problems. Queues are important in DES because flow is defined
through a system. Models are made and based on objective data collected/measured. These features
are useful in manufacturing and supply chain processes. This approach may not be useful for
problems that are not related to the flow of resources outside of manufacturing and supply chain.
Another approach may be required.(Pawlewski et al,2012)
o Agent-based Simulation (ABS): ABS is useful for a better understanding of decentralized systems
where the representation or modeling of many individuals is important. ABS is new to manufacturing
and the supply chain and offers something potentially highly viable. There is no concept of queues
and flows. These features complement and are used in many aspects of supply chain
management.(Siebers et al,2010)
o Monte Carlo Simulation (MCS): The popularity of Monte Carlo Simulation is due to its versatility
and ease of use. The MCS method is a quantitative method that involves assigning individual risk
types to the probability of occurrence. The weak point of this method is that it does not include the
probability of randomness when determining the probabilities of various project risks. MCS is used in
SCR Managemet together with artificial intelligence applications.(Rymarczyk and Kłosowski,2017)
o System Dynamic Simulation (SDS): SDS is based on feedback control theory. The Model treats the
working object as a dynamic system and is characterized accordingly. The basic principle is: use
system modeling to provide a basis for Strategy and decision-making, send the model to the computer,
and verify its validity. It is used in SCR Assessment and management. (Feng,2012)

FUTURE DIRECTION AND CHALLENGES

To test the reliability of the supply chain, the quantity, quality, and time components are calculated
with the weighted average. The multi-objective optimization model is used for order determination and
supplier selection (Vishnu et al., 2019). The Universal Generating Function (UGF) should be used to model
different suppliers in supply chain risk management throughout the entire supply chain. With this approach,
supplier selection, preventing risks that may arise in deliveries and making savings easier to manage (Mohib
& Deif, 2020). Requests from customers, models created using Poisson distribution, and supplier selection
are integrated with Simulation Optimization. Thus, the stock status of the supplier can be followed in direct
proportion to the order (Göçken et al., 2018). When any outage scenario is considered as a problem, supply
planning should be done in the worst-case scenario with two time-indexed stochastic MIP models DFR and
OFR models (Sawik, 2016).
There may be disruptions in different parts of the food supply chain due to uncertain events. As a
result, serious economic losses may occur. In supply chains from grower to consumer, these disruptions can
be optimized with the Agent-based model. In the future, this algorithm can be adapted considering different
supply chain risks (Bottani et al., 2019). For grains, losses occur during operations such as transportation
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

408
and storage in the supply chain process after harvest until they reach the consumer. It proposes models to
bring these losses to the optimum level by considering Brazil, and in the future, research can be made on
this subject by taking into account different transport vehicles (An & Ouyang, 2016). A similar loss is
experienced in the fresh grape production supply chain in China. Unlike grain, the product is perishable
here, making the supply chain fragile. It is emphasized that the supply chain should be paid attention not
only to its intactness but also to its sustainability. Risks are said to be optimized with data using a neural
network created through machine learning (Jianying et al., 2021). It is very important to control the quality
of wheat, which is another type of grain, in terms of food safety. In Canada, these controls are made using
the Kernel Visual Distinction (KVD) system, but this system is very costly. It can be modified for different
mediators as well as control risk optimization using the proposed methodology (Ge et al., 2015). As a result
of the awareness of people, the interest in environmentally friendly products is increasing. Therefore, not
only the companies themselves but also the suppliers they work with must be environmentally friendly. It
explains how to optimize the Partner for Sustainability approach to minimize this risk (Miranda-Ackerman
et al., 2019). The deterioration of food products and the changing prices and demands require inventory
management under these uncertainties. The proposed approach has been shown to be effective in balancing
risk and cost. In the future, the use of the same methodology for different products' problems may be
investigated (Violi et al., 2020). Failure to meet product standards in the food industry poses a great risk.
Companies have developed policies to reduce these risks. However, these policies increase the processing
process. The proposed multipurpose approach offers a highly sustainable perspective as well as providing
safe food supply (Voldrich et al., 2020). One of the risks in food safety of food supply chains is monitoring
chemical hazards. An optimization method has been developed to determine sampling patterns to monitor
chemicals considered harmful in dairy products in the Netherlands (Wang et al., 2020).
A common problem in supply chain risk management, the newsvendor problem is approached with
the mean-risk optimization model. While solving the problem, parameters such as order costs, capacities,
costs of unsold products of the suppliers were modeled using linear programming. It is aimed to maximize
the gain by modeling the NP-MUS-E (Newsvendor Problem with Multiple Unreliable Suppliers-
Expectation)(Merzifonluoglu, 2015) .
Looking at purchasing part of the supply chain, in the study conducted to reduce the risk of purchase
cost, a multiobjective optimization model was created and Pareto-optimal solutions were found with the
NSGA-II algorithm (Li et al., 2019).
In another article, the actual application of supply chain risk management is associated with examples.
Examples for problem-solving with postponement strategy in Make To Order (MTO) and Make To Stock
(MTS) systems are shown (Sodhi et al., 2016).
The study on risk management in the bulk coal export logistic chain in South Africa was divided into
five different sections (Infrastructure, macro-economic, operational risks, natural disasters, people) and the
approaches to be applied in these themes were specified (Botha & Badenhorst-Weiss, 2019).
One of the main items of supply chain risk management is risk visibility. It is seen that the visibility
of risk and the interactions between risks can be better understood using a multi-purpose NP-hard model
(Nooraie & Mellat Parast, 2015). Today, biofuel production draws attention as it is sustainable and
renewable. Production of biofuels from the biomass supply chain can see both natural and human-induced
damage easily. A mathematical model has been developed to design a supply chain network considering
these risks. In the solution of this model, algorithms such as beam optimization are used (Khezerlou et al.,
2021). It is important to be environmentally friendly in biofuel supply chains. Therefore, a mathematical
model has been developed that both maximizes net benefit and reduces harmful gas emissions in supply
chains (Delkhosh & Sadjadi, 2020). Uncertainty in the biomass supply chain can lead to significant risks.
The Monte Carlo Simulation optimization model can be used to understand the effects of uncertainty and
variability on the supply chain (Shabani & Sowlati, 2016). Biomass raw material goes through different
processes in biofuel production. This is called the biomass supply chain. It can be interrupted by risks such
as natural disasters, financial crises, as well as uncertainties such as biomass supply and demand for biofuels.
Uncertainties such as these can be minimized using hybrid robust stochastic probabilistic programming
(RSPP) (Mousavi Ahranjani et al., 2020). It is seen that the risk created by the uncertainty in the biomass
price and demand in the biodiesel supply chain can be optimized by adding to the stochastic mixed-integer
programming model (Azadeh & Vafa Arani, 2016). The optimization problem of the supply chain in
biomass-based bioethanol and biopower production can be solved with a MILP modeling perspective
(d'Amore & Bezzo, 2017).
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

409
 The health sector, which has a direct effect on human life, is one of the important points that
distinguish it from other sectors, with its structure that does not make any mistakes, and the urgent service.
Hospitals, which are the main unit of health services, have a supply chain that includes more components
(such as food, cleaning supplies, bed sheets, surgical gowns) compared to other sectors (Acar & Bozaykut,
2019). Inventory accumulation and obsolescence in the hospital and health sector are several times higher
than in the industry / retail sectors. To provide an efficient health service, many components of the supply
chain must be provided correctly (Zepeta et al., 2016). The methodologies used in the procurement
processes in the hospital supply chain were examined. Cronbach's Alpha values were calculated to weigh
the reliability of the questionnaire. One-way variance (Oneway Anova) and T-test were applied to find out
whether the data differed statistically significantly (Karataş, 2018). When we examine the risks, the blood
supply chain in the health sector, this problem, addressed in a study conducted to provide the locations of
central and temporary blood centers and blood supply in case of a disaster, has been made in the Robust
approach optimization model. A self-adaptive imperialist competitive algorithm (SAICA) was used to solve
the proposed problem and invasive weed optimization (IWO) algorithm was used to find the efficiency of
this algorithm (Haghjoo et al., 2020). In another study on the blood supply chain, considering the pandemic
period, a study was carried out using the Multi-objective mixed-integer mathematical model for more
efficient planning of hospitals, fixed blood centers, and mobile blood centers (Karadağ et al., 2021).
While managing the Software Supply Chain, many risks such as software errors, production delays,
increased production costs may occur. Bayesian Network method can be used to predict these risks (Librantz
et al., 2020). Four different models are used in a study conducted to identify and minimize risks in IT supply
chains. (Maximized coverage, maximized worst-case, minimized regret, maximized conditionally expected
coverage) These models are given examples on companies and it is stated which model will be used in
which scenario and what kind of problems will be solved (Zheng & Albert, 2019).

CONCLUSION

A supply chain is a set of relationships with many components, such as manufacturers, suppliers,
distributors, customers. When different sectors were considered in Supply Chain Risk Management, it was
observed that in addition to having common risks that may occur in the sectors, each sector has its risks and
sensitive points. Ensuring sustainability at sensitive points in the supply chain is crucial. Measures should
be taken to address possible risks to this sustainability. In this context, under the heading of supply chain
risk management, possible risks should be determined by different methods and avoided before they occur.
As it can result from the table (Table 1), there have been many parameters depending on what the
articles have focused on. These parameters have a reflection and relation in the everyday life as well,
meaning Operations Research processes are a part of daily life. As for the methods, some many different
methods and models shaped our table.
To sum up should be avoided by foreseeing possible risks in the supply chain before they arise. The
first step in preventing these risks is to anticipate the risks and calculate the probabilities of these risks.
Appropriate operations research methods should be selected for this forecasting process. The second step is
to identify and implement the most efficient ways to prevent the emergence of predicted risks with the help
of the mentioned methods. In this way, possible losses can be minimized.
The COVID-19 pandemic has put entire economies in a difficult position and has had a devastating
impact on nearly every industry, demonstrating the value of risk management. In this context, based on the
findings of this research, the need for a robust, flexible, and IoT-based management model in supply chain
risk management that covers the whole process has been observed in addition to the IoT-based monitoring
model. Therefore, it is aimed to develop the research in this direction.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

410
 REFERENCES
Abdel-Basset, M., & Mohamed, R. (2020). A novel plithogenic TOPSIS-CRITIC model for sustainable supply
chain risk management. Journal of Cleaner Production, 247, 119586.
Acar, Z., & Bozaykut, T. (2019). A General Review of Turkish Healthcare Supply Chain Management. International
Journal of Research in Business and Social Science.
Ali, I., & Shukran, K. (2016). MANAGING SUPPLY CHAIN RISKS AND VULNERABILITIES THROUGH
COLLABORATION: PRESENT AND FUTURE SCOPE. The Journal of Developing Areas, 50(5), 335-342.
An, K., & Ouyang, Y. (2016). Robust grain supply chain design considering post-harvest loss and harvest timing
equilibrium. Transportation Research Part E: Logistics and Transportation Review.
Aqlan, F. (2016). A software application for rapid risk assessment in integrated supply chains. Expert Systems with
Applications.
Aqlan, F., & Lam, S. (2015). A fuzzy-based integrated framework for supply chain risk assessment. International Journal
of Production Economics.
Avci, M. G., & Selim, H. (2018). A multi-objective simulation-based optimization approach for inventory replenishment
problem with premium freights in convergent supply chains. Omega, 80, 153-165.
Azadeh, A., & Vafa Arani, H. (2016). Biodiesel supply chain optimization via a hybrid system dynamics-mathematical
programming approach. Renewable Energy.
B., F., C., T., H., D., & J., S. (n.d.). Quantitative models for managing supply chain risks: A review. European Journal of
Operational Research, 247(1), 1-15. https://www.sciencedirect.com/science/article/abs/pii/S0377221715003276
Baryannis, G., Dani, S., Validi, S., & Antoniou, G. (2018). Decision Support Systems and Artificial Intelligence in Supply
Chain Risk Management. In Revisiting Supply Chain Risk (p. 19). Springer.
https://link.springer.com/chapter/10.1007/978-3-030-03813-7_4
Baryannis, G., Validi, S., Dani, S., & Antoniou, G. (n.d.). Supply chain risk management and artificial intelligence: state
of the art and future research directions. International Journal of Production Research, 57(7), 25. 2018
Beltrão, L. M., & Carvalho, M. T. (2019). Prioritizing construction risks using fuzzy AHP in Brazilian public
enterprises. Journal of Construction Engineering and Management, 145(2), 05018018.
Biswas, I., Raj, A., & Srivastava, S. K. (2018). Supply chain channel coordination with triple bottom line approach.
Transportation Research Part E: Logistics and Transportation Review, 115, 213-226.
Botha, A., & Badenhorst-Weiss, J.A. (2019). Risk management in a bulk coal export logistic chain: A stakeholder
perspective. Journal of Transport and Supply Chain Management.
Bottani, E., Murino, T., Schiavo, M., & Akkerman, R. (2019). Resilient food supply chain design: Modelling framework
and metaheuristic solution approach. Computers and Industrial Engineering.
Brusset, Xavier & Teller, Christoph. (2016). Supply Chain Capabilities, Risks, and Resilience. International Journal of
Production Economics. 184. 10.1016/j.ijpe.2016.09.008.
BURAN, A. Ç., & AĞCA, A. (2019). THE EFFECT OF COLLABORATION ON MITIGATING SUPPLY CHAIN
RISKS: A RESEARCH ON MANUFACTURING COMPANIES. DUMLUPINAR UNIVERSITY JOURNAL OF
SOCIAL SCIENCES, 19.
Calis, B. (2016). Agent-Based Simulation Model for Profit Maximization. Journal of Management and Information
Science, 4(1), 26-33.
Chaudhuri, A., Boer, H., & Taran, Y. (2018). Supply chain integration, risk management and manufacturing flexibility.
International Journal of Operations & Production Management.
Choirun, A & Santoso, Imam & Astuti, Retno. (2020). Sustainability risk management in the agri-food supply chain:
literature review. IOP Conference Series: Earth and Environmental Science. 475. 012050. 10.1088/1755-
1315/475/1/012050.
Chu, C.-Y., Park, K., & Kremer, G. (2020). A global supply chain risk management framework: An application of text-
mining to identify region-specific supply chain risks. Elsevier, 17.
https://www.sciencedirect.com/science/article/abs/pii/S1474034620300227
Ciccullo, F., Pero, M., & Caridi, M. (2017). Exploring the hidden potential of product design to mitigate supply chain risk.
Claypool, Erin. (2015). Design for Supply Chain: An Analysis of Key Risk Factors. Industrial Engineering and
Management. 04. 10.4172/2169-0316.1000156.
Costas, J., Ponte, B., de la Fuente, D., Pino, R., & Puche, J. (2015). Applying Goldratt’s Theory of Constraints to reduce
the Bullwhip Effect through agent-based modeling. Expert systems with Applications, 42(4), 2049-2060.
d'Amore, F., & Bezzo, F. (2017). Managing technology performance risk in the strategic design of biomass-based supply
chains for energy in the transport sector. Energy.
DANACI, B., HELVACIOĞLU, Ş., & KORKUT, E. (2019). TERSANELERDE TEDARİK ZİNCİRİ YÖNETİMİ. GİDB
Dergi, 1(14), 29-40. http://gidbdergi.itu.edu.tr/tr/pub/issue/53664/717424
Daud A R, Putro U S and Basri M H 2015. Risks in milk supply chain; a preliminary analysis on smallholder dairy
production. Livestock Research for Rural Development. Volume 27, Article #137. Retrieved April 22, 2021, from
http://www.lrrd.org/lrrd27/7/daud27137.htm
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

411
Delkhosh, F., & Sadjadi, S.J. (2020). A robust optimization model for a biofuel supply chain under demand uncertainty.
International Journal of Energy and Environmental Engineering.
Demir, P., & Kayali, N. (2020). PROACTIVE RISK MANAGEMENT APPROACH IN REDUCING THE
AGRICULTURAL SUPPLY CHAIN RISKS. PressAcademia, 3(7), 13. 10.17261/Pressacademia.2020.1294
DEMİRKOL, İ., ÜNĞAN, M. C., & AYANOĞLU, M. (2015). THE IMPACT OF SUPPLY CHAIN RISKS ON FIRM
PERFORMANCE: A SURVEY RESEARCH IN AUTOMATIVE INDUSTRY. İşletme Bilimi Dergisi, 3(1), 18.
Diabat, Ali & Govindan, Kannan & Panicker, Vinay. (2012). Supply chain risk management and its mitigation in a food
industry. International Journal of Production Research - INT J PROD RES. 50. 1-12.
10.1080/00207543.2011.588619.
Dias, Gabriela Costa, Leal Junior, Ilton Curty, & Oliveira, Ualison Rébula de. (2019). Supply chain risk management at
seaport container terminals. Gestão & Produção, 26(3), e4900. Epub August 12,
2019.https://doi.org/10.1590/0104-530x4900-19
Dubey, R., Gunasekaran, A., Childe, S. J., Papadopoulos, T., & Wamba, S. F. (2017). World class sustainable supply chain
management: critical review and further research directions. The International Journal of Logistics Management.
Ekwall, D., & Lantz, B. (2016). Supply Chain Risk Analysis and Assessment: Cargo Theft. Transportation Journal, 55(4),
400-419. doi:10.5325/transportationj.55.4.0400
Fan, Y., & Stevenson, M. (2018). A review of supply chain risk management: definition, theory, and research
agenda. International Journal of Physical Distribution & Logistics Management.
Feng, Y. (2012). System dynamics modeling for supply chain information sharing. Physics Procedia, 25, 1463-1469.
Fernandes, A. C., Sampaio, P., Sameiro, M., & Truong, H. Q. (2017). Supply chain management and quality management
integration: A conceptual model proposal. International Journal of quality & reliability management.
Gautam, Aditya & Prakash, Surya & Soni, Umang. (2018). Supply chain risk management and quality: a case study and
analysis of Indian automotive industry. International Journal of Intelligent Enterprise. 5. 194-212.
10.1504/IJIE.2018.091189.
Ge, H., Gray, R., & Nolan, J. (2015). Agricultural supply chain optimization and complexity: A comparison of analytic vs
simulated solutions and policies. International Journal of Production Economics.
Giannakis, M., & Papadopoulos, T. (2016). Supply chain sustainability: A risk management approach. International Journal
of Production Economics, 171, 455-470.
Gottlieb, Simone & Ivanov, Dmitry & Das, Ajay. (2019). Case studies of the digital technology impacts on supply chain
disruption risk management. 10.1007/978-3-658-25412-4_2.
Göçken, M., Boru, A., & Dosdoğru, A. (2018). İki Aşamalı Tedarik Zincirinde Eş Zamanlı Stok Kontrolünün ve Tedarikçi
Seçiminin Simülasyon Optimizasyonu Yaklaşımı ile Analizi. Karaelmas Fen ve Mühendislik Dergisi.
Haghjoo, N., Tavakkoli-Moghaddam, R., Shahmoradi-Moghadam, H., & Rahimi, Y. (2020). Reliable blood supply chain
network design with facility disruption: A real-world application. Engineering Applications of Artificial
Intelligence.
Ho, W., Zheng, T., Yildiz, H., & Talluri, S. (2015). Supply chain risk management: a literature review. International Journal
of Production Research, 53(16), 5031-5069.
Hossain, N. U. I., Jaradat, R., Marufuzzaman,, M., Buchanan, R. K., & Rinaudo, C. (2019). Assessing and Enhancing Oil
and Gas Supply Chain Resilience: A Bayesian Network Based Approach. IISE Annual Conference.
Ivanov, D. (2017). Simulation Applications to Structural Dynamics in Service and Manufacturing Supply Chain Risk
Management. In Simulation Applications to Structural Dynamics in Service and Manufacturing Supply Chain Risk
Management (p. 32). Springer. https://link.springer.com/chapter/10.1007/978-3-319-69305-7_8
Jiang, Min & Yang, Liu & Zhao, Xiye. (2019). Research on Risk Management of Fresh Agricultural Products Supply
Chain Based on Fuzzy Analytic Hierarchy Process. International Journal of Asian Social Science. 9. 516-524.
10.18488/journal.1.2019.910.516.524.
Jianying, F., Bianyu, Y., Xin, L., Dong, T., & Weisong, M. (2021). Evaluation on risks of sustainable supply chain based
on optimized BP neural networks in fresh grape industry. Computers and Electronics in Agriculture.
Kain, R., & Verma, A. (2018). Logistics management in supply chain–an overview. Materials today: proceedings, 5(2),
3811-3816.
Karadağ, İ., Keskin, M.E., & Yiğit, V. (2021). Re-design of a blood supply chain organization with mobile units. Soft
Computing.
Karataş, G. (2018). Hastanelerde Tedarik Zinciri Yönetiminde Karşılaşılan Sorunlar Üzerine Bir Alan Araştırması. İşletme
Anabilim Dalı Yüksek Lisans Tezi.
Kaur, M., Singh, K., & Singh, D. (2019). Synergetic success factors of total quality management (TQM) and supply chain
management (SCM). International Journal of Quality & Reliability Management.
Kern, D., Moser, R., Hartmann, E. and Moder, M. (2012) ‘Supply risk management: model development and empirical
analysis’, International Journal of Physical Distribution & Logistics Management, Vol. 42, No. 1, pp.60–82.
Khezerlou, H.S., Vahdani, B., & Yazdani, M. (2021). Designing a resilient and reliable biomass-to-biofuel supply chain
under risk pooling and congestion effects and fleet management. Journal of Cleaner Production.
Kırılmaz, O., & Erol, S. (n.d.). A proactive approach to supply chain risk management: Shifting orders among suppliers to
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

412
 mitigate the supply side risks. Journal of Purchasing and Supply Management, 23(1), 54-65.
https://www.sciencedirect.com/science/article/abs/pii/S1478409216300061
Kilubi, I. (2016). The strategies of supply chain risk management–a synthesis and classification. International Journal of
Logistics Research and Applications, 19(6), 604-629.
Kim, K., Kim, H., Kim, S. K., & Jung, J. Y. (2016). i-RM: An intelligent risk management framework for context-aware
ubiquitous cold chain logistics. Expert Systems with Applications, 46, 463-473.
Li, G., Zheng, J., & Zhang, M. (2019). Parallel optimization method for product configuration and supplier selection based
on interval. IOP Conference Series: Earth and Environmental Science.
Librantz, A.F.H., Costa, I., Spinola, M.D.M., de Oliveira Neto, G.C., & Zerbinatti, L. (2020). Risk assessment in software
supply chains using the bayesian method. International Journal of Production Research.
Mangla, S., Kumar, P., & Barua, M. (2015). Risk analysis in green supply chain using fuzzy AHP approach: A case study.
Resources, Conservation and Recycling.
MANHART, P., SUMMERS, J., & BLACKHURST, J. (2018). A META-ANALYTIC REVIEW OF SUPPLY CHAIN
RISK MANAGEMENT: ASSESSING BUFFERING AND BRIDGING STRATEGIES AND FIRM
PERFORMANCE. Journal of Supply Chain Management, 56(3), 34. 10.14514/BYK.m.21478082.2018.6/1.84-
117
Merzifonluoglu, Y. (2015). Impact of risk aversion and backup supplier on sourcing decisions of a firm. International
Journal of Production Research.
Miranda-Ackerman, M.A., Azzaro-Pantel, C., Aguilar-Lasserre, A.A., Bueno-Solano, A., & Arredondo-Soto, K.C. (2019).
Green supplier selection in the agro-food industry with contract farming: A multi-objective optimization approach.
Sustainability (Switzerland).
Mohib, A.M.N., & Deif, A.M. (2020). Supply chain multi-state risk assessment using universal generating function.
Production Planning and Control.
Moktadir, M. A., Dwivedi, A., Khan, N. S., Paul, S. K., Khan, S. A., Ahmed, S., & Sultana, R. (2021). Analysis of risk
factors in sustainable supply chain management in an emerging economy of leather industry. Journal of Cleaner
Production, 283, 124641.
Mousavi Ahranjani, P., Ghaderi, S.F., Azadeh, A., & Babazadeh, R. (2020). Robust design of a sustainable and resilient
bioethanol supply chain under operational and disruption risks. Clean Technologies and Environmental Policy.
Munir, M., Jajja, M. S. S., Chatha, K. A., & Farooq, S. (2020). Supply chain risk management and operational performance:
The enabling role of supply chain integration. International Journal of Production Economics, 227, 107667.
Mzougui, I., Carpitella, S., Certa, A., El Felsoufi, Z., & Izquierdo, J. (2020). Assessing Supply Chain Risks in the
Automotive Industry through a Modified MCDM-Based FMECA.
Naor, M., Coman, A., & Wiznizer, A. (2021). Vertically Integrated Supply Chain of Batteries, Electric Vehicles, and
Charging Infrastructure: A Review of Three Milestone Projects from Theory of Constraints Perspective.
Sustainability, 13(7), 3632.
Natarajarathinam, M., Capar, I., & Narayanan, A. (2009). Managing supply chains in times of crisis: a review of literature
and insights. International Journal of Physical Distribution & Logistics Management.
Nieuwenhuyzen, M.V., Niemann, W., & Kotze, T.G. (2018). Supply chain risk management strategies : a case study in the
South African grocery retail industry
Nooraie, S.V., & Mellat Parast, M. (2015). . A multi-objective approach to supply chain risk management: Integrating
visibility with supply and demand risk. International Journal of Production Economics.
Oliveira, J.B., JİN, M., Lima, R.S., Kobza, J.E., & Montevechi, J.A.B. (2019). The role of simulation and optimization
methods in supply chain risk management: Performance and review standpoints. Elsevier, 92, 28.
https://www.sciencedirect.com/science/article/abs/pii/S1569190X18301734
Oliveira, U. R. d., Marins, F. A. S., Rocha, H. M., & Salomon, V. A. (2017). THE ISO 31000 STANDARD IN SUPPLY
CHAIN RISK MANAGEMENT. Journal of Cleaner Production, 151, 35.
http://dx.doi.org/10.1016/j.jclepro.2017.03.054
Panova, Y., & Hilletofth, P. (2018). Managing supply chain risks and delays in construction project. Industrial Management
& Data Systems.
Omnia Partners. (n.d.). A Beginner’s Guide to Supply Chain Risk Management IDENTIFY, ASSESS, MITIGATE.
Https://Www.Riskmethods.Net/.
https://info.omniapartners.com/hubfs/PRIVATE_SECTOR/Supplier%20Pages/riskmethods/beginners-guide-to-
scrm%20w_%20logo.pdf
Özfirat, M. K., Özkan, E., Kahraman, B., Şengün, B., & Yetkin, M. E. (2017). Integration of risk matrix and event tree
analysis: a natural stone plant case. Sādhanā, 42(10), 1741-1749.
Paswan, V. K., Mukherjee, A., & Krishna, B. V. (2020). Productive Implementation of Supply Chain Management with
Total Quality Management. International Journal of Innovative Research in Science, Engineering and Technology,
4684-4689.
Pawlewski, P., Golinska, P., & Dossou, P. E. (2012). Application potential of Agent Based Simulation and Discrete Event

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

413
 Simulation in enterprise integration modelling concepts. ADCAIJ: Advances in Distributed Computing and
Artificial Intelligence Journal, 1(1), 33-42.
Puche, J., Ponte, B., Costas, J., Pino, R., & De la Fuente, D. (2016). Systemic approach to supply chain management
through the viable system model and the theory of constraints. Production planning & control, 27(5), 421-430.
Qazi, A., Akhtar, P., & Wieland, A. (2018). Risk Matrix Driven Supply Chain Risk Management: Adapting Risk Matrix
Based Tools To Modelling Interdependent Risks and Risk Appetite. Computers & Industrial Engineering, 38.
https://doi.org/10.1016/j.cie.2018.08.002
Qian, X., Chan, F. T., Zhang, J., Yin, M., & Zhang, Q. (2020). Channel coordination of a two-echelon sustainable supply
chain with a fair-minded retailer under cap-and-trade regulation. Journal of Cleaner Production, 244, 118715.
Rajagopal, V., Venkatesan, S. P., & Goh, M. (2017). Decision-making models for supply chain risk mitigation: A
review. Computers & Industrial Engineering, 113, 646-682.
Ramesh, K. T., & Sarmah, S. P. (2020). Impact of supply risk management on firm performance: a case of the Indian
electronics industry. International Journal of Productivity and Performance Management.
Ranjan, A., & Jha, J. K. (2019). Pricing and coordination strategies of a dual-channel supply chain considering green quality
and sales effort. Journal of Cleaner Production, 218, 409-424.
Raspotnig, C., & Opdahl, A. (2013). Comparing risk identification techniques for safety and security requirements. Journal
of Systems and Software, 86(4), 1124-1151.
Rostamzadeh, R., Ghorabaee, M. K., Govindan, K., Esmaeili, A., & Nobar, H. B. K. (2018). Evaluation of sustainable
supply chain risk management using an integrated fuzzy TOPSIS-CRITIC approach. Journal of Cleaner
Production, 175, 651-669.
Rymarczyk, T., & Kłosowski, G. (2017). Supply chain risk management by Monte Carlo method. Informatyka,
Automatyka, Pomiary w Gospodarce i Ochronie Środowiska, 7.
Saglam, Y. C., Çankaya, S. Y., & Sezen, B. (2020). Proactive risk mitigation strategies and supply chain risk management
performance: an empirical analysis for manufacturing firms in Turkey. Journal of Manufacturing Technology
Management.
Sato, Y., Tse, Y. K., & Tan, K. H. (2020). Managers' risk perception of supply chain uncertainties. Industrial Management
& Data Systems.
Sawik, T. (2016). On the risk-averse optimization of service level in a supply chain under disruption risks. International
Journal of Production Research.
Schmidt, C. G., & Wagner, S. M. (2019). Blockchain and supply chain relations: A transaction cost theory perspective.
Journal of Purchasing and Supply Management, 25(4), 100552.
Schulte, J., Villamil, C., & Hallstedt, S. I. (2020). Strategic sustainability risk management in product development
companies: Key aspects and conceptual approach. Sustainability, 12(24), 10531.
Shabani, N., & Sowlati, T. (2016). Evaluating the impact of uncertainty and variability on the value chain optimization of
a forest biomass power plant using monte carlo simulation. International Journal of Green Energy.
Shahbaz, Muhammad & RM Rasi, Raja Zuraidah & Zulfakar, Mohd & Ahmad, Md & Abbas, Zaheer & Mubarak,
Muhammad. (2018). A NOVEL METRIC OF MEASURING PERFORMANCE FOR SUPPLY CHAIN RISK
MANAGEMENT: DRAWBACKS AND QUALITIES OF GOOD PERFORMANCE. Journal of Applied and
Fundamental Sciences. 10. 967-988. 10.4314/jfas.v10i3s.83.
Shojaei1, P., & Haeri, ,. S. A. S. (2018). Development of supply chain risk management approaches for construction
projects: A grounded theory approach. Shiraz University, 37.
https://www.sciencedirect.com/science/article/abs/pii/S0360835218305850
Shou, Y., Hu, W., Kang, M., Li, Y., & Park, Y. W. (2018). Risk management and firm performance: the moderating role
of supplier integration.
Siebers, P. O., Macal, C. M., Garnett, J., Buxton, D., & Pidd, M. (2010). Discrete-event simulation is dead, long live agent-
based simulation!. Journal of Simulation, 4(3), 204-210.
Singh, K., & Misra, D. (2017). Theory of constraints for sustainable inventory management in downstream supply chain.
Journal of Management (JOM) Volume, 4, 1-10.
Sodhi, M.S., Son, B., & Tang, C.S. (2016). Researchers' perspectives on supply chain risk management. Production and
Operations Management.
Šoško, G. B., Grgurević, D., & Buntak, K. (2019). RISK MANAGEMENT AS A FACTOR OF INCREASING OF
COMPETITIVENESS AND MORE EFFICIENT SUPPLY CHAIN MANAGEMENT. International Journal for
Quality Research, 13(2).
Šukalová, V., & Ceniga, P. (2015). Application of the theory of constraints instrument in the enterprise distribution system.
Procedia economics and finance, 23, 134-139.
Sutduean, J., Joemsittiprasert, W., & Jermsittiparsert, K. (2019). Uncertainty as an antecedent of the supply chain risk:
Does supply chain flexibility matter in risk mitigation?. Humanities & Social Sciences Reviews, 7(2), 503-509.
Şen, E., Körük, E., Serper, N., & Çalış Uslu, B. (2019). Big Data Analytics and Simulation for Better Strategic
Management. Journal of Current Research on Engineering, Science and Technology, 5(2), 1-12.
Tanjung, W. N., Atikah, S. A., Hidayat, S., Ripmiatin, E., Asti, S. S., & Khodijah, R. S. (2019, May). Risk Management
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

414
 Analysis Using FMECA and ANP Methods in the Supply Chain of Wooden Toy Industry. In IOP Conference
Series: Materials Science and Engineering (Vol. 528, No. 1, p. 012007). IOP Publishing.
TEPE KÜÇÜKOĞLU, M. (2020). Supply Chain Risk Evaluation: A Study on Identification, Grouping and Prioritization
of Risks. JOURNAL OF BUSINESS RESEARCH-TURK, 17.
Tran, T. H., Dobrovnik, M., & Kummer, S. (2018). Supply chain risk assessment: a content analysis-based literature review.
International Journal of Logistics Systems and Management, 31(4), 562-591.
Treiblmaier, H. (2018). The impact of the blockchain on the supply chain: a theory-based research framework and a call
for action. Supply Chain Management: An International Journal.
Trkman, P., Oliveira, M. P. V. d., & McCormack, K. (2016). Value-oriented supply chain risk management: you get what
you expect. Emerald Insight, 116(5), 25. https://www.emerald.com/insight/content/doi/10.1108/IMDS-09-2015-
0368/full/html
Tsang, Y. P., Choy, K. L., Wu, C. H., Ho, G. T., Lam, C. H., & Koo, P. S. (2018). An Internet of Things (IoT)-based risk
monitoring system for managing cold supply chain risks. Industrial Management & Data Systems.
Um, K. H., & Kim, S. M. (2019). The effects of supply chain collaboration on performance and transaction cost advantage:
The moderation and nonlinear effects of governance mechanisms. International Journal of Production Economics,
217, 97-111.
Ummi, Nurul & Ferdinant, Putro & Ade, Irman & Gunawan, Akbar. (2018). Integration house of risk and analytical
network process for supply chain risk mitigation of cassava opak chips industry. MATEC Web of Conferences.
218. 04022. 10.1051/matecconf/201821804022.
Violi, A., Laganá, D., & Paradiso, R. (2020). The inventory routing problem under uncertainty with perishable products:
An application in the agri-food supply chain. Soft Computing.
Vishnu, C.R., Das, S.P., & Sridharan, R. (2019). Reliable supplier selection and order allocation model for managing risks:
A simulation-based optimization approach. IEEE International Conference on System, Computation, Automation
and Networking.
Vishnu, C.R., Sridharan, R., & Kumar, P. N. R. (2019). Supply chain risk management: models and methods. International
Journal of Management and L methd Making.
Voldrich, S., Wieser, P., & Zufferey, N. (2020). ptimizing the trade-off between performance measures and operational
risk in a food supply chain environment. Soft Computing.
Wang, Z., van der Fels-Klerx, H.J., & Lansink, A. G. J. M. O. (2020). Optimization of sampling for monitoring chemicals
in the food supply chain using a risk-based approach: The case of aflatoxins and dioxins in the dutch dairy chain.
Risk Analysis.
Wiengarten, F., Humphreys, P., Gimenez, C., & McIvor, R. (2016). Risk, risk management practices, and the success of
supply chain integration. International Journal of Production Economics, 171, 361-370.
Wijaya, O. Y. A. (2021). Risk Management Mitigation in the New Normal Era. Budapest International Research and Critics
Institute-Journal (BIRCI-Journal) Vol, 4(1), 1088-1097.
Yang, Bo & Xie, Le. (2019). Bayesian Network Modelling for “Direct Farm” Mode based Agricultural Supply Chain Risk.
Journal of Physics: Conference Series. 1288. 012056. 10.1088/1742-6596/1288/1/012056.
Yang, J., & Xie, H. (2020). Antecedents and consequences of supply chain risk management capabilities: an investigation
in the post-coronavirus crisis. International Journal of Production Research, 59(5), 1573-1585.
https://doi.org/10.1080/00207543.2020.1856958
Yin, C., & McKay, A. (2018). Introduction to Modeling and Simulation Techniques. In Proceedings of ISCIIA 2018 and
ITCA 2018. Leeds.
Yu, K., Cadeaux, J., Luo, N., Qian, C., & Chen, Z. (2018). The role of the consistency between objective and perceived
environmental uncertainty in supply chain risk management. Industrial Management & Data Systems.
Zepeta, E.D., Nyaga, G.N., & Young, G.J. (2016). Supply chain risk management and hospital inventory: Effects of system
affiliation. Journal of Operations Management.
Zhao, Xiaofei & Wang, Ping & Pal, Raktim. (2020). The effects of agro-food supply chain integration on product quality
and financial performance: Evidence from Chinese agro-food processing business. International Journal of
Production Economics. 231. 107832. 10.1016/j.ijpe.2020.107832.
Zheng, K., & Albert, L.A. (2019). A robust approach for mitigating risks in cyber supply chains. Risk Analysis.
Zineb, E., Brahim, B., & Houdaifa, A. (2017). The Impact of SCRM strategies on supply chain resilience: A quantitative
study in the Moroccan manufacturing industry. Abdelmalek Essaadi
University,6(4),6.http://ijisscm.bsne.ch/ojs.excelingtech.co.uk/index.php/IJSCM/article/download/1665/1665-
6369-1-PB.pdf
Zineb, E., Brahim, B., & Houdaifa, A. (2017). The impact of SCRM strategies on supply chain resilience: A quantitative
study in the Moroccan manufacturing industry. International Journal of Supply Chain management, 6(4), 70-76.
Zubayer, M.A.A., Mithun Ali, S., & Kabir, G. (2019). Analysis of supply chain risk in the ceramic industry using the
TOPSIS method under a fuzzy environment. Journal of Modelling in Management.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

415
 NEW META-HEURISTIC APPROACH FOR SOLVING THE VRPTW
IN SUPPLY CHAIN MANAGEMENT

Nesrine BIDANI1, Hela MOALLA FRIKHA2, Adnan YASSINE3

Abstract ⎯The decision maker can encounter difficult decision problem in the presence of multiple
customer’s request. In fact, the Vehicle Routing Problem VRP is an NP- hard problem such as this paper is in
this context. Indeed, the main contribution of our research work is the development of a new specific
approach, based on meta-heuristic methods, to overcome our problematic in order to optimize a good
approach following of methods allowing to solve the initial decision making problems of the company.
This paper allows solving a VRP with Time Windows (VRPTW) in the context of E-logistic and E-commerce
based on multi-objective optimization.For this reason, we will propose to offer our model to the manager
allowing monitoring the obtained results before applying it by itself. Achieving the objectives of this proposed
approach will facilitate the routine work in several online sales companies and provide optimal approaches.
In fact, our approach consists to achieve these two main objectives: on the one hand emphasize customer
management based on customer satisfaction in order to reply to his expectations and have a lasting
relationship with him for him to be faithful. On the other hand, take an interest in the management of the
transport chain by optimizing the two objectives concerning costs and delivery rounds home.
In this paper, we will present our proposed solving approach with a meta-heuristic method to optimize our
VRPTW problem using instances from the literature. Results will show that our proposed algorithm will
achieve an optimal solution with a reasonable execution time.
Keywords ⎯E-logistic, E-Commerce, Meta-heuristic method, Supply Chain Management, Transportation,
VRPTW.

INTRODUCTION
The supply chain is the actions chain in order to satisfy a customer’s need. Also, it concerns the series of
activities and organisations with financials, informations and physicals flows from initial actor to final
customer. Although, logistic field is the strategic management of the total supply-chain or the time related
arranging of resources. The general reason of logistics is succeeding by high customer satisfaction. It must
offer a good quality service with short and acceptable costs. In our work, taking account into the information
and physical flow between the actors of chain, Figure 1 illustrates that the first action in the Supply Chain
Management (SCM) of our studied problem is the demand sent by a customer through an online demand.
Then, the E-commerce society prepares it in the form of a package. Then, it offers this package to the
distributor in order to transport it to the customer at the lowest cost within a time interval accepted by the
customer to provide good service in order to satisfy their expectations. In SCM field, this paper focus on
Vehicle Routing Problem (VRP) that is very essential in operational research and in logistic domain. In
addition, our studied problem is an NP-hard problem and has many variants allowing for some differents
criterias. VRP can be defined as follow: it is a combinatorial optimization problem that exists in a variety of
distribution systems. It is also a problem of selecting the shortest route from a depot to serve a number of
nodes by considering transport capacity. The problem deals with allocation of vehicles to serve several
customers. Many works for VRP are been studied in literature. Indeed, one of the VRP variants is VRP with
Time Windows (VRPTW) that is our studied problem. The VRPTW has an additional constraint, which is

[email protected], Corresponding author, Department of Quantitative Methods, OLID Laboratory, ISGI Sfax, University of
Sfax, Tunisia
2 [email protected], Department of Quantitative Methods, OLID Laboratory, ISGI Sfax, University of Sfax, Tunisia
3 [email protected], LMAH, University of Havre, France

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

416
called time window, where each customer has different available time and the vehicles must visit the
customers in their available time. In fact, the VRPTW is a complex problem and has many applications in
real-life situations that require minimizing distribution costs. Indeed, our research work focuses in the case of
VRPTW in E-logistic and E-commerce fields based on multi-objective optimization. In addition, this
optimization is based on requests for transport between industry and customers taking into account cost, time,
satisfication services and capacity constraints. So, in this paper, in order to solve this problem and evaluate the
performance of our optimization approach, we present a related works of these optimization algorithms in
order to find the optimal solution of VRP and VRPTW. After analyzing many existing studies, we can see and
conclude that our proposed approach is able to solve our studied problem. In order to achieve the objectives of
this proposed approach, we should facilitate the routine work in several online societies by applying optimal
techniques. Therefore, our two main objectives to be realized are to minimize the cost of transport and
maximize of customer service quality.

Figure 1. Supply Chain Management of our studied problem.

The remainder of this paper is organized as follows: Section 2 presents a brief review of studied problems in
SCM and a related works of meta-heuristics algorithms. Then, we describe, in section 3, the meta-heuristics
algorithms with the presentation of GA and SA techniques. In Section 4, we will propose an approach
allowing solving the decison problem by applying a mathematical formulation and meta-heuristic methods.
Section 5 demonstrates numerical examples and discusses the obtained computational results. Finally, we will
briefly conclude the realized work in Section 6 and open some perspectives.

LITURATURE REVIEW

In this part, we present a literature of review of meta-heuristics methods application in our context. In fact, the
VRPTW is a complex problem and has many applications in real-life situations that require minimizing
distribution costs. The VRPTW can be defined as follow: there are a number of customers C, several vehicles
V, and a depot d. Each customer has units of demand and the vehicles have maximum units to be carried. The

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

417
aim of the problem is find the set of routes starting and ending at d. Each customer in C must be visited by
exactly one vehicle in V. An amount of time is required by the vehicle on the road and during servicing the
customer. An additional constraint called time window forces the vehicle to service the customers in their
available time. The vehicle must wait if it arrives earlier than the opening time of the customer. However, late
arrival of the vehicle is not permitted. Various objective functions have been addressed in the existing studies.
However, most of the studies consider reducing the number of vehicles and the total travel distance of the
vehicles simultaneously.
VRP and VRPTW problems in Supply Chain Mangement
Many researches increasing interest in the last decade summarize a recent literature in VRP problem in SCM.
Álvarezand Munari (2016) prepared research paper addresses the vehicle routing problem with time windows
and multiple deliverymen, a variant of the vehicle routing problem which includes the decision of the crew
size of each delivery vehicle, besides the usual scheduling and routing decisions. This problem arises in the
distribution of goods in congested urban areas where, due to the relatively long service times, it may be
difficult to serve all customers within regular working hours. Given this diffculty, an alternative consists in
resorting to additional deliverymen to reduce the service times, which typically leads to extra costs in addition
to travel and vehicle usage costs. The objective is to define routes for serving clusters of customers, while
minimizing the number of routes, the total number of assigned deliverymen, and the distance travelled. Two
metaheuristic approaches based on Iterated Local Search and Large Neighbourhood Search is proposed to
solve this problem. The performance of the approaches is evaluated using sets of instances from the literature.
Cassettari (2018) is focused on a variant of VRP, the CVRP when applied to natural gas distribution networks.
Managing natural gas distribution networks includes facing a variety of decisions ranging from human
resources and material resources to facilities, infrastructures, and carriers. Despite the numerous papers
available on VRP, there are only a few that study and analyze the problems occurring in capillary distribution
operations such as those found in a metropolitan area. So, he introduces a new algorithm based on the Saving
Algorithm heuristic approach, which aims to solve a CVRP with time and distance constraints. This joint
algorithm minimizes the transportation costs and maximizes the workload according to customer demand
within the constraints of a time window. The effectiveness of the approach is demonstrated by the results from
a real case study in a natural gas distribution network. Adhi et al (2019) proposed a new metaheuristic
algorithm to solve VRP in order to achieve optimal solution. This new optimization algorithm that is Partial
Comparison Optimization (PCO) was developed to solve combinatorial optimization problems such as VRP.
In this study, they based on PCO to solve the problems that existed in the origin VRP. To prove PCO is a
good metaheuristics for solving VRP, several of instances of symmetrical VRP were selected from the VRP
library to evaluate its performance. The numerical results obtained from the calculation indicated that the
proposed optimization method could achieve results that almost similar with the best-known solutions within a
reasonable time calculation. It showed that PCO was a good metaheuristics to solve VRP. Vidalet al., (2020)
provided a concise overview of existing and emerging problem variants. Model sare typically refined along
three lines: considering more objectives that are relevant and performance metrics, integrating vehicle routing
evaluations with other tactical decisions, and capturing fine grained yet essential aspects of modern supply
chains. They organized the main problem attributes within this structured framework. They discuss recent
research directions and pinpoint current short comings, recent successes, and emerging challenges.

GA and SA meta-heuristics algorithms

Hosny et al (2010) proposed a research work and tried to handle the difficult constraints using a new solution
representation and simple neighbourhood moves that will maintain the feasibility of solutions throughout the
search. The proposed solution method is tried within two meta-heuristic approaches, a Genetic Algorithm
(GA) and a Simulated Annealing (SA). Based on the performance of the two algorithms on a number of
benchmark problems, they concluded about which among the two algorithms seems more appropriate for
solving the problem. Focusing on the existing problem in e-commerce logistics system, Yanhui et al., (2013)
formulated a location-inventory-routing problem model with no quality defects returns. To solve this NP-hard
problem, an effective hybrid genetic simulated annealing algorithm (HGSAA) is proposed. Results of
numerical examples show that HGSAA outperforms GA on computing time, optimal solution, and computing
stability. The proposed model is very useful to help managers make the right decisions under e-supply chain
environment. Nasab el al., (2013) prepared a research work when the location-routing problem with fuzzy
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

418
demands (LRPFD) is considered which may arise in many real life situations in logistics management, and a
fuzzy chance constrained program is designed to model it, based on the fuzzy credibility theory. A hybrid
simulated annealing (SA) based on heuristic incorporated with stochastic simulation is developed and
proposed to solve the problem. The efficiency of the solution procedure is demonstrated via comparing its
performance with those of some other existing solution procedures from literature using a standard benchmark
set of test problems. Ariyani et al., (2018) proposed in research work when genetic algorithm is hybridized
with simulated annealing algorithm to solve the problem. Genetic algorithm is employed to explore global
search area and simulated annealing is employed to exploit local search area. Four combination types of
genetic algorithm and simulated annealing (GA-SA) are tested to get the best solution. The computational
experiment shows that GA-SA1 and GA-SA4 can produced the most optimal fitness average values with each
value was 1.0888 and 1.0887. However GA-SA4 can found the best fitness chromosome faster thanGA-SA1.

Impact of GA and SA for VRP

Many studies are presented in literature for the application of the GA and SA algorithms to solve their
problems with diverse VRP variants. Thonethong (2016) proposed a new metaheuristic approach by
Simulated Annealing (SA) algorithm that is developed to generate delivery routes in which both demands can
be satisfied in the same delivery routes while minimizing the transportation cost. The algorithm was tested
with standard problem instances of CVRP. The results from testing the algorithm using numerical examples
shows that there is a trade off between additional cost of allowing delivery of transshipment demands on the
same trip and the benefit of reducing delivery lead time of transshipment demand. Rabbaniet al., (2017)
studied a new kind of green vehicle routing problem called multi-compartment green vehicle routing problem
(MCGVRP) is presented. The aim of this problem is to minimize the cost of changes in load, speed and
payment to drivers by considering pollution emitted by vehicle. Two meta-heuristic algorithms including
genetic algorithm (GA) and Simulated Annealing (SA) are selected to solve the presented problem along with
a hybrid metaheuristic algorithm. So, the results obtained by these methods are compared with each other. In
the study of Siswanto (2019), a discussion of the creation of routing system and proposed business processes
for the delivery company by applying the Asymmetric Capacitated Vehicle Routing Problem (ACVRP) using
Simulated Annealing metaheuristic algorithm with a trade-off between the total distance and the maximum
distance of one route 0.1; 0.5; and 0.9. The routing system optimization is adjusted to the needs of the decision
maker to focus its route determination, whether it is focused on the total distance or the maximum distance of
one route. The proposed business process can saves as much as 2 hours.

Impact of GA and SA for VRPTW

Mahmudy (2014) introduced a research paper that addresses the VRPTW by using an improved simulated
annealing algorithm. Special functions to effectively exploring neighbourhood solutions are developed. The
functions are required to deal with the large search space of the VRPTW and enhance the power of the
simulated annealing to obtain better solutions. The proposed approach is evaluated in comparison with well-
known benchmark problems available in the literature. A set of computational experiments prove that the
improved simulated annealing could produce promising results in the average of computational time of 82.29
seconds. Kumar et al.(2015) used the heuristics and metaheuristics algorithms which are of the combinatorial
optimization type. Although, they prepared a literature review of VRPTW, TDVRP, and a metaheuristic such
as the genetic algorithm that was conducted. In this work, they used the implementation of Matlab and
HeuristicLab optimization software of the VRPTW and its extension. The time-dependent VRPTW
(TDVRPTW) has been carried out, using the model as well as metaheuristics such as the genetic algorithm
(GA). The results were comparable to the earlier algorithms developed and in some cases the current
algorithm yielded better results in terms of total distance travelled and the average number of vehicles used.

DESCRIPTION OF META-HEURISTICS ALGORITHMS

In this part, we present a literature review of meta-heuristics methods. According to Khalilpourazari and
Khalilpourazary (2017), Khalilpourazari & Arshadi Khamseh (2017), Dhiman G.&. Kumar (2017), Mirjalili
& Lewis (2016), there are many available classifications for metaheuristic algorithms in the literature. So,
in this paper, we focus on two important categories of Meta-heuristic Algorithms such as Evolutionnary
© 19th International Logistics and Supply Chain Congress
October 21-22, 2021, Gaziantep, Turkey

419
Algorithms and Physics based methods. In addition, we focus to apply Genetic Algorithm (GA) among the
first category and Simulated Annealing (SA) among the second category.

GA algorithm
GA is an evolutionary algorithm category that often performs optimal solutions to many problems because
this category does not make any hypothesis about the adaptive landscape or basic fitness. This algorithm
has many advantages like: It is a foundation for many evolutionary algorithms and simulates the natural
process of evolution like Selection, Mutation and Crossover (Goldberg, 1989). This algorithm uses
chromosomes in order to exploit and explore the solution space in determining and enhancing the value of
near-optimal solutions. Also, GA is a well-known algorithm that has been improved in many ways, is easy
to implement, is fast convergency and can be used independently to solve a given problem. It does not
depend on other algorithms or heuristics, It has the ability to handle random types of objectives and
constraints, GA uses simple operators and can be used to solve problems that have high computational
complexity.
SA algorithm
SA is a physic based algorithm that is made the optimization using the physics rules in the world. It is an
inspired algorithm by the heating and cooling of metals. It focuses on the notion of temperature that is an
important parameter on applying this method. In addition, it is usually reducing by a few predefined factors at
the end of each iteration. Also, it is relatively easy to implement, even for complex problems, can deal with
arbitrary systems and cost functions and generally gives a best solution, statistically guarantees finding an
optimal solution (Kirkpatrick et al, 1983).

Figure 2 presents the architecture of the both of GA and SA algorithms. It illustrates the different steps of each
method.

Figure 2. Diagram of architecture of (a) GA and (b) SA algorithms.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

420
 PROPOSED APPROACH

To optimize our problem, we should choose a technique by applying some existed or new proposed methods.
In this paper, we focus in VRPTW in E-Logistics in supply chain field. We desire find the best solution with
the best performance and the best CPU to achieve the main objectives of our studied problem and obtain the
optimum solution with metaheuristic algorithms. The resolution of VRPTW problem is important in E-
commerce and E-Logistic domain. Therefore, in future work, many directions can be serving as prior
knowledge for implementation in order to propose an algorithm based on meta-heuristics techniques.
Problem Description

In this section, a mathematical model for the VRPTW is presented. For a society X, we consider K Vehicles
offer one or more packages to N customers according to customer’s requests in the limited and accepted time
by the customer. Our objective is to minimize the total cost and maximize the quality of customer service. In
addition, we try an approach in order to obtain a feasible routing, the optimum routing, number of vehicles
and optimal cycle time for the both of customers and society transport. Therefore, the goal of this research
work is to apply a meta-heuristic approach to solve our studied problem to obtain an optimal solution by
applying a best algorithm using differents optimization techniques to solve our problem using some instances.

Figure 3. Example of optimal solution to VRP variant.

Modeling Of Problem By A Mathematical Formulation

Our problem is a transport on demand considering the time (VRPTW) and the quality of customer service.
We try to formulate our problem by a multi-objective mathematical model taking into account the following
assumptions and notations that are shown in table 1 as follow:

Table 1. Assumptions and notations of Mathematical formulation

▪ K : Total Number of vehicle available (such as k=1,..K) (Set of Vehicle)

▪ N : Number of node (Customer and deposits) (such as i, j= 0, .., N+1) (Set of all nodes)
Set, Indices ▪ 0, N+1: Initial depot and final of set of vehicles.

▪ 𝑐𝑖𝑗𝑘 : Cost of transport to the arc (i, j),

Parameters (Data) ▪ 𝑐𝑎𝑝𝑘 : Capacity of vehicle k, with k є {1, ..., K}.
▪ 𝑤𝑖 : Weight of package to be delivered of customer i.
▪ 𝐴𝑘𝑖 : Arrival time of vehicle k in node i.
▪ 𝑡𝑖 : Service time associate in node i.
▪ 𝑙𝑖 : Maximal time of reception of package to be delivered by the node i.
▪ 𝑇𝑚𝑎𝑥 : Maximal time of use of vehicle k.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

421
 ▪ 𝛽𝑖 : Coefficient of delay penalty if vehicle k exceeds 𝑙𝑖 .
▪ 𝛽𝑖 = 𝜀𝑐𝑖𝑗𝑘 et 𝑃e𝑛𝑎𝑙𝑡𝑦 = max (0, 𝛽𝑖 (𝐴𝑘𝑖 − 𝑙𝑖 )
▪ 𝑀𝑒𝑘 : Mechanic available for the vehicle k.

▪ 𝑥𝑖𝑗𝑘 : Binary variable such as i, j ∈ 𝑁, 𝑘 ∈ 𝐾 𝑒𝑡 𝑖 ≠ 𝑗

Decision Variables ▪ 𝑥𝑖𝑗𝑘 = 𝟏 𝑖𝑓 𝑡ℎ𝑒 𝑐𝑜𝑠𝑡𝑢𝑚𝑒𝑟 𝑗 𝑖𝑠 𝑣𝑖𝑠𝑖𝑡𝑒𝑑 𝑎𝑓𝑡𝑒𝑟 𝑡ℎ𝑒 𝑐𝑜𝑠𝑡𝑢𝑚𝑒𝑟 𝑖 𝑤𝑖𝑡ℎ 𝑡ℎ𝑒 𝑣éℎ𝑖𝑐𝑙𝑒 𝑘, 𝟎 𝑒𝑙𝑠𝑒
▪ In other words:
𝑥𝑖𝑗𝑘 = 𝑻𝒓𝒖𝒆 𝑖𝑓 𝑡ℎ𝑒 𝑛𝑜𝑑𝑒 𝑗 𝑖𝑠 𝑣𝑖𝑠𝑖𝑡𝑒𝑑 𝑎𝑓𝑡𝑒𝑟 𝑡ℎ𝑒 𝑛𝑜𝑑𝑒 𝑖 𝑤𝑖𝑡ℎ 𝑡ℎ𝑒 𝑣éℎ𝑖𝑐𝑙𝑒 𝑘, 𝑭𝒂𝒍𝒔𝒆 𝑒𝑙𝑠𝑒

Objective function Our objective function, Equation 1 expressed in the next section, allows to:
▪ Minimizes the total cost of transport.
▪ Maximizes the quality of customer service: in other words minimizes the dissatisfaction of
customer or minimizes the penalty (reduce delay in delivery) by respecting the delivery
time.

Multi-Objective Mathematical Model

𝑁+1 𝑁+1 𝑁+1

𝐾
𝑘 𝑘
𝑀𝑖𝑛 ∑ ∑∑ 𝑐𝑖𝑗 𝑥𝑖𝑗 + ∑ 𝑚𝑎𝑥 (0, 𝛽𝑖𝑘 (𝐴𝑘𝑖 − 𝑙𝑖 )) (1)
𝑘=1
𝑖=0 𝑗=0 𝑖=1

∑ 𝑥 𝑁 𝑘
∑𝑁
𝑖=0 𝑗=0 𝑖𝑗
≤1∀𝑘 ∈𝐾 (2)

∑𝑁 𝑘
𝑗=1 𝑥0𝑗 = 1 ∀𝑘 ∈𝐾 (3)
∑𝑁 𝑘
𝑖=0 𝑥𝑖,𝑁+1 = 1 ∀𝑘 ∈𝐾 (4)
∑𝑁+1 𝑘
𝑖=1 𝑥𝑖𝑗 = ∑𝑁+1
𝑖=1 𝑥𝑗𝑖𝑘 ∀𝑗 ∈𝑁 (5)
∑𝑁+1 𝑁+1 𝑘
𝑖=0 ∑𝑗=0 𝑤𝑖 𝑥𝑖𝑗 ≤ 𝑐𝑎𝑝𝑘 ∀𝑘 ∈𝐾 (6)
(𝐴𝑘𝑖+ 𝑘
< 𝐴𝑗𝑘
𝑡𝑖 ) 𝑥𝑖𝑗 ∀ 𝑖 ∈ 𝑁, ∀ 𝑘 ∈ 𝐾 (7)
𝐴𝑘𝑖 ≤ 𝑙𝑖 ∀ 𝑖 ∈ 𝑁, ∀ 𝑘 ∈ 𝐾 (8)
ST ∑𝑁+1 𝑁+1 𝑘
𝑖=0 ∑𝑗=0 (𝐴𝑖 + 𝑡𝑖 ) ≤ 𝑇𝑚𝑎𝑥 ∀𝑘 ∈𝐾 (9)
∑𝐾
𝑘=1 𝑀𝑒𝑘 ≥ 1 (10)
∑𝐾
𝑘=1 𝑘 ≥ 1 (11)
(𝐴𝑘𝑖 + 𝑡𝑖 ) ≤ 𝑇𝑚𝑎𝑥 (12)
𝑘
𝑥𝑖𝑗 ∈ {0,1} (13)

Discussion

The explication of constraint is as follow:

Equation 2: the vehicle k passes through at most once per customer.
Equation 3: the vehicle k comes out only once from the depot ‘’0’’.
Equation 4: After delivery, the vehicle k returns only once to the final depot which is the same from the
beginning.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

422
Equation 5: fuzzy constraint shows that the input of node i = the output of node i.
Equation 6: the maximum capacity of the vehicle k must not be exceeded.
Equation 7: The arrival time of the customer i + service time does not exceed the arrival time of the next node.
Equation 8: The arrival time of the vehicle k at node i is at the maximum li of the time window.
Equation 9: The working time of the vehicle k must not be exceeded by the passage of all the nodes.
Equation 10: At least one mechanic available for the vehicle k which is down.
Equation 11: At least one vehicle k available to achieve a service.
Equation 12: The maximum delivery time must not be exceeded.
Equation 13: Binary values.

This formulation is a static model to be solved by a meta-heuristic method and not exact method because of
the second objective in our objective function that is non-linear. That’s why these characteristics give reason
for the model to be effectively hard to be solved using exact algorithms. To pact with the complexity and find
optimal results in a realistic computational time, we want apply meta-heuristic approaches which are
extensively used for which the GA and the SA. In addition, the constraints of our model are linear.

RESOLUTION, IMPLEMENTATION AND RESULTS

In this section, we present solving approach with SA and GA that are Meta-heuristic methods to optimize our
VRPWT problem using Soloman’s Instances. The resoltuion of problem is investigated using the data in
literature. To find optimum solutions, the GA and the SA were coded in MATLAB 7.14.0.739, and run on a
laptop equipped with an Intel Core i5 processor working at a speed of 2.67 GHz and 4 GB of RAM.
Results show that evolutionary method achieved significant accuracy (performance) for better solution.
In order to obtain the values of the algorithms’ parameters, each algorithm is applied through pilot runs, each
time changing the parameters in their corresponding ranges and obtaining the response values.
The parameters settings of two employed algorithms are given in Table 1.

Table 2. GA and SA parameters and its values.

Meta-heuristic Parameters Values

Algorithms

▪ Population size 80
GA ▪ Crossover 0.9
▪ Mutation 0.05
▪ Number of generations 1000

▪ Annealing function Fast annealing

▪ Reannealing interval 100
SA ▪ Temperature update function Exponential temperature update
▪ Initial temperature 100
▪ Acceptance probability function SA acceptance
▪ Start point Set that all variables take lower bound
values

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

423
 GA for VRPTW Resolution
Figure 4 presents the solution of problem by applying the GA.

Figure 4. GA for VRPTW Resolution.

SA for VRPTW Resolution

Figure 5 shows the solution of problem by applying the SA algorithm.

Figure 5. SA for VRPTW Resolution.

Revised SA for VRPTW Resolution

Figure 6, 7 and 8 present the solution of problem by applying the Revised SA for three different instances.

Figure 6. Solution by Revised SA algorithm for Instance (8*3).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

424
 Figure 7. Solution by Revised SA algorithm for Instance (5*24).

Figure 8. Solution by Revised SA algorithm for Instance (8*70).

Discussion
In order to compare the performances of the applied of algorithms, each problem instance has been run many
iterations. Fistly, by applying GA, we obtained the solution presented in figure 4. Secondly, concerning SA
application solution that is illustrated in figure 5, we realized 100 iterations to find the best cost 375. And
thirdly, by applying revised SA, the number of iterations is 1200 times with the parameters settings given in
Figure 6, 7 and 8. Such as the obtained objective function values of costs of 220.16, 326.7 and 373.09, are
reported in Figure 6, 7 and 8 respectively. Table 2 represents the best obtained solutions by the GA and the SA
accompanied with other approaches given in the literature for diverses costs. Therefore, we can affirm and
outperform that the revised SA algorithm founds the best-known solutions.
The solution of the proposed mathematical model reach that the existing techniques are based on the
complexity and the required computational efforts associated with the model. So, efficient meta-heuristic
techniques are necessary to solve the NP-hard model for big problems usually found in real situations.
In this paper, effective solution approaches based on GA, SA and revised SA algorithms for binary variables
are presented for solving the proposed multi-objective optimization model. The results indicate the efficiency
of the applied metaheuristic methods in solving the proposed model. All the applied methods showed a
remarkable performance and best solution qualities. In addition, the metaheuristic algorithms can efficiently
touch a big size of instances in a reasonable computational time.
In fact, the application of revised SA presents an acceptable proposed approach and produces the optimal cost,
which may indicate its superiority in searching for solutions of similar problems. For that reason, we can be
served as a useful decision support tool for managers of society.
However, in the dynamic TSP and VRP, we do not know all the information for planning the routes when the
routing process begins. In particular, after constructing the initial routes, the information can be updated due
to various external factors. To transform our problem from static model to dynamic model, it is necessary to

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

425
add the time index. We want improve this solution by proposing hybrid method between SA and another
meta-heuristic algorithm.

CONCLUSION
The managers of society want follow a good decision in the case of difficult decision problem in the
presence of multiple customers’ request. Indeed, the context of this paper is in the VRPTW in E-Logistics and
E-commerce domain that is an NP- hard problem. That’s why we want finding the best soltuion. In order to
overcome our problematic and to optimize using a good method that can solve the decision making problems
of the society, we proposed an approach based on the multi-objective mathematical formulation of studied
problem and the application of meta-heuristic algorithms. Therefore, we obtain the best solution with the best
performance and the best CPU in order to realize the main objectives of problem. The results show that our
proposed algorithm achieves an optimal solution with a reasonable execution time. Therefore, in future work,
in order to propose an algorithm based on meta-heuristics algorithms among their three cited category, we are
in front of many directions which can be serving as prior knowledge for implementation. In addition, to create
more efficient algorithm for global optimization.most of the times, the hybrid metaheuristic techniques are
efficient and robust in comparison to the basic versions of the algorithms that are hybridized because they can
profit from advantages of several applied metaheuristic algorithms.

REFERENCES

Adhi et al (2019),“A new metaheuristics for solving vehicle routing problem: Partial Comparison Optimization”, Annual Conference on Industrial and
System Engineering (ACISE) 2019,IOP Conf. Series: Materials Science and Engineering 598 (2019) 012023, doi:10.1088/1757-899X/598/1/012023

A. K. M. Foysal Ahmed and Ji Ung Sun, (2018) Bilayer Local Search Enhanced Particle Swarm Optimization for the Capacitated Vehicle Routing
Problem

Amalia Kartika Ariyani, Wayan Firdaus Mahmudy, Yusuf Priyo Anggodo Hybrid Genetic Algorithms and Simulated Annealing for Multi-trip Vehicle
Routing Problem with Time Windows International Journal of Electrical and Computer Engineering (IJECE)

Aldair Álvarez (2016) Pedro MunariMetaheuristic approaches for the vehicle routing problem with time windows and multiple deliverymen
Sukanya Thonethongsimulated annealing algorithm for vehicle routing problem with transshipment, a thesis submitted, Academic year 2016

Budi Nur Siswanto, Afferdhy Ariffien, Ilham Jayakusuma (2019) SISTEM ROUTING PROSES DELIVERY MENGGUNAKAN SIMULATED
ANNEALING (Studi Kasus: PT. X) Jurnal Teknologia Aliansi Perguruan Tinggi (APERTI) BUMN Vol. 2, No.1, Februari 2019, ISSN 2654-5683

Caceres-Cruz et al (2014) Rich Vehicle Routing Problem: Survey; Article in ACM Computing Surveys · December 2014, DOI: 10.1145/2666003

Han and Wang (2018), “A Survey for Vehicle Routing Problems and Its Derivatives”, IOP Conference Series: Materials Science and Engineering,
IMMAEE 2018.

Dhiman G.&. Kumar V , “Spotted hyena optimizer: A novel bio-inspired based metaheuristic technique for engineering applications “, Advances in
Engineering Software 114 (2017) 48–70

ErdeliT and CariT (2019),A Survey on the Electric Vehicle Routing Problem: Variants and Solution Approaches, Hindawi Journal of Advanced
Transportation, Volume 2019, Article ID 5075671, 48 pages, https://doi.org/10.1155/2019/5075671

Eberhart R. & Kennedy J.. “A New Optimizer Using Particle Swarm Theory”, In proceedings of the Sixth International Symposium on Machine and
Human Science, pp. 39-43, 1995

Feliu JG (2013), Modèles et méthodes pour la logistique urbaine: les problèmes de tournées de véhicules à deux échelons, Cahier de Recherche LET.

Goldberg D. E.. “Genetic Algorithms in Search, Optimization, and Machine Learning”, ADDISON-WESLEY PUBLISHING COMPANY, 1989
Kirkpatrick S., Gelatt D. Jr., & Vecchi M. P., "Optimization by simulated annealing", Science, 220(4598), pp 671–680, 1983

Hosny (2010) Comparing Genetic Algorithms and Simulated Annealing for Solving the Pickup and Delivery Problem with Time Windows

Khalilpourazari S, Mohammadi M (2016) Optimization of closed-loop Supply chain network design: a Water Cycle Algorithm approach. 12th Int Conf
on Indust Eng 41-45. IEEE. doi: 10.1109/INDUSENG.2016.7519347

Khalilpourazari S, Khalilpourazary S (2017) A Robust Stochastic Fractal Search approach for optimization of the surface grinding process. Swarm
Evolu Comput. doi: 10.1016/j.swevo.2017.07.008

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

426
Khalilpourazari S, Khamseh A.A. (2017) Bi-objective emergency blood supply chain network design in earthquake considering earthquake magnitude:
a comprehensive study with real world application. Ann Oper Res 1-39. doi:10.1007/s10479-017-2588-y

Khalilpourazari S, Khalilpourazary S (2016) Optimization of production time in the multi-pass milling process via a Robust Grey Wolf Optimizer.
Neural Comput Appl 1-16. doi:10.1007/s00521-016-2644-6

Khalilpourazary S, Kashtiban PM, Payam N (2014) Optimization of Turning Operation of St37 Steel Using Grey Relational Analysis. J Comput Appl
Res Mech Eng 3:135-144. doi: 10.22061/JCARME.2014.67

Kumar(2015) A Time-Dependent Vehicle Routing Problem with Time Windows for ECommerce Supplier Site Pickups Using Genetic Algorithm,
Article in Intelligent Information Management · January 2015

Leonora Bianchi, Mauro Birattari, Marco Chiarandini, Max Manfrin, Monaldo Mastrolilli, Luis Paquete, Olivia Rossi-Doria, Tommaso Schiavinotto
(2000), Hybrid Metaheuristics for the Vehicle Routing Problem with Stochastic Demands Journal of Mathematical Modelling and Algorithms
manuscript.

Li Y, Guo H, Wang L, Fu J (2013). A hybrid genetic-simulated annealing algorithm for the location-inventory-routing problem considering returns
under e-supply chain environment. ScientificWorldJournal. 2013;2013:125893. doi:10.1155/2013/125893

Lucia Cassettari, Melissa Demartini *, Roberto Mosca, Roberto Revetria and Flavio Tonelli, (2018) A Multi Stage Algorithm for a Capacitated Vehicle
Routing Problem with Time Constraints

Mirjalili S, Mirjalili SM, Hatamlou A (2016) Multi-verse optimizer: a nature-inspired algorithm for global optimization. Neural Comput Appl 27:495-
513. doi: 10.1007/s00521-015-1870-7

N. Ghaffari-Nasab et al. / Scientia Iranica, Transactions E: Industrial Engineering 20 (2013) 919–930

Nathalie De Jaegere, Mieke Defraeyea and Inneke Van Nieuwenhuyse (2013) The Vehicle Routing Problem: State of the Art Classification and Review
Silva et al (2005), “A Logistic Process Scheduling Problem: Genetic Algorithms Or Ant Colony Optimization”, 2005 Ifac

Olatz Arbelaitz and Clemente Rodriguez (2014) Comparison of systems based on evolutionary search and simulated annealing to solve the VRPTW
problem International Journal of Computational Intelligence and applications vol. 04, No. 01, pp. 27-39 (2004)

Song Mei-xian, Jun-qing Li, Yun-qi Han, Yu-yan Han, Li-li Liu, Qun Sun (2020), Metaheuristics for solving the vehicle routing problem with the time
windows and energy consumption in cold chain logistics

Stodola (2018) Using Metaheuristics on the Multi-Depot Vehicle Routing Problem with Modified Optimization Criterion, Algorithms 2018, 11, 74; doi:
10.3390/a11050074

SMAILI.Yamina, SENOUSSI .Ahmed, MOUSS. Nadia. Kinza 2013; Metaheuristic Approach For Solving The Vehicle Routing Problem: Application
In Pharmaceutical societyCoDIT'13.

Thibaut Vidal, Gilbert Laporte, Piotr Matl (2020), A concise guide to existing and emerging vehicle routing problem variants arXiv: 1906.06750v2
[cs.DM] 5 Apr.

Rabbani M.; Z. Tahaei; H. Farrokhi-Asl; N. Akbarin Saravi Using meta-heuristic algorithms and hybrid of them to solve multi compartment Vehicle
Routing Problem 2017 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM)

Wayan Firdaus Mahmudy (2014)Improved simulated annealing for optimization of vehicle routing problem with time windows (VRPTW)110
KURSOR Journal Vol. 7, No. 3, October 2014, page 109-116.

Wang et al (2017); Wang Fengming, Shi and Wen, “Optimization of Vehicle Routing Problem with Time Windows for Cold Chain Logistics Based on
Carbon Tax”, Sustainability 2017, 9, 694; doi:10.3390/su9050694

Yun-qi Han, Jun-qing Li, Zhengmin Liu, Chuang Liu and Jie Tian (2020) Metaheuristic algorithm for solving the multi-objective vehicle routing
problem with time window and drones International Journal of Advanced Robotic Systems March-April: 1–14

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

427
ANALYSIS OF THREAT IN GLOBAL SUPPLY CHAINS CREATING
LOGISTICS BARRIERS DURING A GLOBAL PANDEMIC

Zbigniew Bentyn1, Sylwia Konecka2

Abstract ⎯ During a global pandemic, operations in the global supply chain have met an unusual risk. The
purpose of this study consist of research and description of a new threat for maintaining a flow in the supply
chain. Spatial distribution of manufacturing and distributive operations revealed a particular vulnerability of
supply chains. Many areas of global supply chains suffer disturbances in everyday operations. Such cases
may cause the creation of logistic barriers, posing a challenge for logistic operators. Analysis of many
research reports reveals critical vulnerabilities of a global supply chain.

Keywords ⎯ global supply chain, logistic barrier, disruption, pandemic

INTRODUCTION
The development of the international economy and consequently global supply chains necessitates analysing
potential threats to their operation. The changing and complex international environment is a source of many
threats and challenges for the functioning of supply chains.
Unfortunately, experiences of the years 2020-21 proved that the catalogue of potential threats is growing, as is
the complexity of these international structures and the demands placed on them. Expectations that
international logistic flows will ensure uninterrupted continuity of supply in the global market result from the
perceptions of end customers who perceive the waiting time for delivery as a time wasted. The impatient
customers form market pressing on supply chain managers to accelerate transport and logistics operations to
improve their competitive position. Covid-19 pandemic has resulted in unprecedented disruption in global
supply chains. National lockdowns, closed airspace and borders, the lack of international air connections,
shortages of certain essential products, driver shortages and cross-border restrictions shrank road freight
capacity. Ocean carriers removed shipping capacity from the market, and air freight capacity dropped because
passenger traffic dried up. Simultaneously we observed fluctuating demand shifting towards e-commerce.
Sudden changes, as in Italy, where e-commerce sales of consumer products rose by 81% in a single week,
increase pressure on distributors and finally logistic operators. Caused by the anti-pandemic measures, long-
term problems with manufacturing operations follow the known pattern of a bullwhip effect in the supply
chain. Subsequent issues with supply effect in decreased availability of products and increased prices for the
end consumer. Many pandemic threats and disturbances create logistic barriers to the supply flow, thus
increasing logistics costs to maintain supply chain operations. The erection of these logistics barriers forces to
innovate and modernise existing supply chains. That is one of the reasons companies operating in a global
environment re-evaluating locations manufacturing operations and signal a strategic shift towards nearshoring
to decrease the risk involved in global operations.

LITERATURE REVIEW AND THEORETICAL FRAMEWORK

To prepare a theoretical base for analysis, there is a need to look closer at definitions of various forms of
threats in the supply chain, which may form a logistic barrier. Ideas of factors confronting the flow in supply
chains are well-grounded in literature and various in form. To describe them in the order it is convenient to
start from the broadest idea of Uncertainty. Operating in conditions of uncertainty is an inherent feature of
human activity. The multitude of factors influencing such activities as well as their potential effects,
significantly increase the risk, which is particularly important in the case of business activities. The operation
of supply chains is, therefore an example of cooperation between companies in conditions of uncertainty. The

1
Zbigniew Bentyn, Poznań University of Economics and Business, Institute of International Business and Economics, Department of
Logistics, Poznań, Poland, [email protected]
2
Sylwia Konecka, Poznań University of Economics and Business, Institute of International Business and Economics, Department of
Logistics, Poznań, Poland, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

428
classic approach to the concept of uncertainty is presented in the definition of J. Goldstein, which compares
uncertainty to an element of the environment characterised by a high level of variability. According to him,
this variability is the reason for the necessity to adapt the organisation (Goldstein J., 1994). An operational
recommendation resulting from the company's strategy is also such an influence towards the area of
uncertainty that results in its reduction (Mason-Jones, R., & Towill, D., 1998). If we assume, after Dixit, A.
and Pindyck, R., that the uncertainty may come from endogenous and exogenous sources for the enterprise, it
should be indicated that while management activities may reduce those arising inside enterprises, those
coming from outside the supply chain remain beyond the control of managers (Dixit, A., Pindyck, R., 1994).
In the case of exogenous threats, only the emergence of the threat and its duration in a given period allows
managers to work out ways to limit its impact on the supply chain. This fact means that gaining experience is
a potential remedy for new, previously unknown threats. An Idea following Uncertainty – Threat, is more
precise yet not measurable. A supply chain threat is the primary form of a condition that can significantly
affect the operation of the supply chain. It's rather general nature is related to the difficulty of predicting what
kind of specific effects for the supply chain it may bring. The catalogue of threats changes, as does the global
economy, along with the resources and opportunities available for given supply chains. Today's threats often
focus on the security of data transmission, their quality, and the sensitivity of IT systems to intrusion attempts,
as well as distorting the transmitted data (Chopra S., Sodhi M., 2004, p. 54). According to T. Kaczmarek, a
threat is described as a certain danger compared to risk, which is only an estimated probability, usually lower
than certainty (T. Kaczmarek, 2008, p. 52). Thus, a threat in the supply chain is a potential risk situation. On
the other hand, the risk is a measurable value of the probability of the effects of a given threat (Kisieleski P.,
Stanek A., 2017). The definition of risk linking the aspects of the supply chain competitiveness with the
occurrence of factors described as a triggering event is the proposal of I. Heckmann et al. They describe the
supply chain's risk as a potential loss for this chain caused in two areas characterizing the supply chain. The
first is preventing the supply chain's value-added performance goals from being met. The second area is
preventing the achievement of value-for-money goals in terms of efficiency. In any case, the occurrence of
trigger events is responsible for the loss formulated in this way (Heckmann I., Comes T., Nicke S., 2015).
Narrowing even further the scope of research on the effects influencing the supply chain helps to understand
another idea - the level of disturbances. D. Ivanov explains the disturbance as a result of the described risk in
the supply chain (Ivanov D., Sokolov B., 2010). It follows that in order to identify a disorder, a causative act is
required to implement the anticipated risk (Barroso et al.). At the same time, it specifies that there is a direct
impact on the normal functioning of the supply chain as a result of a specific event. This influence can be
manifested in two ways. In the first case, the normal operation of the supply chain is disrupted. In the second,
the stability of the organisation or the entire supply chain is compromised. Regarding the concept of
previously defined risk, he argues that the disorder can trigger a predicted or unforeseen event (Barroso, A. P.,
Machado, V. H., Cruz Machado, V., 2008). Taking the perspective of the effects of the disturbance, S.
Konecka believes that a disturbance is the inability to carry out planned activities or a certain critical number
of activities enabling the further functioning of the supply chain (Konecka S., 2015). An idea closing this
comparative analysis is Disruption. Handfield R, McCormack K. (2008) define disruption as major delays in
production or distribution nodes that have consequences in the activities of other nodes in the supply chain.
Disruptions are usually a bottleneck at one of the nodes, which in turn spreads through the entire supply chain.
Every single event, such as fire, problems with the quality of manufactured products, machine failures,
delayed customer orders, may cause significant disruptions in the entire supply chain. Problems related to
identified supply chain disruptions can have a number of consequences. They also include influencing
investors through the media, looking for information that may affect the value of shares. It is widely assumed
that vulnerable supply chains increase operating costs and decrease their efficiency. As a result, delays in the
production and transportation processes in the global supply chain are expected (Hendricks K., Singhal V.,
2003). This opinion further emphasises the essence of the disruption - delay. According to Blackhurst et al.,
Reducing the response time to a delay can significantly reduce the level of risk of disruption. Quick response
is possible thanks to increased transparency of operations in the supply chain (visibility) (Blackhurst, et al.,
2005).
The broadest concept of uncertainty creates an area in which a specific threat is revealed. Risk arises when a
threat takes the probable form of a triggering event. This event, in turn, leads to a specific disturbance
resulting in a disruption in the supply chain. Thanks to this order, a more precise and realistic picture of the
impact of many malicious factors on the supply chain is created. However, one should be sceptical about the
effect of the description of the last, in a sense, final category, i.e. disturbance. It seems that reducing the effect

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

429
of all dangers in the area of the supply chain operation only to the delay effect is insufficient. After all, the
consequences of the events threatening the supply chain, especially on a global scale, are much richer. For this
reason, there is a need for a more general idea of an event influencing the flow in the supply chain- the logistic
barrier. A logistics barrier is an obstacle that hinders or prevents the flow or logistic process in the supply
chain in such a way that it directly or indirectly reduces the flow of goods, reducing the cross-section of the
logistics channel or the speed of the products flowing in it (Bentyn, 2022). Such an approach allows allocating
and analysing particular events affecting operations in a global supply chain with the most critical factor in
focus: the flow of merchandise, money and information.

GLOBAL SUPPLY CHAIN INFLUENCED BY THE PANDEMIC CONDITIONS

The global pandemic caused by the SARS-CoV-2 pathogen revealed particular vulnerabilities of the global
supply chain. Spreading a virus and new mutations may affect daily operations across the world, causing
many disruptions in the global supply chain and forming logistics barriers to the flow. The feature that
distinguishes logistic barriers and allows to classify them is the place of their occurrence in the supply chain.
Recognition of place of the appearance of a logistic barrier contributes to the determination of the entity
responsible for its occurrence. It enables the allocation of geographical, organisational and competence-
related activities that may neutralise its effects. In addition, the strength and nature of the impact it exerts on
the supply chain depend on the location of the logistic barrier (Fig.1). An internal problem with storage will
cause particular consequences, and in comparison, a different consequence will be caused by a political barrier
arising in the area of the continent or a communication blockade of a city or region endangered by the spread
of the Covid-19 virus, for example. The simplified diagram of the supply chain and its relations illustrate three
main areas of incidence for logistic barriers. The broadest area is the economic, political and social
environment where all supply chain operations occur. Logistic external barriers are created there. In the area
of relations between the links in the supply chain, inter-organisational barriers can be allocated, and within the
links of the supply chain, internal logistics barriers arise.
The quick spread of the virus from the most important manufacturing hub of the world China to three critical
distributive hubs in Europe, the USA and Japan, impacted global structures of the supply. A trigger event in
the form of covid restriction in the Wuhan area started to disrupt the manufacturing schedule of this vital area.
Epidemiologic blockade of logistic flows in the Hubei province, together with the essential city of Wuhan (11
million inhabitants), is an example of an external logistic barrier. The Chinese authorities, deciding to close
the region on January 23.2020, prevented the movement of goods and people in this important industrial
centre and logistics hub. It is worth noticing that this is where the OBOR (One Belt One Road) project begins,
aimed at accelerating logistics operations between Asia and Europe. On April 8, Wuhan Airport resumed
flight operations after 76 days of closure. Thus, it is a symbolic end of the external barrier, bringing
consequences in the form of delays in deliveries and the emergence of logistics costs and costs of restoring
product manufacturing processes. According to the development of a pandemic state, many internal logistics
barriers revealed themselves. Problems with connecting partners in supply chain may be allocated to the inter-
organisational logistics barriers. The disruptions in the operation of Chinese ports caused by the reduction in
the number of crews resulted in congestion and the necessity to move to other free ports or the postponement
of many transport operations. Cargo storage costs were rising, reducing the profitability of cooperation in the
supply chain. Logistics operator Maersk announced an additional congestion charge of $1,000 per container
for all refrigerated cargo towards Shanghai and Xingang. The Sea-Intelligence agency estimated at that time
the costs of adapting transport routes, taking into account the resulting logistic barriers, at $350 million
weekly.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

430
 Logistics barriers internal for the supply chain
Inside of a link in supply chain Between links on the supply chain
Internal logistic barriers Inter-organizational logistics barriers

Business, political and social environment

Logistic barriers external for the supply chain

Figure 1. Classification of logistics barriers depending on allocation in the supply chain.

Unfortunately, the nature of a virus affecting human labour caused disruption inside of manufacturing sites.
Such cases may qualify as an internal logistics barrier because they appeared inside a link in a given supply
chain. For example, the presence of the Covid-19 pathogen necessitates the protection of employees against
the effects of a disease. Stopping all operations in factories in the automotive industry is also associated with
the particular requirements that this type of activity places on partners in the supply chain. Just-in-time
operations help to minimise storage costs, but at the same time increasing the risk of factory downtime if
suppliers are unable to complete the scheduled delivery of parts.
An example is the activity of the automotive concern Fiat and Chrysler (FCA). Before the downtime, the
factory in Tychy produced an average of 1,040 cars per day. In such conditions, a three-week downtime
reduces the number of cars produced by over 21,000. This effect reduces plant efficiency and creates problems
for car dealers and customers further down the supply chain
All such events pose a significant test of the performance of global supply chains as seen from both a local
and a global perspective. The awareness of the sensitivity of the structures of international chains to the
occurrence of various types of logistic barriers should lead to a deeper understanding of them. Critical
vulnerability of the global supply chain lies among following factors:
1. Spatial distribution of operations. The impressive reach of a global supply chain poses a risk of delays
of logistics operations, forming a phenomenon described in the literature as the bullwhip effect. Operations
before pandemic carried in the just in time mode were after impact involuntarily transformed into just in case
scenarios. That fact increased the costs of logistics.
2. Manufacturing operation. Still, most of the operations require human labour. Such dependency is
particularly dangerous in pandemic conditions. Preventing from spreading the virus and at the same time
maintaining a manufacturing potential seem contradictory in factories relying on human labour. A reasonable
solution in the form of automatisation of operations may be difficult and time-consuming.
3. Logistics operation ̶ transport and warehousing. Similarly, transport operation, as well as
warehousing, require human work. Luckily in those areas, the advance in automatisation processes is
significant. Employing robots inside a warehouse may effectively increase operability during the pandemic.
On the other hand, transportation still needs technical solutions and law regulations to employ automated
transport modes.
4. Distribution. Supermarkets as a location potentially increasing virus spreading may be a target of
restrictions. No wonder customers turned towards e-distribution solutions. Since a relatively rapid increase in
customer interest in e-commerce, this area needs time to develop an efficient solution ready to serve a mass
market.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

431
5. Low restrictions. Lockdowns create prohibited areas for the movement of people and cargo. Such
incidents may become a new normal because of arising new mutations. Transport costs may increase, and
delays in the delivery time may influence scheduling and customer satisfaction level.
6. Capacity constraint. Shifting in demand among different channels of distribution may affect logistic
operations. A limited number of containers already caused problems in the ability to ship requested
merchandise. Forcing business owners to purchase and operate their own containers creates inefficient
solutions, further decreasing the effectiveness of global operations.
Above delimitation of exposure areas in the global supply chain may serve as a checklist for global operators
looking to resilience and maintaining a certain level of operations. Thanks to the independent organisations
researching global supply chains, it is possible to confront their reports with the experiences of existing
logistic operators.

RESEARCH REPORTS DELIVERING A DIAGNOSIS

With regard to the pandemic and its impact on supply chains, several reports (GEP, 2020; CISA, 2020; BCI,
2020; Interos, 2020; Gartner, 2020; PwC, 2020) have appeared confirming the occurrence of selected factors.
GEP conducted a survey among 400 executives in European and American global companies. A report by the
US Department of Homeland Security (DHS) in August 2020 conducted research on a sample of 50
companies in the area of information and communication technology, diverse in terms of size. At the request
of the Business Continuity Institute (BCI, 2020), research was conducted in which over 350 companies from
19 industries and 77 countries participated. Serving global customers in the supply chain, finance,
cybersecurity and compliance areas, Interos conducted a survey in October 2020 of 450 chief executives of
risk and compliance, logistics, IT, and procurement and operations executives at companies with revenues in
excess of $ 1 billion. These companies came from the United States in the following industries: aerospace,
defense, finance, processing, manufacturing, technology and energy. From February to March 2020, Gartner,
Inc. also collected data from 260 global supply chain leaders responsible for chain management in many
industries, including technology, industry and food. Both the Americas and Europe were the study areas. The
next report was based on a series of surveys prepared and distributed by PricewaterhouseCoopers (PwC).
Respondents are CFOs from the United States and Mexico. Four-fifths of them worked for Fortune 1000
companies.
All analyzed reports indicate that the pandemic has more or less disrupted supply chains. According to GEP
45% of companies experienced significant disruptions in their supply chains. According to DHS 53% reported
a moderate impact of the COVID-19 pandemic on their organisation's supply chains and 33% indicated minor
difficulties. There have been no companies that have experienced catastrophic consequences that would
paralyze their business. 60% of broadcasters, 56% of communication service providers, and 40% of IT service
providers have experienced moderate disruptions in their supply chain. However, each broadcaster faced at
least minimal disruptions from COVID-19. According to BCI the epidemic affected 73% of enterprises,
disrupting the chains on the supply side and 64% on the demand side. In the case of those polled who used
cargo transport by air for deliveries, 54.1% said that the pandemic had a significant detrimental impact on
their organizations. The International Air Cargo Association also reported that global cargo capacity decreased
by 35%. Conversely, a quarter of those interviewed experienced "serious disruptions" in their shipments by
sea. Ships were not allowed to enter ports and 120 of 126 countries imposed crew changes restrictions. Rail
transport has had the least impact on its operation. Directors surveyed by Interos also found that they
encountered delays in order fulfillment ̶ 52% of enterprises. Research by Gartner, Inc. found that only 21%
of enterprises had a non-disruptive supply chain. A resilient supply chain is understood as a chain monitored
on many levels of suppliers, which in the event of disruptions does not lose efficiency and can quickly react to
events such as Brexit, trade war or COVID-19. PwC research on the impact of the epidemic on the supply
chain in March 2020 found that 34% of executives thought that supply chain problems were the most
significant, and 30% considered making changes to their existing supply chains.
Obviously, these disruptions resulted in financial losses, 64% of enterprises reported 6% to 20% loss in
revenues in 2020 (GEP, 2020). The pandemic showed that 50.5% of companies did not have plans to take into
account the disruptions in their supply chains caused by it. However, the losses caused by it led 53.2% to plan
to create contingency plans in the future and 32.3% to adjust their current plans to avoid similar disruptions.
Responses to questions about supply chain transparency have shown that most supply chain incidents arise
from disruptions in the flow of information and goods at second and third-tier suppliers. During the pandemic,

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

432
these disruptions were magnified. Which contributed to high financial losses of enterprises. According to the
Interos research, the increase in costs also resulted in a significant increase in prices by suppliers (42%).
First of all, the root causes of the large-scale negative effects of the pandemic on supply chains were seen in
their globalization and sourcing in countries with low labor costs. The collected data shows the evolution of
the trend of relocating sourcing and manufacturing activities outside China. According to Gartner's research,
33% of enterprises have already done so or are planning to do so in the next three years. The COVID
pandemic was just another reason for this, as global supply chains were disrupted long before COVID-19
emerged. The 2018-19 trade war between China and the United States showed the weaknesses of global
supply chains between these countries. These activities, exacerbated by the pandemic, resulted in the
formation of the idea of regionalization of production, which allows for the strengthening of the company's
network. The desire to strengthen the company's chain is the second reason, right after the increase in
transport costs - caused by customs duties imposed by the US and Chinese governments in recent years, to
move production from China to closer locations, such as Mexico. The problem of customs duties is so serious
that the costs of chain service increased by up to 10% for over 40% of American companies.
Directors surveyed by Interos also indicated the need to look for new suppliers due to supplier downtime
caused by health restrictions (40%) and to move to suppliers from other geographic locations due to import
restrictions (38%). When asked about the vulnerable points of chains, 46% of respondents indicated
dependence on unreliable partners or governments, and 44% blamed the lack of adequate short-term and long-
term planning. 36% each indicated geographic concentration and a lack of understanding of interconnections
in global supply chains. Additionally, the majority (72%) of respondents believed that COVID-19 would be
one of the reasons to move production to the United States in the near future. At the same time, 78% believed
that the process of shifting production from the Far East to the United States would be continued in the longer
term.
Likewise, DHS respondents pointed to communication problems with their suppliers at the start of the
pandemic. Many Chinese suppliers have been locked in their homes for several weeks due to a lockdown. In
the case of companies that had their production lines in China, many of them, as in previous years, replenished
their stocks before the Chinese New Year, when some factories can stop up to 3 working days. If not for
advance orders, supply shortages would have been even more severe. However, the sources of components
such as semiconductors are in Asia, which poses a risk of geographic concentration. Parts incl. processors,
graphics cards and RAM memory are produced in the vast majority of five countries around the world (USA,
Japan, Taiwan, South Korea, China), which results in low chain resistance to local perturbations. In this case,
lockdowns introduced in the above countries resulted in plant downtime and significant global delays in
production and shipping. Changing a semiconductor manufacturer is not easy for IT service providers. Since
the level of customization of these components is very high, production requires appropriate technological
processes, and these differ between producers. The disruption of the supply and demand curve resulted in a
significant increase in prices.
BCI respondents also declared that they plan to limit supplies from eastern Asia ̶ 30% of enterprises. More
than two-thirds of manufacturers will try to use suppliers closer to their production plants as part of
strengthening the supply chain. The report also outlines the importance of having geographically diversified
suppliers and suggests that diversifying production may be a good way to minimize the risk of disruption.
More than half of the producers did not conduct research on their suppliers of level 2 and above. The lower
the supplier level, the less knowledge of supplier locations is. The events of 2020 meant that 60% of
enterprises intend to conduct a more detailed analysis of their chains. In addition, 45.3% of organizations
intend to locate all of their second tier vendors, compared with 36% who do now. During the outbreak of the
pandemic, many organizations faced delivery delays, first in Asia, then in Europe, and then in regions of
South America. The report emphasizes the fact that as many as 57.2% of enterprises want to diversify their
sources of supply, they will try to source goods locally. In March 2020, the institute's coronavirus
preparedness report was released, which showed that 16.3% of organizations had started sourcing some or all
of their goods more locally. Two months later, the figure had risen to 36.4%. At the same time, some
respondents speculated that the industry might move away from the JIT concept to produce products with
fewer product variants, which would allow for stocking and local production. 19.6% of the respondents
replied that after the pandemic they would maintain more inventory
Shippers and communication service providers (DHS, 2020) have listed inventory management as having the
greatest impact on supply chain disruptions. Many companies during the pandemic increased their inventory
levels to protect themselves against potential future disruptions related to the coronavirus, and in some cases,

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

433
e.g. in the sector of graphics cards that are used to mine cryptocurrencies, increased demand. The vast
majority of companies declared that they would increase their orders and inventories. This is linked to the
uncertainty surrounding the pandemic. Such an approach, however, exacerbates the current shortages in the
supply chain and may extend the lead times.
IT service providers have identified reasons for supply chain disruptions such as impacts on transport and
logistics or suppliers in different regions. Some companies order products based on a risk analysis. In the case
of high value products or long lead times, they try to choose different sources of supply. However, their
diversification is very difficult due to the aforementioned difficulties. In many cases, larger public firms had
better insight into their subordinate suppliers than private firms. Many IT companies may have hundreds of
downstream vendors for one product, and they often don't know who their tier 2, 3, or 4 vendor is.
In addition to these effects of the pandemic, approximately 38% of companies with revenues in excess of $ 1
billion have found "significant damage to their brand reputation" as a direct result of supply chain disruptions.
It also found that in the face of the pandemic, despite spending millions on ERP solutions, most global
companies are unable to effectively manage complex supply chains (GEP, 2020). During the pandemic, due to
the lack of maintaining stocks, using the just-in-time practice, because it is cheaper and does not require
having warehouse facilities, smaller companies were more susceptible to disruptions and interruptions in
supply chains.
The above-mentioned studies also indicated the potential possibilities of coping with the events that supply
chain managers had to deal with during the pandemic. Having a resilient supply chain comes at a higher cost.
Companies wishing to combine the advantages of JIT management, which emphasizes the delivery of required
materials in the required time and quantity, with Just-in-Case management, which emphasizes the planning
and preparation of various scenarios, must commit to risk management, so as to be able to assess the
willingness of the organization to take risks and compare this risk with other goals, e.g. cost effectiveness. The
respondents indicated regionalization of production as one of the methods of dealing with disruptions in
supply chains in the future. More than a quarter of respondents (Gartner, 2020) said they had already
regionalized or re-localized production to be closer to demand. As advantages of this practice, they emphasize
that regional supply chains can mitigate delays and shortages during periods of disruption. Some companies
are also planning hybrid methods, such as manufacturing in the Far East and moving only the final assembly
closer to the customer.
The results of a survey conducted in April (PwC, 2020) indicated that 56% of companies plan to develop
additional, alternative sourcing options for their supply chains. So you can see an increase of 26% in one
month. A quarter of 330 CFOs said supply chain strategies are critical to rebuilding or increasing corporate
revenue streams after the coronavirus pandemic.
Prior to COVID-19, 58% of respondents had made or were in the process of making major changes to
strengthen their supply chains, 20% planned such changes, and only 4% were not interested in making any
improvements to the chain. The pandemic situation forced 40% of companies to make further big changes,
42% made small adjustments, 18% of companies planned further changes. Two companies that did not plan
changes before the pandemic changed their position on this issue (Interos, 2020).

SELECTED STUDIES ON THE IMPACT OF THE COVID-19 PANDEMIC ON SUPPLY

CHAINS IN POLAND
Research on the impact of the pandemic on the operations of the company and supply chains was also carried
out in Poland by Polish scientists, the conclusions drawn from them should be compared with those coming
mainly from American sources. Marzantowicz, Nowicka & Jedliński (2020) undertook the description of
supply chain disruptions at the beginning of the spread of the COVID-19 pandemic. The survey was
conducted in March 2020 among Polish managers from the production, commercial and service sectors. The
most significant, in the opinion of managers, was the introduction of restrictions and a drop in the number of
orders. The reduced number of orders may be due to the closing of stores, which was one of the methods of
fighting the pandemic. The managers' attention was also drawn to the issue of prolonged shipments and
declining demand from European markets.
Enterprises encountered the following adversities through the pandemic:
– delivery delays ̶ 72%,
– failure to complete the shipment plan on time ̶ 63%,

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

434
– delays in the execution of orders due to the reduced number of available employees ̶ 54%,
– change of supplier required due to logistic problems at the supplier ̶ 45%,
– extended production time due to delays in the execution of orders with suppliers ̶ 45%,
– reduced access to transport outsourcing ̶ 45%,
– production stoppage for more than a week ̶ 27%.

The most popular methods of dealing with supply chain disruptions during a pandemic were increased
cooperation with suppliers - 82%. By exchanging information regularly, both sides were able to limit the
negative impact of the existing risk of chain disruptions. In addition, 45% of respondents used macroeconomic
analysis to predict the supply of a given product, so that they could order it in advance and create an
appropriate time buffer.
Managers' ideas for changes that will be required to create more resilient supply chains were to use
technologies that reduce uncertainty and the number of unpredictable factors ̶ 72%, change the structure of
supply chains to more hermetic ones ̶ 45%, renegotiate current contracts with suppliers, yes, in order to
jointly be able to react better to similar events in the future ̶ 45%, change the inventory management system,
e.g. increase inventory levels instead of the JIT approach ̶ 45%, increase the amount of technology use,
digitization of supply chains ̶ 45%, change the transport policy, e.g. . resignation from logistics outsourcing
for the benefit of own fleet ̶ 36%. Taking into account all the adversities that occur during a pandemic, the
most resistant to disruptions chains are those that can quickly adapt to the current situation, are reactive and
can quickly introduce internal instruments to increase their resilience, using digital technology to manage risk
and minimize the negative impact of disturbances on the chain. Research also confirms that the shorter the
chains, the easier it is to manage and control the risk in them.
Subsequent studies concerned the transport, forwarding and logistics industry. They are worth considering due
to the significant impact of traffic bans in certain periods of the pandemic and the serious impact of this
industry on flows in the supply chain. Publications showing the impact of the pandemic on supply chains,
logistics and transport are slowly appearing in the scientific world literature (Ivanov, 2020; Michail and
Melas, 2020; Paché, 2020; Pinto, 2020; Trautrims, 2020; Rejeb, A. et al. 2020; Marzantowicz et al., 2020;
Niewiadomski, 2020; Russell et al., 2020; Rojas-Rueda et al., 2021). In order to systematize and broaden the
knowledge in the presented research area, the results of own empirical research on Polish road carriers were
presented. Similar analyzes were carried out by Chinese researchers (Liu et al., 2020), who in the context of
the COVID-19 outbreak in China, they observed five impacts on China’s logistics industry, which are sharp
drop in logistics demand, shortage of transportation capacity, disruption of logistics network, change of
service mode, and increase in operating costs and number of loss making enterprise.
The subjects of the research were road transport companies operating on the foreign and Polish market, which
are registered in Poland and offer primarily transport services, then forwarding, storage and logistics services.
The respondents pointed to the drastic decrease in transport, which was the result of the decline in orders. In
total, over 80% of the respondents recorded a decrease in transport, including a significant decrease in over
50%.
This effect is a result of the economic condition deterioration of enterprises operating in the basic spheres of
the economy that are served by the transport sector. In April and May 2020, the decrease in orders on average
was: in accommodation and catering 63.15%, in retail trade 31.55%, in construction 26.45%, in industrial
processing 25.05% and in wholesale 25.15% . In the transport and warehousing industry alone, orders fell by
37.5%, which in fact were much greater in the transport industry. After decreases in the volume of transport
recorded at the turn of March and April, their return to the volume in March did not take place until
September 2020.
According to own research, in response to the question about which product groups were affected by the
decrease in transport - 29.55% of respondents indicated a decrease in transport for products from the
automotive industry. On the other hand, the increase in transport in the food industry was indicated by 15.9%
of the respondents, at the same time it was reported that there was no increase in any industry in 70.45% of the
respondents. It should be taken into account that the respondents' answers concern the situation at the turn of
May and June 2020, when alternatives to changing the structure of the transported articles were only searched,
at that time, first of all, declines in total transport were noted. The decrease in demand for deliveries in the
HORECA, automotive, aviation, heavy industry, household appliances, clothing and footwear industries was
supplemented by the growing demand for transport services in the electronics, packaging, pharmaceutical, e-

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

435
commerce industries, including e-grocery. Taking into account the fact that at the threshold of 2021, many
industries are reviving again, it can be expected that the changes in the transport structure will not deepen.
The respondents were also asked to specify the most significant negative impact of the Covid-19 pandemic on
transport activities. The respondents pointed to a general decline in revenues and profitability resulting from a
decline in demand for orders, a decline in the volume of transported loads in contractual and occasional
exchange relationships, and difficulties in obtaining return loads. In addition, the following have become a big
problem: prolonged payments over 60/90 days and more, long debt collection procedures due to a weak
judicial apparatus, unfair practices related to alleged bankruptcies, general stagnation caused by small
amounts of cargo. Among the answers, a problematic issue was also the reduction of employees’ wages in
order to limit layoffs, the inability to use the existing rolling stock potential, reluctance to perform tasks by the
older generation’s staff, uncertainty, and even fear. The described factors reinforce each other. For example,
as a result of too few loads, prices fall below the break-even point. Disturbances in the liquidity of enterprises
prevent the repayment of current receivables and undertaking investments aimed at increasing the
competitiveness of enterprises.

CONCLUSION AND DISCUSSIONS

Summing up, the presented theoretical analytical scheme is reflected in the quoted results of research on the
impact of the covid-19 pandemic on the supply chain and the barriers arising in it. The epidemic itself has a
social background and is a kind of external barrier. However, the pandemic affects both individual enterprises
and industries, e.g. transport, forwarding and logistics, which can be treated as internal barriers to the supply
chain. Disruptions in supply chains are reflected in a reduced level of demand for transport services. Strategies
in supply chains regarding their globalization or regionalization, the use of JITs and the amount of maintained
stocks also have a direct impact on the structure of goods transported and the scope of orders carried out for
individual industries. Therefore, when analysing the impact of the covid -19 pandemic on supply chains, it can
be considered from a broader perspective than just from the point of view of the created barriers. The broader
aspect of the impact of the coronavirus outbreak on the supply chain could include the following impacts:
price increases due to shortages of goods and semi-finished products for production, increased volumes of
orders for semi-finished products and components to create a buffer stock, extended customs clearance time,
lack of quarantined employees, drop in profits in many individual companies.

REFERENCES
Barroso, A. P., Machado, V. H. & Cruz Machado, V. (2008). A supply chain disturbances classification. Proceedings of
the International Conference on Industrial Engineering and Engineering Management: Volume 1–3 (pp.1870–1874).
Singapore.
BCI (2020, July 28). Covid-19: The Future of Supply Chain, Retrieved from: https://www.resilience360.dhl.com/wp-
content/uploads/2020/07/20200714_The-Future-of-Supply-Chain.pdf
Bentyn, Z. (2022). Types, importance and impact of logistic barriers on the functioning of the links of the international
supply chain. Poznań, Poland: Publishing House of the University of Economics and Business.
Blackhurst, J., Craighead, C., Elkins, D., & Handfield, R. (2005), An empirically derived agenda of critical research
issues for managing supply-chain disruptions. International Journal of Production Research, 43(19), 4067–4081.
Chopra, S., Sodhi M. (2004). Managing Risk To Avoid Supply-Chain Breakdown. MIT Sloan Management Review.
CISA (2020, July 28). Building a More Resilient ICT Supply Chain: Lessons Learned from the COVID-19 Pandemic.
Retrieved from https://www.cisa.gov/publication/ict-supply-chain-lessons-learned-covid-19.
Dixit, A., Pindyck, R. (1994). Investment under uncertainty. Princeton, NJ, Princeton University Press.
GEP (2021, June 22). The Post-Pandemic 2020 Outlook: Key Focus Areas for Procurement and Supply Chain. Retrieved
from https://www.gep.com/webcasts/the-post-pandemic-2020-outlook-key-focus-areas-for-procurement-and-supply-
chain.
Goldstein, J. (1994). The Unshackled Organization: Facing the Challenge of Unpredictability through Spontaneous
Reorganisation, CRC Press.
Handfield, R., McCormack, K. (2008). Supply Chain Risk Management. Minimising disruptions in global sourcing. New
York: Auerbach Publications Taylor & Francis Group.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

436
Heckmann, I., Comes, T., Nicke, S. (2015). A critical review on supply chain risk – Definition, measure and modelling.
Elsevier, Omega, 52, 119-132.
Hendricks, K., Singhal, V. (2003). The effect of supply chain glitches on shareholder wealth. Journal of Operational
Management, 21, 501–522.
Ivanov, D. (2020). Predicting the impacts of epidemic outbreaks on global supply chains: a simulation-based analysis on
the coronavirus outbreak (COVID-19/SARS-CoV-2) case. Transportation Research Part E, 136, 101922, 1-14.
Ivanov, D., Sokolov, B. (2010). Adaptive Supply Chains Management. London: Springer-Verlag.
Kaczmarek, T.T. (2008). Ryzyko i zarządzanie ryzykiem. Ujęcie interdyscyplinarne. Warszawa: Difin.
Kisielewski, P., Stanek, A. (2017). Klasyfikacja zagrożeń w łańcuchu dostaw. Wybrane metody analizy i oceny ryzyka.
Autobusy, 6.
Konecka, S. (2015). Ryzyko zakłóceń w zarządzaniu łańcuchami dostaw. PhD thesis, Poznań: Poznań University of
Economics and Business.
Liu, W., Liang, Y., Bao, X., Qin, J. & Lim, M.K. (2020). China's logistics development trends in the post COVID-19 era.
International Journal of Logistics Research and Applications.
Marzantowicz, Ł., Nowicka K. & Jedliński, M. (2020). Smart „Plan B” – in face with disruption of supply chains in
2020. Logforum 16 (4), 2.
Mason-Jones, R. & Towill, D. (1998). Shrinking the supply chain uncertainty circle, Control, The Institute of Operations
Management, 24(7), 17–22.
Michail, N.A., Melas, K.D. (2020). Shipping markets in turmoil: An analysis of the Covid-19 outbreak and its
implications. Transportation Research Interdisciplinary Perspectives, 7 (100178), 1-9.
Niewiadomski, P. (2020). Corporate maturity desiderata in the face of the COVID-19 pandemic – the digital plane of
logistics microfoundations. Logforum 16 (4), 3.
Paché, G. (2020). The „Day After” Covid-19 Pandemic: Logistical Disorders in Perspective, Review of European
Studies, 12(3), 1-8.
Pinto, C.A. R. S. (2020). Knowledge management as a support for supply chain logistics planning in pandemic cases.
Brazilian Journal of Operations & Production Management, 17(03).
PricewaterhouseCoopers (PwC). (2021, June 22) COVID-19 Pulse Survey, Retrieved from
https://www.pwc.com/us/en/library/covid-19/pwc-covid-19-cfo-pulse-survey.html
Rejeb, A., Rejeb, K. & Keogh, J. G. (2020). Covid-19 and the food chain? Impacts and future research trends. Logforum
16 (4), 1.
Rojas-Rueda, D., Morales-Zamora, E. (2021). Built Environment, Transport, and COVID-19: a Review, Current
Environmental Health Reports.
Russell, D., Ruamsook, K. & Roso, V. (2020). Managing supply chain uncertainty by building fexibility in container port
capacity: a logistics triad perspective and the COVID‑19 case. Maritime Economics & Logistics.
Trautrims, A., Schleper, M. C., Cakir, M.S. & Gold, S. (2020). Survival at the expense of the weakest? Managing
modern slavery risks in supply chains during COVID. Journal of Risk Research, 19 (pp.1466-4461).

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

437
 APPROACH OF A B2B DISTRIBUTION NETWORK OPTIMIZATION AS
VEHICLE ROUTING PROBLEM
SOLVING AS A VRP LIMIT TIME IN LAST CUSTOMER
M.E. David Camacho Fonseka1

Abstract— Currently, the supply chain continues to play a very important role in the economic environment,
customers are increasingly looking for greater accuracy in delivery times and customer experience plays a more
significant role in logistics each time, whoever has this greater focus will be able to dominate this market each time.
The vehicle routing problem is one of the most important problems in operations research, due to its applicability in
situations where the distribution of products to geographically dispersed customers is required using a fleet of
vehicles, from which it owes its name. This article addresses the solution of reaching customers on time with a time
constraint on the last customer delivery.

Keywords—Vehicle routing problem, Transportation management, Supply Chain Optimization.

INTRODUCTION
I.
The supply chain is defined as the set of activities, facilities and means of distribution necessary to carry out the
entire sales process of a product. This is, from the search for raw materials, their subsequent transformation and to the
manufacture, transport and delivery to the final consumer, this is how it becomes increasingly important for the
moments of purchase decision in the clients, so it also becomes a critical element for what is known today as improving
the customer experience.
So supply chain management is an integrative approach to manage material and information flows with suppliers
and customers, as well as between different functions within a company in an optimal way.
This is because any logistics network has as its main objectives:
 Minimize the total cost of the network (investment in configuration and cost network operating) and;
 Maximize customer satisfaction.

This has made corporations control their costs more than ever where challenges that companies face are the
difficulties when implementing solutions focused on the supply chain and that are affected by inadequate management
that leads to generating costs instead of optimizing them.
For this, the delivery process must be aligned with the corporate strategy and the objectives of the organization's
logistics system, so these types of decisions at a practical level are crucial to guarantee the success of companies [1].
The intention of this paper is to present a case presented in a company where the delivery on time is a key factor to
generate value in its customers and the design in the last mile is critical to consider the service level. How was
addressing the problem was as a vehicle routing problem (VRP) in this paper was developed a solution considering the
parameter as time limit in the last customer critical as part to get the service level, it was added this parameter as a
restriction in the classic CVRP.
The remainder of this paper is organized as follows. In section 2, we review some papers on the problem and VRP
concepts. In Section 3, we develop a formulation for the VRPLTLC. In Section 4, we run experimentation with data to
validate the solution. Finally, we provide some concluding remarks in Section 5.

1
David Camacho Fonseka,, Sr. Manager Customer Experince, Grainger, Mexico, [email protected]

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

438
 II. LITERATURE REVIEW

A. Classic Vehicle Routing Problem

The classic vehicle routing problem is one of the most popular problems in combinatorial optimization, and its
study has led to different techniques of general applicability such as exact and heuristic analyzes. A VRP consists of
determining a set of routes that begin and end in geographically dispersed distribution centers, so that vehicles visit a
set of customers, with a specific fleet of vehicles [2].
In figure 1 by Bertazzi and Grazia [3] shows a solution with two routes of a VRP instance where the distribution
center is at node 0. Again, we do not indicate the costs of the edges and assume that the routes in bold and dotted lines
are the only optimal solutions. This means that the single 0-1-2-3-4 route is not feasible, due to customer demands and
vehicle capacity. If possible, due to the triangular inequality, the cost of the edge (2,3) would not be greater than the
cost of the two edges (0,2) and (0,3). Routes 0-1-2-0 and 0-3-4-0 can be traveled by a single vehicle, one after the
other, or by two different vehicles. For example, suppose that all customers must be seen on the same day and that a
vehicle can drive 8 hours a day. If the time required to travel the two routes is less than 8 hours, a single vehicle can
travel both routes on the same day. On the other hand, if each route requires a long time (more than 4 hours) to be
traveled, two vehicles will be necessary.

Fig.1 A VRP solution (Bertazzi & Grazia, 2013)

B. VRP variants
Within the literature you can find different models of the VRP, additionally there are different types of problems
that are determined by the values that the parameters or variables that make up the problem can take [4]. The main
routing problems that relate to the case study are presented below:
Capacitated VRP (CVRP). Due you have a group of people geographically distributed in certain areas, separated
from the point of arrival at a known distance, collected by homogeneous vehicles with limited capacity [5].
VRP with Time Windows (VRPTW). The objective is to minimize the fleet, time and costs related to the trip. To
provide a better service, narrow time windows are managed, benefiting customers by reducing waiting times [6]. In
Restrepo et al. [7] establishes that each customer's service must start within an associated time window and the vehicle
must stop at the consumption center for instants of time.
Split Delivery VRP (SDVRP). It is a relaxation of the VRP where the same customer is allowed to be served by
different vehicles, if this reduces the total cost. In Rocha et al. [4], it is established that this condition is important if the
sizes of the customers' orders fill the capacity of the vehicle. This relaxation is very important if customer order sizes
are very large compared to vehicle capacity [8].
VRP with Stochastic Demands (VRPSD). It is a variation of the CVRP where each demand or requirement of the
clients is uncertain.
Dynamic Vehicle VRP (DVRP). In Toth and Vigo [9] they present this problem where the capacity restriction is
replaced by the maximum length (or time) restriction and the length of the arcs for each route cannot exceed the
maximum length of the route. If there is a restriction in both the vehicle's capacity and the maximum distance, the
problem is called DCVRP.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

439
C. VRP studies cases
Literature review present different authors which according the problem type present a solution.
Table 1 below present the following information:
Problem Authors Solution Comments
type
Milk
distribution
Tarantilis y
HFVRP Heuristic and the
Kiranoudis
construction
sector.
American
Alshamrani Red Cross
VRPPD Metaheuristic
et. Al blood
distribution.
Lubricating
oil waste
Repoussis
VRPPD Metaheuristic collection
et. Al
and
recycling.
Large scale
LSVRP Zeng et. Al Heuristic in real traffic
conditions.
Tobacco
VRPTW Hu & Ma Metaheuristic distribution
problem.
Gas station
refueling
F. Cornillier
VRPTW Metaheuristic problem with
et. Al
time
windows.
Safe
L. Zhang et. transportation
VRPTW Heuristic
Al of hazardous
materials.
Table 1. VRP study cases

III. MODELLING AND FORMULATING VRP LIMIT TIME IN THE LAST CUSTOMER FLEET DESIGN (VRPLTLC)

A. Formulation

This problem took as a base the classic capacitated VRP.

Below presented the following parameters:
𝐼 = Set of nodes or clients.
𝐾 = Set of vehicles.
𝐷𝑖 = Customer demand 𝑖.
𝑄𝐾 = Vehicle capacity 𝐾.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

440
UK = Binary variable ∈ {0,1}, it is equal to 1 when a vehicle is used.
𝑇𝑖𝑗 = Time to transit from node 𝑖 to node 𝑗.
ℎ𝑚𝑎𝑥= Maximum stipulated time limit for the route.
𝐶𝑖𝑗 = Cost of going from node 𝑖 to node 𝑗.
Also, we have the following variables:
X𝑖jk = Binary variable ∈ {0,1}, it is equal to 1 when it travels from node 𝑖 to node 𝑗 and 0 when it does not.
h𝐾 = Total travel time of the vehicle 𝐾.
yi = Artificial variables to break the subtour.

The formula states as follow:

Minimize:

(1)
With the restrictions:

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

441
 In equation 1 is presented the formulation minimize the cost adding a penalty when the vehicle exceeds the time
limit allowed for each vehicle.
In restriction 2 it imposes that for each j and each vehicle k, all the times a customer arrives is equal to all the times
he leaves with the same vehicle, known as the flow balance.
In restriction 3 it imposes that, for all j, that is, for each customer, you can only visit once with a single vehicle.
In restriction 4 it imposes that, for all the j, that is, for each customer, it only allows to leave once with a single
vehicle.
In restriction 5 it imposes that each vehicle can arrive at the warehouse only once.
In restriction 6 it imposes that each vehicle can leave the warehouse only once.
In the restriction 7 it imposes all the demand that I carry on the vehicle must not exceed its capacity.
In restriction 8 it counts the maximum time that vehicle k is passing to penalize it, that is, the maximum time of the
route is taken away from the time it takes for vehicle k to return to the warehouse, just to count how long it took in
reaching the last customer and removes the limit that was established is where it counts the hk to penalize it.
In restriction 9 it refers to the expression for the subtours elimination.
In restriction 10 and 11 refer to the expressions of valid inequalities to close the loop and quick search solution.

B. Heuristic Nearest Neighbor

Given the results that will be shown in the section, it was decided to establish a heuristic solution. The simplest
heuristics to solve a problem are known as greedy, they are algorithms that build a solution, the idea of these is simply
to find a solution quickly from a very simple reasoning.
It takes into consideration the closest node that has not been visited and is visitable, which if it meets the demand
capacity, becomes the new origin until it meets the capacities, returning to the deposit.
The nearest neighbor heuristic in the problem works as follows:
 Mark all clients 𝑖 as having not been visited.
 Select a client that has not visited.
 Visit to the closest client who has not been visited and who can be visited.
 Visible is defined when a node 𝑖, 𝑗 has been visited with a vehicle 𝐾 which has the vehicle load capacity and
time to serve it.
 Customer 𝑖 is marked as a new origin.
 The same is repeated looking for the closest customer 𝑖, 𝑗.
 When it no longer has load capacity or time then it returns to the deposit 𝑖 = 0.
 Use another vehicle 𝐾.
 It stops when there are no more customers to visit.

C. Heuristic Local Search

Heuristic methods can build a solution or start from an existing one and try to improve it, a method used for this
is the local search, in the local search change is made using certain criteria, if the method is improved, you can
decide to keep the solution or make another change.
Local search procedures are applied to improve an already obtained solution. In these procedures, a set of
neighboring solutions is defined and part of a primary solution is then replaced by a neighboring solution with lower
cost. The procedure is repeated until you cannot improve the solution.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

442
 The implementation (2-opt) takes 2 elements, a journey begins when the first element is chosen (i = 1) and
makes the exchange with the following once finished, the next element continues forward until finished.

IV. EXPERIMENTATION
3 stages were considered in order to evaluate the performance of the designed solution.
 Stage 1 reaching of the mathematical model.
 Stage 2 a comparison is made between the mathematical model with fleet design and heuristic.
 Stage 3 shows the results between the comparison of the solution proposal comparing it with the data obtained
from the study company.
The program use was Visual Studio 2019, C++ with CPLEX 12.10 as optimizer and Python 3.8 for the graphics.

A. Stage 1
For the validation of the mathematical model described in section 3, it was solved with real data from customer
locations with a variety of different numbers of customers (10, 20, 30, 40, 50, 60, 70 & 80).
Processing
Client GAP (%)
time (sec)
10 0 2
20 0 25
30 17 600
40 35 600
50 35 600
60 34 600
70 39 600
80 - -
Table 1. Stage 1 results
This stage showed that the model with more than 50 customer due computing time and the computer specifications
is not able to be used so was decided to design some heuristic.
In this experimentation it is shown that above 20 clients it is difficult to obtain optimal solutions, from 30 clients it
begins to have sub optimal solutions with a GAP greater than 17% and above 70 clients it does not even show us a
feasible solution. Therefore, the results are shown in the next stage of the heuristic model proposed in section 3.

B. Stage 2.1
A comparison is made between the mathematical model and the heuristic developed to see the behavior with the
same level of customers where a solution could be feasible with the mathematical model.

Mathematical model = 600 seg limit Heuristic

Client Z (hr) GAP Ave. Std Dev. seg Z (hr) seg
10 5.80 0.0 0.0 1 5.89 0.02
20 11.35 0.0 0.0 18 15.28 0.03
30 15.45 17% 14% 600 18.24 0.07
40 18.47 34.48% 15.37% 600 19.28 0.09
50 15.99 16.43% 10.78% 600 16.89 0.04

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

443
 60 24.49 33.53% 6.41% 600 23.84 0.06
70 26.62 38.80% 14.56% 600 25.94 0.04
80 x x ND x 25.87 0.08
Table 2.1. Stage 2 results
All the results shown complied with the time restriction in the last client. It is insisted that this restriction was
complied with because of the importance in the level of service of the study carried out.
As mentioned in the previous stage above 70 customers the exact model can no longer process with the equipment
where the tests were performed is where it does not mark results.
The heuristic presented to have fast processing times, less than 1 minute, as well as to present solutions approximate
to that of the exact model.

C. Stage 2.2
At this stage, the heuristic model processes larger instances in a shorter time, in this experimentation shows the
average obtained with a higher number of customers (80, 85, 90 and 95) with real company data.
Table 2.2 shows the data obtained where the first column represents the number of customers, the second column
the number of vehicles used, the column Z(hr) ave. is the average of the solutions reached by the heuristic model in the
different instances, the processing time column (seconds) on average of the time it took for the model to be processed.

Heuristic
Client K
Z (hr) Ave. seg
80 2 25.093 0.061
85 3 33.081 0.050
90 3 34.871 0.052
95 3 36.153 0.056
Table 2.2. Heuristic with real data from company

The results show the solutions reached for the instances above 80 customers with 3 vehicles and comply with the
restriction of the delivery schedule of 10 hours and be within the service level.
Graphical features made in Python 3.8 to illustrate behavior in the following figure.

Table 2.3. Heuristic with 80 customers

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

444
D. Stage 3
At this stage, data from the company was taken in a day of deliveries to compare the results that may have been
obtained by using the methodology proposal against the data that was obtained from the company directly.
The following results are presented in tables 3 and 4 where the first column represents the scenarios of the heuristic
model and the current one of the study company, the second column represents the number of customers, the third
column the number of vehicles used, column Z(hr) the solution reached by the model heuristic, the cost column
represents the percentage of improvement based on the cost that paid to supply customers, the column % service level
represents the percentage calculated based on deliveries supplied in time (1 day).

Client K Z (hr) % Cost improvement Service level

80 3 26.33 25% 100%
80 ND ND ND 63%
Table 3.1. Compare actual with heuristic model 80 customers

Client K Z (hr) % Cost improvement Service level

85 3 37.2 22% 100%
85 ND ND ND 81%
Table 3.2. Compare actual with heuristic model 85 customers

The data that is observed in the table we can see that the heuristic model may meet the level of service expected by
the company in addition to having a less use of vehicles compared to the company's results.

V. CONCLUSIONS
In this work, it was possible to design a mathematical model that considers the level of service, which is the time
limit to ensure deliveries within the hours that the company offers to customers.
It was concluded that a mathematical model due to its computational times to propose it in the study company is
complicated since the optimal waiting time to process routing models is 15 min., From here it derived the development
of a heuristic of those studied in the literature.
Within the literature, different authors were studied where the most widely used heuristics are those of the 2 Opt
local search, to use this type of heuristics, a method must first be used to develop an initial solution so that based on
this solution the local search.
When comparing both solutions, there were only decimal points of difference between one and the other, with the
local search taking the shortest time.
Among the results obtained with results from the company's clients (80-95 clients), optimal solutions are obtained
with 3 vehicles.
When comparing the results of the heuristic solution and the results of the company was found that the solution
reaches 100% of the expected service level with a lower use of vehicles for the cases shown in the tables 3.1 and 3.2.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

445
 REFERENCES

[1] Hollmann, H. Supply Chain network design under uncertainty and risk. Submitted for the Degree of Doctor of
Philosophy. 2011.

[2] Barán, B., & Hermosilla, A. Comparación de un Sistema de Colonia de Hormigas y una Estrategia Evolutiva para
el Problema del Ruteo de Vehículos con Ventanas de Tiempo en un Contexto Multiobjetivo. San Lorenzo,
Paraguay: Centro Nacional de Computación, Universidad Nacional de Asunción. 2001.

[3] Bertazzi, L. y M. Grazia, Inventory routing problems with multiple customers. EURO Journal on Transportation
and Logistics · August 2013.

[4] Rocha, L.; González, C. y Orjuela, J. (2011). Una revisión al estado del arte del problema de ruteo de vehículos:
Evolución histórica y métodos de solución. En: Ingeniería, Vol. 16, No. 2, pág. 35 - 55 .

[5] Arias, J. Aplicación de un modelo de optimización en la planeación de rutas de los buses escolares del colegio
Liceo de Cervantes Norte. Pontificia Universidad Javeriana, Facultad de Ingeniería, Departamento de procesos
productivos. Bogotá. 2010.

[6] Jean-Francois Cordeau, Michel Gendreau, Gilbert Laporte, Jean-Yves Potvin y Frédéric Semet, “A guide to
vehicle routing heuristics”, The Journal of the Operational Research Society, Vol. 53, No. 5, 2002, pp 512- 522.

[7] Jorge Hernán Restrepo, Pedro Daniel Medina y Eduardo Arturo Cruz, “Un problema logístico de programación
de vehículos con ventanas de tiempo”, Scientia et Technica – Universidad Tecnológica de Pereira, Vol. 14, No
39, 2008.

[8] M. Dror and G. Laporte and P. Trudeau, ”Vehicle Routing with Split Deliveries”, Discrete Applied Mathematics
50, 239-254 (1994).

[9] Paolo Toth y Daniele Vigo, “The Vehicle Routing Problem”. Society of Industrial and Applied Mathematics
(SIAM) monographs on discretemathematics and applications, Philadelphia, USA, 2002, pp 1-23, 109-149.

© 19th International Logistics and Supply Chain Congress

October 21-22, 2021, Gaziantep, Turkey

446
 ARTIFICIAL INTELLIGENCE AND SUSTAINABLE SUPPLY CHAIN
FINANCE IN OMNICHANNEL LOGISTICS
Monika Bal1, Kinga Pawlicka2

Abstract ⎯ Enterprises striving to maximize the effectiveness of the supply chain operation should also
strive to improve its sustainability. However, very often a big barrier for them is the observance of
sustainable development practices simultaneously ensuring better financial and environmental results.
One of the main assumptions of Sustainable Supply Chain Finances (SSCF) is the optimization and
strengthening of business processes while minimizing the negative impact on the environment. In
recent years, the importance of modern solutions consistent with the idea of sustainable development
and using Artificial Intelligence (AI) in improving the efficiency of supply chain management has also
increased. Hence, two goals were adopted. The first one, of a theoretical nature, consists in
determining the possibilities of supporting the implementation and development of SSCF with the use
of AI technology. The second, of a practical nature, concerns the presentation of ways to improve the
supply chain sustainability achieved with the use of modern supply chain management solutions - the
implementation of SSCF with the help of AI on the example of the company from the clothing industry
using omnichannel.

Keywords ⎯ artificial intelligence, omnichannel logistics, supply chain finance, sustainable supply chain

INTRODUCTION
Sustainable Supply Chain Finances
The aspects of supply chain management that have particularly attracted the attention of both researchers
and business in recent years (according to authors research based on Scopus database) are Supply Chain Finance
(SCF) and Sustainable Supply Chain Management (SSCM). SCF is most often defined as solutions implemented
with the help of financial institutions or technologies, the ultimate goal of which is to adapt financial flows to
physical and information flows in the supply chain (Hofmann, 2003; Camerinelli, 2009; Lamoureux and Evans,
2011; Wuttke, Blome, and Henke, 2013). M. Steeman (2014) defines SCF as a financial arrangement applied
by at least two supply chain partners and supported by a chain leader to improve overall financial performance
and reduce the overall level of supply chain risk. The concept of SSCM is defined as supply chain management
in a way that integrates the goals and requirements of sustainable development defined by the company,
suppliers, customers, and external stakeholders (e.g., consumers, decision makers, associations). Among these
goals, there are distinguished economic, social, environmental, and ethical that all members of the supply chain
must achieve in order for the supply chain to be sustainable (Fritz, 2019, p. 2).
The assumptions and goals of SCF and SSCM are linked by the concept of Sustainable Supply Chain
Finances (SSCF). The essence and main challenge of the SSCF is to solve the problem of the lack or difficult
access of the supply chain members, who are not its leaders, to financial resources for financing investments in
sustainability of the products (Karaosman et al. 2017). Usually it is connected with difficulty of that companies
to access the capital and related high costs of capital, while often small suppliers from the beginning of the
supply chain are the links in the supply chain where sustainability issues need improvement and would bring
the greatest benefits (de Boer et al., 2017). SSCF is therefore the implementation of SCF in financing products
improving the level of companies’ sustainability.
Due to the asymmetric distribution of access to capital and the sustainability level of individual links in the
supply chain, the implementation of sustainable supply chain finance solutions is particularly important for the
links at the very beginning of the supply chain, although the entire supply chain achieves long-term benefits
(Fig. 1). An important aspect presented in the illustration is the distinction between “direct” and “extended”

1Poznań University of Economics and Business, Institute of International Business and Economics, Department of Logistics, Poznań,
Poland, [email protected]
2 Poznań University of Economics and Business, Institute of International Business and Economics, Department of Logistics, Poznań,

Poland, [email protected]

447
supply chain, because tools facilitating access to capital for the implementation of innovative working capital
solutions are offered by chain leaders to the links forming a direct chain, while the lack of such assistance for
indirect suppliers causes the so-called SME financing "missing middle" (Milder, 2008, Alibhai, Bell, and
Conner, 2017).

Figure 1. The Supply Chain Network’s Asymmetrical Distribution of Sustainability Risks and Access to
Finance
Source: De Boer et al., 2017, p. 2.

Use Of Artifıcial Intelligence in Supply Chain Management

Artificial intelligence (AI) is one of the tools and technologies used to manage the supply chain, which can
contribute to achieving significant benefits for the development of enterprise innovation. Artificial intelligence
is understood as a field dealing with the creation of machines and algorithms, the functioning of which is based
on the characteristics of human intelligence, including the ability to adapt independently to changing conditions,
abstract understanding and continuous learning (Wyskwarski 2015, Stawiarska 2016). AI can function as
hardware and / or software in a system representing human intelligence, this type of representation is already
visible in all business processes and operations (Vipul 2009, Min 2010). Artificial intelligence influences not
only the development and practical implementation of new solutions, but also the development of knowledge
and skills of employees through their participation in training, which is necessary to perform official tasks and
duties (Olan et al. 2021).
Supply chain management activities, including those related with finance, require interaction between
employees and machines. The manner of interaction varies depending on the complexity of data and the degree
of human capital use. The effect of this cooperation led to the development of cobots, i.e. robots used to interact
with humans (El Zaatari, Marei, Li, Usman 2019). This results in blurring the boundaries where repetitive
activities were assigned to machines, while people were engaged in what requires creative thinking, anticipation
and sensitivity. This trend is in line with the Industry 5.0 concept (Özdemir & Hekim, 2018).
Artificial intelligence theories focus on the creation and use of machine learning that can support business
operations and the supply chain (Fan et al. 2020, Radhakrishnan et al. 2018, Li et al. 2017). In the case of the
supply chain, the use of AI will be of particular importance in areas related to the use and analysis of data, i.e.
the forecasting process and analytics. Artificial Intelligence provides predictive tools that enable demand and
inventory management to improve the decision-making process (Huang, Fu 2019).

448
 New machine learning techniques can be used to manage inventory, which allow for the detection of
incorrect data (incorrect data search), automation of seasonality detection, or assigning parameters for inventory
to better understand changes taking place in the supply chain and coordinate, for example, sales at the logistics
center level and / or store. Inventory management is a determinant of the proper functioning of the supply chain,
as too little inventory can lead to loss of customers and reduced sales, while excess inventory requires more
money, more transport or more work. As a consequence, the coordination and control of inventories becomes
one of the most important criteria of the SSCM, for which the company's inventory records must be kept and
updated in an optimal way. Moreover, most of the raw materials and semi-finished products, especially for the
clothing industry, are imported, therefore strengthening the SSCM translates into an improvement in the supply
of raw materials. The purchase of materials and raw materials is influenced by currency fluctuations, interest
rates, inflation, trade regulations, or a crisis (political, environmental), as well as changes in customer demand,
technological progress, bargaining power of key suppliers as well as changes in delivery times and the quality
of raw materials. Therefore, in order to maintain a competitive position, enterprises should develop an effective
management procedure that includes the control of purchases, storage and transportation of raw materials, as
well as the development of an ecologically sustainable process and reduction of purchasing costs (Tseng et al.
2018).

METHODOLOGY, APPLICATION AND COMPARISON

Implementation of AI based on the example of a clothing company
Determinants of AI such as human intelligence, data analysis, prediction and optimization are essential for
designing sustainable supply chain finance (Patnaik 2015; Zahraee, Khalaji Assadi, and Saidur 2016).
According to G. Yu, F. Li and Y. Yang (2017), the main problem of supply chain finances is the issue of linking
financial processes with material flows. The solution is AI, which strengthens partnership and information
exchange between customers and suppliers in the supply chain by sharing available financial resources
(Venkatesh et al. 2019). AI brings together FDIs and supplier networks for financial services sharing, especially
given the supply chain finance criteria and regulations of financial institutions and brokers (Olan et al. 2021).
Full availability of information and financial data in individual links of the chains ensures supervision and
optimization of transactions. Enterprises operating under such assumptions achieve the necessary balance, equal
access to resources and unified measures of performance monitoring, which contributes to generating benefits
for all network participants.
Based on financial data for individual links in the supply chain, there are generated conclusions that improve
the quality of services and customer service. The aim of the enterprises is to develop extended analytics covering
the handling of current data and control of the analytical capabilities of companies operating within the SC, as
well as the implementation of the decision-making process for a dynamic and unpredictable environment.
Accelerating decision cycles with AI for supply chain finance operations enables real-time scheduling using
unstructured data. This results in a significant shortening of decision cycles, which synchronizes the planning
and execution of tasks.
As part of the second objective, there was undertaken a theoretical description of the implementation of AI,
which will support the implementation of SSCF in the clothing company with a global reach founded in Poland.
The study concerns the company that operates under the omnichannel strategy integrating traditional and online
sales and currently does not apply the SSCF guidelines. AI is used to support different business processes, not
directly connected with SSCF. After a detailed case study of the enterprise, guidelines for the implementation
of the SSCF (based on AI) were proposed and recommendations were formulated in the form of potential
benefits and threats resulting from the implementation of the project. Data used for the case study comes from
the company's website as well as press releases and reports.
The case company has a multi-brand strategy and its offer includes brands targeted at different groups in
terms of wealth, age and lifestyle. Enterprise is operating over 30 years and currently employing more than
20,000 people in its offices, logistics centers and sales networks globally. At the end of 2020, the company's
offer was available in a total of 38 countries, including 30 on-line, and had network of over 1,800 stores with a
total area of over 1 million m2. Company focuses on social responsibility and other principles of sustainable
development, addressing these issues in its strategy and declared values. Issues such as environmentally friendly
production (proper water management and energy acquisition, reduced plastic), chemical safety of production,
elimination of packaging that is not subject to closed circulation and ecological solutions in the buildings of the

449
headquarters and sales network are given special care. The facilities of the analyzed company, including stores,
have BREEAM and LEED certificates. However, these areas do not directly concern the supply chain finances.
The case company is listed on the Warsaw Stock Exchange as part of the WIG20 index and belongs to the
MSCI Poland index. Over 50% of sales revenues is generated in Poland. During the SARS-CoV-2 pandemic,
the company recorded a three-digit increase in revenues from online stores, which, however, was not able to
fully compensate for lost sales opportunities in stationary stores. Multiple lockdowns announced both in Poland
and abroad resulted in a decrease in overall sales by approx. 15% compared to the previous year. Closing
stationary stores, reducing the volume of orders with suppliers, transferring some collections to subsequent
seasons and unfavorable exchange rates on prices in selected quarters resulted in company achieving a lower
gross margin by 1.5 pp. compared to the same period of the previous year. The effects of the pandemic also
affected operating costs, which decreased by 2.3% y / y in 2020/21. This decrease was the result of lower rent
costs obtained thanks to exemption from the obligation to pay rents in shopping malls during both lockdowns
pursuant to a resolution of the Polish government. As a result, in the financial year 2020/21, company recorded
a net loss of PLN 190 million compared to the net profit achieved a year earlier in the amount of PLN 486
million.
In order to implement AI in the clothing company, there should be analyzed certain activities, which were
formulated in the form of four points constituting the principles for the implementation of this project.
First, there is need to develop a business case for AI implementation. The company’s board should collect
and present the reasons why AI is being implemented, such as value for money, the scope of work and the need
to undertake the project at any given time. Preparing a valuable business case and including it in the company's
strategy will increase the chances of obtaining maximum benefits from the implementation of AI and avoid
problems with overestimating the return on investment or omitting the potential risk.
Second, a comprehensive database must be built. It is impossible to use artificial intelligence for any area
without access to appropriate information sets. Therefore, the company should standardize and catalog its own
data, including archived data. It is also worth checking if and how they can obtain further data sets that can
bring a competitive advantage, e.g. information on consumer moods or the location of various events in real
time (Rewolucja AI, 2017). In many companies, synchronizing data from various systems (in this case - WMS,
a system cooperating with an RFID code reader, goods allocation system) is a big challenge, for which good
solution is Big Data Analytics (BDA). BDA is a set of tools, algorithms, simulations and optimization that can
be applied to Big Data in order to analyze them and extract unknown, hidden, important and useful relations,
patterns and information (Adams, 2010). BDA can improve the supply chain sustainability, used in areas such
as supporting modern sustainable production, designing sustainable ship routes and schedules, or assessing
environmental performance (Chalmeta & Barqueros-Muñoz, 2021).
Third, it is recommended to assess resources and potential implementation models. Appropriate knowledge
and experience are necessary in the implementation of each project, which is why company under examination
should consider using expert services, e.g. by outsourcing or the experience of consulting companies. Integration
within the enterprise is just as important as acquiring the right know-how on the market. AI solutions most often
relate to seemingly independent departments of an organization, such as warehousing, distribution, finance and
accounting, or customer service. However, they should be treated holistically. The choice of the project
management model is also a very important aspect. For AI and the apparel industry, Agile seems to be the most
appropriate approach. It divides projects into smaller tasks and stages. Dividing the work into smaller parts
gives agile teams the opportunity to analyze the initial results, re-evaluate the work done and adopt an iterative
approach at each stage of the project (Serrador, Pinto 2015).
Fourth, all processes should be integrated with particular emphasis on the financial aspect. After considering
the conclusions from the pilot AI implementation, it is worth synchronizing them with the company's core
business processes. Doing so will allow companies to fully achieve the benefits assumed in the business case.
Integration should be holistic and usually requires process changes - their automation or thorough redesign in
cooperation with people responsible for a given area (Rewolucja AI, 2017). In the case of AI, the most important
thing is to ensure clear standards of human-machine and / or robot cooperation. The most difficult changes in
the organization concern human capital and not innovative technological solutions. Therefore, the company
should prepare employees for the implementation of AI through training and full access to information about
the stage of project work.
After the correct implementation of AI, it is possible to achieve a number of benefits, manifested in gaining
a sustainable competitive advantage as a result of effective forecasting of demand and reducing the costs of
operations in the supply chain (Someh et al. 2020). On this basis, it can be concluded that the use of AI for the

450
implementation of the omnichannel strategy is particularly desirable, as it allows to predict trends, optimize
logistics operations for two distribution channels, synchronize prices and personalize promotions. Some of them
anticipate the needs of customers by sending them the goods without waiting for the order confirmation. The
potential benefits of AI implementation for a clothing industry company include:
• Increase in innovation, particularly in robotics and automation dedicated mainly to warehouse
processes.
• Achieving a higher margin.
• Better forecasting of customer needs.
• Increase in sales.
• More effective procurement management and cost control.
Robotization and automation, guaranteed by AI, contribute to the return on productivity, which can be used
by analyzed company. A good benchmark is Amazon's adaptation of KIVA robots (based on artificial
intelligence) for sorting goods in logistics centers. The use of this type of solutions increased the efficiency of
the daily service level from 700 thousand up to 1.5 million products. As a result, an employee can handle 300
products intended for shipment within an hour, which is three times more than the original with an accuracy of
99.99%.
Research shows (Syam, Sharma 2018; Kumar et al. 2019, Gao 2020) that companies that base their
development strategy on AI achieve much higher margins than companies that do not make such investments.
Process automation (especially production and storage) for goods from the fast fashion market, where products
are often delivered to stores, ensures increased efficiency, which is a necessary condition to increase the margin.
It is important to emphasize that company operating on such market will obtain tangible benefits only if they
decide to implement it significantly. If any of the key elements are removed (solid foundations in the form of
high level of digitization, full-scale implementation of AI or strategic direction of AI) - the profits will become
much lower.
AI can help company accurately forecast customer needs in real time. The analyzed company could better
coordinate omnichannel processes by using regularities and dependencies identified through the analysis of
large data sets. AI draws information from various sectors of the enterprise, such as: databases of past
transactions with suppliers, social media trends, purchasing patterns, history of visited websites, facial
expression analysis and weather forecasts. The machine learning algorithm in the case of the clothing industry
in predicting the level of sales should be based on the costs of waste returns (related to reverse logistics), lost
sales and the size of orders adjusted to the demand.
According to data from the McKinsey Global Institute (2020), for online commerce, personalization
combined with dynamic pricing can lead to an increase in sales by up to 30%. The learning systems that could
be proposed by the company should focus on personalizing the promotion, adjusting it to customers, season or
trends. One of the simpler solutions in this regard are beacons, sending offers to customers' smartphones and at
the same time tracking their consumer choices.
The use of AI allows company to optimize the costs of the global supply chain and flexibly respond to
changes in the environment. Company from the case study can perfectly use this in the event of another
(hypothetical) closure of stores in shopping malls due to a pandemic or other unpredictable situation and
efficiently switch to online sales, without unnecessary inventory or having to move the collection over time (as
was the case with lockdown in 2020). Digitally integrating manufacturers' systems with suppliers using AI will
provide full information on the availability of production capacity, the availability of materials for production,
and will also let company to optimize the availability of transport, as well as extraordinary events and failures.
The listed benefits for the clothing industry company resulting from the implementation of AI can be used
to improve all dimensions of sustainability (social, environmental and economic), examples of which are shown
in Table 1.

Table 1. Implications for the supply chain sustainability resulting from the implementation of AI
Benefit from AI Implications for supply chain sustainability
implementation
Increase in warehouse The use of solutions that save resources - energy, fuel, water, etc.
processes innovation Improving the efficiency and productivity, quality of service as well as working
conditions, increasing safety, e.g. by reducing the risk of accidents

451
 Achieving a higher The excess margin can be used for investments in the social and environmental
margin and increase in dimension of sustainability
sales Improving profitability, economies of scale benefits through greater sales
Better forecasting of Better matching of the assortment to the demand will reduce the problem of
customer needs overproduction in the clothing industry, resulting in dead inventory harmful to
the environment and constituting an unnecessary cost. Reducing returns reduces
the cost, including the environmental cost, of implementing and handling them
More effective Reduction of costs resulting from disruptions in the supply chain, e.g.
procurement maintaining larger stocks, organizing additional transport
management and cost Improving the stability of employment
control
Source: Own elaboration.
Implementation of Sustainable Supply Chain Finances for a clothing company
An additional justification for the implementation of AI for company from clothing industry is the support
for the implementation of SCF. When implementing SCF, companies typically use analytical models to help
determine the order in which a program is introduced to vendors. These models are typically a transformation
of the Peter Kraljic matrix where supplier segmentation is based on factors such as relative strategic importance
for the buyer and risk, such as supply chain disruption in the event of supplier cessation. The second popular
segmentation model focuses on the potential value that SCF can bring to a supplier and buyer base together.
The analyzed criteria are the difference in the credit ratings of individual suppliers and the buyer, the suppliers'
cost of capital rate, and total expenses for all suppliers with a specific credit rating (de Boer, van Bergen,
Steeman, 2015, p. 33; Chakuu, Masi, Godsell, 2017, pp. 753).
The implementation of sustainable supply chain finance, however, has some specific features that should
guide the analyzed company. According to de Boer et al. (2017), in this case, both the capabilities and required
financial resources of analyzed suppliers should be additionally specified. For this purpose, company’s board
should divide the solutions for its suppliers into "rewarding", addressed to suppliers with a better rating and
greater financial resources, and "supporting" addressed to prospective suppliers with limited possibilities of
independent implementation of solutions consistent with the idea of sustainability, for reasons such as low rating
and limited resources. A supporting SCF solution may be, for example, fixed assets financing or collaborative
innovations, while an example of a rewarding solution is selective or performance based reverse factoring.
The model proposed by R. de Boer et al. (2017) (Figure 3) combines the previously mentioned popular
vendor portfolio analysis methods supplemented with criteria such as risk of sustainable investment, and ease
of access to capital analyzed to determine for which supplier to apply reward mechanisms, support mechanisms,
where to replace suppliers, and where other solutions than SSCF may be more appropriate.

452
 Figure 2. Implementation of SSCF solutions in various suppliers of the supply chain
Source: De Boer et al, 2017, p.4.

CONCLUSION
The advantage of using the model by company from clothing industry is a more precise adjustment of
the expected benefits from the implementation of the SSCF to the specific needs of suppliers, resulting from the
position in the chain, rating, ownership and ease of obtaining various types of resources. Therefore, it would be
necessary to collect a large amount of sensitive information about suppliers and their complex analysis. The use
of integrated systems, such as WMS or SAP, as in the example company, is helpful in achieving this. However,
the difficulty related to the implementation of the model in company’s supply chain will be the need to obtain
data from cells such as cotton plantations and processing plants, chemical plants or factories belonging to the
extended and ultimate supply chain and located in developing countries. In order for this process to be effective,
it should therefore be supported by IT solutions that facilitate the collection and analysis of Big Data. Another
challenge is qualitative and sensitive nature of needed data, which may require use of artificial intelligence.

SOURCES
Adams, M.N. (2010). Perspectives on data mining. Int. J. Mark. Res. 52, 11–19.
Chalmeta, R.; Barqueros-Muñoz, J.-E. Using Big Data for Sustainability in Supply Chain Management. Sustainability
2021, 13, 7004. https://doi.org/ 10.3390/su13137004
El Zaatari, S., Marei, M., Li, W., & Usman, Z. (2019). Cobot programming for collaborative industrial tasks: An overview.
Robotics and Autonomous Systems, 116, 162-180.
Fan, J., Fang, L., Wu, J., Guo, Y., & Dai, Q. (2020). From brain science to artificial intelligence. Engineering, 6(3), 248-
252, doi:10.1016/j.eng.2019.11.012.
Gao, Z. (2020). The application of artificial intelligence in stock investment. In Journal of Physics: Conference Series (Vol.
1453, No. 1, p. 012069). IOP Publishing.
Huang, Y., & Fu, J. (2019). Review on application of artificial intelligence in civil engineering. Computer Modeling in
Engineering & Sciences, 121(3), 845-875.
Olan, F., Liu, S., Suklan, J., Jayawickrama, U., & Arakpogun, E. O. (2021). The role of Artificial Intelligence networks in
sustainable supply chain finance for food and drink industry. International Journal of Production Research, 1-16.
Özdemir, V., & Hekim, N. (2018). Birth of industry 5.0: Making sense of big data with artificial intelligence,“the internet
of things” and next-generation technology policy. Omics: a journal of integrative biology, 22(1), 65-76.

453
Kumar, V., Rajan, B., Venkatesan, R., & Lecinski, J. (2019). Understanding the role of artificial intelligence in personalized
engagement marketing. California Management Review, 61(4), 135-155.
Min, H. (2010). Artificial intelligence in supply chain management: theory and applications. International Journal of
Logistics: Research and Applications, 13(1), 13-39.
Patnaik, D. (2015). Theorizing change in artificial intelligence: inductivising philosophy from economic cognition
processes. AI & SOCIETY, 30(2), 173-181, doi.org/10.1007/s00146-013-0524-5.
Radhakrishnan, A., David, D. J., Sridharan, S. V., & Davis, J. S. (2018). Re-examining supply chain integration: a resource
dependency theory perspective. International Journal of Logistics Systems and Management, 30(1), 1-30.
Li, R., Dong, Q., Jin, C., & Kang, R. (2017). A new resilience measure for supply chain networks. Sustainability, 9(1),
144, doi:10.3390/su9010144.
Yu, G., Li, F., & Yang, Y. (2017). Robust supply chain networks design and ambiguous risk preferences. International
Journal of Production Research, 55(4), 1168-1182, doi.org/10.1080/00207543.2016.1232499.
Venkatesh, V. G., Zhang, A., Deakins, E., Luthra, S., & Mangla, S. (2019). A fuzzy AHP-TOPSIS approach to supply
partner selection in continuous aid humanitarian supply chains. Annals of Operations Research, 283(1), 1517-1550.
Vipul, J., (2009). Editorial Note for the Special Issue on ‘Artificial Intelligence Techniques for Supply Chain Management’.
Engineering Applications of Artificial Intelligence, 22 (6): 829–831. doi:10.1016/j.engappai.2009.01.009.
Stawiarska, E. (2016). Logistyczne systemy informatyczne wykorzystujące sztuczną inteligencję w branży
motoryzacyjnej. Organizacja i Zarządzanie: kwartalnik naukowy.
Someh, I., Wixom, B., & Zutavern, A. (2020). Overcoming organizational obstacles to artificial intelligence value creation:
propositions for research. In Proceedings of the 53rd Hawaii International Conference on System Sciences, doi:
10.24251/HICSS.2020.712.
Serrador, P., Pinto, J. K. (2015). Does Agile work?—A quantitative analysis of agile project success. International journal
of project management, 33(5), doi.org/10.1016/j.ijproman.2015.01.006.
Syam, N., & Sharma, A. (2018). Waiting for a sales renaissance in the fourth industrial revolution: Machine learning and
artificial intelligence in sales research and practice. Industrial marketing management, 69, 135-146.
Rewolucja AI. Jak sztuczna inteligencja zmieni biznes w Polsce (2017). Raport McKinsey & Company, Forbes Polska.
Tseng, M. L., Wu, K. J., Hu, J., & Wang, C. H. (2018). Decision-making model for sustainable supply chain finance under
uncertainties. International Journal of Production Economics, 205, 30-36.
Wyskwarski, M. (2015). Metody sztucznej inteligencji w organizacji inteligentnej. Zeszyty Naukowe. Organizacja i
Zarządzanie/Politechnika Śląska.
Zahraee, S. M., Assadi, M. K., & Saidur, R. (2016). Application of artificial intelligence methods for hybrid energy system
optimization. Renewable and Sustainable Energy Reviews, 66, 617-630, doi.org/10.1016/j.rser.2016.08.028.
Steeman, M. (2014). The Power of Supply Chain Finance - How companies can apply collaborative finance models in their
supply chain to mitigate risks and reduce costs. Zwolle (NL): Windesheim University of Applied Sciences.
Karaosman, H., Brun, A. and Morales-Alonso, G. (2017). “Vogue or Vague: Sustainability Performance Appraisal in
Luxury Fashion Supply Chains,” in Sustainable Management of Luxury, Springer Singapore, pp. 301–330.
De Boer, R. & Bals, L. & Tate, W. & Gelsomino, L. & Steeman, M. & Bals, C. (2017). Exploring the Financial Flows in
Sustainable Supply Chains. The International Conference on Production Research (ICPR2017) vol 5
Milder, B. (2008). “Closing the Gap: Reaching the Missing Middle and Rural Poor through Value Chain Finance,” Enterp.
Dev. Microfinance, vol. 19, no. 4, pp. 301–316.
Alibhai, S., Bell, S. and Conner, G. (2017). “What’s Happening in the Missing Middle? Lessons From Financing SMEs,”
Washington D.C.
Fritz, M.M.C. (2019). Sustainable Supply Chain Management. In book: Responsible Consumption and Production.
Encyclopedia of the UN Sustainable Development Goals. Springer, Cham DOI: 10.1007/978-3-319-71062-4_21-1.

454
 THE EFFECT OF DIGITALIZATION LEVEL AND SMART LOGISTICS
PRACTISES ON LOGISTICS PERFORMANCE

Asst.Prof. Gökhan AKANDERE1

Abstract ⎯ The focus of the article is to investigate the impact of the effective implementation of smart
applications on the logistics performance of Turkish logistics enterprises, with the effect of digitalization
level/maturity. First, the literature on "Digitalization Level/Maturity" (DL), "Smart Logistics Practises" (SLP)
and "Logistics Performance" (LP) was reviewed. In order to determine the regression weights and
significance of the structures, factor analysis was performed on the data collected from 385 participants.
Adoption patterns of smart logistics applications were analyzed using the structural equation modeling (SEM)
approach. As a result of the research, it was determined that the increase in the DL level supports the SLP
application of the enterprises, the increase in the DL level indirectly causes an increase in the LP and the
increase in the SLP application level will also increase the LP. DL is positively associated with both SLP and
LP. It is also positively associated with SLP and LP. The results show that businesses can improve LP by
running SLP activities and raising DL levels. As a result of testing the hypotheses, it was found that DL affects
SLP, SLP affects LP, DL affects LP via SLP.

Keywords ⎯ Digitalization Level/Maturity, Smart Logistics Practises, Logistics Performance, SEM

1
Selcuk University, Vocational School of Social Sciences, Department of Management and Organization, [email protected]

455
 INTRODUCTION
The concepts of Industry 4.0 and logistics 4.0 were developed to help German companies improve their
performance by reducing their costs. With the current digitalization trend in the spirit of Industry 4.0, logistics
businesses require the transformation of their processes with logistics 4.0 roots in order to gain sustainable
development and market share. This is one of the aims of the fourth industrial revolution, which accelerates
the development of the digitalization capability of enterprises. The goal of Logistics 4.0 is not to involve
people in their work, but to avoid mistakes and have faster processes where information can be shared
effortlessly and in real time. There will always be a need for the involvement of people who control processes
and take control of any system failures.
In this context, the article focuses on the effect of the application level of smart logistics activities on
logistics performance with the effect of digitalization. First, the literature on "Digitalization Level/Maturity",
"Smart Logistics" and "Logistics Performance" was reviewed. Factor analysis was performed on the collected
data to determine the regression weights and significance of the structures. Adoption patterns of smart
logistics applications were analyzed using the structural equation modeling (SEM) approach. SEM evaluates
the level/maturity of digitalization and the observable and unobservable nature of smart logistics activities'
applications through empirical analysis. The focus of the study is to investigate the effect of the effective
implementation of smart applications on the logistics performance of Turkish logistics enterprises, with the
effect of digitalization level/maturity. The remainder of the article is organized as follows: a literature search
on digitalisation, smart logistics applications is presented. Then, the proposed model and hypothesis are
explained. Next, the research methodology and experimental results are presented. It summarizes the analysis
and discussion, followed by the conclusion.

LITERATURE
Digitalization Level/Maturity
Digital Transformation has emerged as an important research topic for academics and businesses. It can
be defined as the ability to transform existing products or services into digital variants and thus offer
advantages over the tangible product (Parviainen et al. 2017). Digital Transformation is a structured capability
that guides a company through its digitalization journey and provides improvement in operational processes
and providing value to customers as a result of the integration of digital technologies and new business models
in all areas with different cultures, processes, structures and strategies (Ismail et al. 2017).
Stoldt et al. (2018) states that businesses can implement two strategies to digitize their processes, such as
gradually transforming their processes and production spaces, or implementing a radical change by replacing
all processes and systems with fully digitized ones. Digital transformation can potentially create new
possibilities for businesses and foster new avenues of development and growth and lead to new understanding
in the value chain. (Parviainen et al. 2017; Ross et al. 2017). A high level of digitalization has a positive
impact on the competitiveness and productivity of businesses. Table 1 shows the studies on the level and
maturity of digitalization in the literature.
In the research, five corporate digitalization levels as Poor, Basic, Medium, Good and Excellent were used
to determine the level of digitalization in operational processes of enterprises (Bickauske et al. 2020);
•Weak/None; The business does not apply or need to apply digital technologies in its operational
processes.
•Basic level; The business uses basic digitalization applications and tools such as internet, mobile phone,
e-document in its operational processes.
•Intermediate level; The business automates various processes with information communication
technology tools.
•Good level; The business fulfills the requirements of Industry 4.0 in most of its operational processes by
transforming its various processes with digitalization.
•Great level; The business uses digitalization tools as software and hardware at a high level in its
horizontal and vertical processes.
Table 1. Digitalization/Maturity Level Literature Review
Publication Subject Aim/Objective Method Findings
Romero- Level of The power of digitalization as an Survey In the Spanish hotel industry, a high
Martínez digitalization in the anti-corruption tool for the hotel level of digitalization has been found to

456
 and García- hotel industry industry is examined. reduce the likelihood of corrupt
Muiña, practices.
(2021)
Albukhitan, Developing a digital A new transformation strategy A guideline for developing a digital
Literature
(2020) transformation that includes a different aspect of transformation for the manufacturing
strategy for the strategy adapted to the nature Review sector was proposed.
manufacturing of the manufacturing sector.
Brozzi et al., Fundamental Developing a fundamental As a result of the analysis of data
Survey
(2021) readiness indicators readiness indicator (KRI) for obtained from 65 manufacturing
for assessing the evaluating businesses and companies, it has been determined that
digital level of researching the overall digital digital tools used in operational
manufacturing level of companies in specific processes are effective because they can
SMEs areas. provide intuitive elements.
Schuh et al., Evaluation of the It aims to enable companies to The rapid assessment model (CCMS)
Model
(2021) level of find the most suitable Industry 4.0 and Industry 4.0 Maturity Index were
digitalization of maturity model for them. Proposal analyzed.
enterprises in the
production sector
Bickauske et Analysis of the level Determining the level of adoption As a result of the research, it has been
Survey
al., (2020) of digitalization of of digitalization and industry 4.0 determined that the difficulties such as
enterprises in the practices of enterprises the lack of information of the
Lithuanian employees, the lack of investment and
production sector the slow rate of return and the lack of
qualified employees negatively affect
the digitalization level of the enterprises.
Billon et al., A multivariate It aims to identify and explain the Variables that explain the level of
Survey
(2010) analysis to identify differences in the adoption of digitalization in 142 countries have been
differences in global information and communication identified.
levels of technologies (ICT) for 142
digitalization developed and developing
countries.
Gierlich et Digitalization To investigate how the roles of As a result of the research, it has been
Qualitative
al., (2019) approaches of SMEs SMEs in ecosystems and determined that the roles of SMEs in
in platform digitalization approaches are Research ecosystems determine the framework of
ecosystems related with a qualitative research digitalization approaches.
approach

Smart Logistics and Logistics Performance

In the last decade, the use and evolution of Information and Communication Technologies (ICT) in
industry has become inevitable mainly because it is vital for increasing organizational efficiency and
competitiveness (Heuser et al., 2008). This has spurred the adoption of ICT in logistics and manufacturing
operations as a result of increased use of the internet providing real-time communication between machines
and humans in industry operations. The digitalization paradigm led to the emergence of a new concept called
smart logistics or logistics 4.0 (Issaoui et al., 2019). Logistics 4.0 is seen as the result and next step of Industry
4.0.
By incorporating a range of activities as part of the supply chain process, such as logistics, demand
forecasting, sales planning, supply requirements, inventory management and product distribution, efficient
and reverse flow of services, goods and related information between a starting point and a destination
according to customers it is defined as cost-effective planning, implementation and control.
McFarlane et al. (2016), smart logistics is to plan and control operational processes with smart tools and
methods. Smart logistics is a logistics system that can increase flexibility, adapt to market changes, and enable
the company to be closer to customer needs (Bigliardi et al., 2020). Uckelmann (2008) defines smart logistics
as the fulfillment of monitoring and control processes using technology to obtain information about material
flow. Smart Logistics provides material and information flows with the tools and smart methods that are the
requirements of Industry 4.0 solutions, the traceability of the products and the identification of its elements,
the detection of the problem, the automatic execution of the selection and solution, the use of cloud-supported
network for information flow, the realization of product and facility locations. It can be defined as planning,
organizing and controlling the steps such as evaluating with real-time big data analysis and creating the most
appropriate route for material and product transportation (Oleśków-Szłapka et al., 2019).
The main areas of applications of an efficient and powerful Logistics 4.0 are as follows (Barreto et al.,
2017; Yin et al., 2018);
• Enterprise Resource Planning (ERP) used in supply chain management,
• Warehouse Management Systems (WMS) used for inventory management,

457
 • Transportation Management Systems (TMS) used for the management of logistics activities in supply
chain management,
• Intelligent Transportation Systems (ITS), which includes new technologies, infrastructures, control
methods and operations in the field of transportation;
• Information Security (IS) used to protect information and infrastructures.
Also in the near future, it will increase the efficiency of management's decision-making, such as the
availability of real-time and sequential information collected through sensor networks, drone points, and
business intelligence systems, and make businesses more flexible and efficient (Barreto et al., 2017). The
benefits of Smart Logistics for businesses are as follows (Bechtsis et al., 2017; Kauf, 2019);
• It eliminates the constant control obligation of employees for products and services.
• It provides transparency for all processes.
• It allows businesses to communicate and interact with their environment.
• It facilitates the delivery of state-of-the-art products and services.
• It ensures that material handling systems are integrated into logistics processes and used appropriately.
• It enables the digitalization of billing, payment or licensing services.
• It increases efficiency and sustainability.
Although the literature has clearly shown the potential and future applications of digitalization, smart
logistics processes still need significant support regarding the transformation, the application of techniques
and technology, as well as the development of relevant business models. Extensive research and development
efforts aim to increase the ability to monitor, monitor and predict the progress of ongoing logistics processes.

Table 2. Smart Logistics and Logistics Performance Literature Review

Aim/Objective Method Findings
Publication Subject
As a result, the models developed
represent reality and can be applied
A simulation-based decision
in different intralogistics
Coelho et al., Simulation-based decision support tool is proposed for the Model
environments, acting as digital
(2021) support for intralogistics right production and distribution Proposal
matching tools to improve
processes for logistics 4.0.
operations without interruption of
real clusters.
Explores how to think and design
an intelligent Intelligent Product- The potential of smart PSS in PI
Pan et al., Smart product-service systems Model
Service System (PSS) for and the need for research have
(2019) in logistics Proposal
service-oriented, intelligent been identified.
interoperable logistics
As a result of the research, the
Jabeur et al., Leveraging Intelligent Proposing a multi-agent-based necessary steps for the joint
Logistics Collaboration with a solution to support collaboration Framework operation of the objects were
(2017) Multi-Agent System-Based between tools and equipment Proposal determined and the risk
Solution used in smart logistics processes management processes were
supported.
Clusters of
Application of Blockchain Explaining and exemplifying the As a result of the application,
Issaoui et al., blockchain
technology in smart logistics applications of Blockchain in classification was made according
(2019) application
processes Smart Logistics to four clusters.
To determine the principles of
Gregor et al., Model The term Smart Connected
Smart Connected Logistics Smart Connected Logistics
(2017) Proposal Logistics has been clarified.
System applications
Evaluation of the level of
Kostrzewski Suggesting a model for the
innovation for selected Model A current applicable model has
et al., (2019) explanation of the term Logistics
solutions regarding logistics Proposal been created.
4.0
4.0
Oleśków- Evaluation of the Logistics 4.0 Grey Systems According to the results of the
Szłapka et Logistics 4.0 Maturity Levels maturity model and maturity research, no firm has reached the
al., (2019) Evaluated According to GDM levels through the GRI decision Theory 4th and 5th level of maturity
model (Management, Integration).
Various benefits, including
Providing a frame of reference
transparency, building trust, and
for blockchain-based green
Korczak and improving collaboration and
Smart Logistics in the logistics to achieve sustainable Model
Kijewska collaboration – Incentive
Development of Smart Cities logistics operations with the Proposal
., (2019) mechanisms, data storage and
integration of the Internet of
transmission as well as
Things and big data.
implementation cost challenges
Identifying the opportunities that It has been determined that
Cimini et al., Exploring human factors in
Industry 4.0 applications in Case Study different and different applications
(2019) logistics 4.0
Logistics can provide for of businesses using the same

458
 businesses technology lead to similar results.
To identify logistics innovation It has been determined that there is
The impact on logistics
Wang et al., capability and its effects on a negative relationship between
innovation capability and Survey-SEM
(2020) supply chain risk in Australian logistics innovation capability and
supply chain risks
cargo firms supply chain risks.
Douaioui and Examining the three industrial
Interaction between Industry Literature It was emphasized that there is no
Mabroukki revolutions that underpin
4.0 and smart logistics Review single truth behind "Industry 4.0".
(2018) Industry 4.0
A conceptual framework was
Applicability of Lean and
Literature review on how lean developed that improves the
Edirisuriya et Green Concepts in Logistics Literature
and green practices can improve performance of logistics operations
al., (2018) 4.0 A Systematic Literature Review
operational performance by applying lean and green
Review
concepts.
A multi-agent system based on A model has been presented that
cloud computing platform is Multi-agent determines the estimation of the
A cloud computing
Benotmane proposed to measure the approach and economic, environmental and
applications in green logistics
et al., (2018) economic, environmental and simulation social costs of logistics activities
service
social performances of the model and the impact of a new strategic
logistics process. solution.

RESEARCH METHODOLOGY
Studies on digitalization level/maturity, smart logistics practices and logistics performance were evaluated
through literature review and synthesis. In addition, in the article, empirical analysis was used to understand
the effect of digitalization level/maturity of enterprises on the logistics performance of smart logistics
applications. Structural equation modeling (SEM) evaluates the unobservable and observable nature of
digitalization level/maturity and smart logistics applications. The proposed work only considers linear
relationships via SEM. However, skewness, kurtosis coefficients, PP-Plot plots and Anderson Darling
normality test were used for the normality of the data. However, the homogeneity of the variances was
checked with the Levene homogeneity test. Data were evaluated with descriptive statistics, Explanatory Factor
Analysis (EFA), Confirmatory Factor Analysis (DFA), and Structural Equation Modeling analyzes. These
analyzes will be carried out using SPSS 21 and AMOS 21 software.
The first part of the questionnaire used in the research consists of descriptive questions, the second part
consists of Bickauske et al., (2020) and Sternad et al., (2018) 4 items to determine the level of digitalization,
the third part consists of Cyplik et al., (2019) and Oleśków-Szłapka et al., (2019) smart logistics practises
were composed of 8 items, and in the fourth part, questionnaire questions were formed by adapting Wang
(2019), Facchini et al., (2020) and Zelbst et al., (2010) performance 8 items.

CONCEPTUAL FRAMEWORK AND PROPOSED HYPOTHESIS

The conceptual model proposed as a result of the evaluation of the studies in the literature is shown in
Figure 1. The model consists of 3 factors: Level of Digitalization (DL), Smart Logistics Applications (SLP)
and Logistics Performance (LP). A pilot study was conducted on twenty logistics companies to validate the
Survey/Group. A five-point Likert scale was used to collect 70 responses. After the pilot study, the 22 sub-
factors that make up the 3 main factors were reduced to 14. Table 3 shows the factor, sub-factors and
hypotheses.

H3 (+)

H1 (+) H2 (+)
Digitalization Smart Logistics Logistics
Level Practices Performance

Figure 1. Conceptual Model ("+" Indicates Positive Hypotheses)

Table 3. Constructs and Hypotheses

459
 No Factor Sub-factors (Items) Reference
H1 Digitalization Level DL1: A basic level of digitalization Sternad et al.,
Level of Digitalization (DL) Affects processes are implemented in our business. (2018);
Smart Logistics Applications (SLP) DL2: Horizontal digitalization processes are Bickauske et
applied in our business. al., (2020)
DL3: Vertical digitalization processes are
implemented in our business.
DL4: In our enterprise, integrated
digitalization processes are applied at
horizontal and vertical levels.
H2 Smart Logistics Practises (SLP) SLP1: Electronic document flow is used in Oleśków-
Smart Logistics Applications Impact our business. Szłapka et al.,
Logistics Performance (LP) SLP2: Automatic placement and collection (2019); Cyplik
systems are used in warehouse processes. et al., (2019)
SLP3: Automated systems and tools are
used in internal transport processes.
SLP4: Warehouse Management System is
used in our business.
SLP5: Transportation processes are planned
automatically (Route optimization, delivery
planning etc.,)
H3 Logistics Performance (LP) LP1: Transportation costs decreased. Wang (2019);
Level of Digitalization (DL) indirectly LP2: Storage costs decreased. Facchini et al.,
affects Logistics Performance (LP) LP3: Transfer costs decreased. (2020); Zelbst
LP4: Inventory costs decreased. et al., (2010)
LP5: Packaging costs reduced.

SAMPLE CHARACTERISTICS
In the research, the questionnaire (questions) was applied to logistics managers and experts in SC
operations. For the survey, Turkish logistics enterprises that have adopted or are in the process of adopting
smart logistics practices in the digitalization process, and the managers and experts of these enterprises
were chosen as the sample of the research. Necessary feedback was received from 385 managers and
experts. Data was collected online over a 6-month period via the Google forms application.
In the research, providing access to logistics companies and managers during the Covid 19 pandemic
was the biggest challenge encountered. However, as a result of the online survey application, it was seen
that 385 of the 400 participants reached were valid. Evaluating demographic data, businesses with more
than 250 employees had the largest proportion of responses (285, 76%). In terms of working capital, the
highest response came from firms with more than $20 million in capital (295, 76.6%). The highest return
on experience came from respondents with more than 16 years of experience (130, 33.8%) and the highest
return on education came from respondents with a bachelor's degree (220, 57.10%).
Table 4. Sample Characteristics
Categories Frequency Percentage (%)
Gender Male 290 75.30
Female 95 24.70
Total 385 100%
Educational qualification High school 65 16.9
Associate degree 40 10.4
Bachelor's degree 220 57.1
Postgraduate 60 15.6
Total 385 100%
Age 21–30 90 23.4
31–40 140 36.4
41–50 115 29.9
51–60 40 10.4
> 60 0 0

460
Total 385 100%
Total years of experience < 5 years 110 28.6
6–9 years 45 11.7
10–15 years 100 26
> 16 years 130 33.8
Total 385 100%
No. of employees in organisation 20–49 40 10.4
50–249 60 15.4
More than 250 285 76
Total 385 100%
Organisation capital (in million US dollars) <5 30 7.8
5–10 45 11.7
11–20 15 3.9
More than 20 295 76.6
Total 385 100%

RESULTS
EFA
In this study, logistics performances of logistics companies within the scope of digitalization level were
measured using a questionnaire-based method. A questionnaire based on a five-point Likert scale was used
to understand the effect of the three variables in the study. As a result of the online survey application, 385
valid responses were received from managers and experts in the supply chain and logistics processes. The
Bartlett and Kaiser-Meyer-Olkin (KMO) test was performed to test the suitability of the data. The results
show that the KMO value is 0.822 and it is satisfactory as it is greater than 0.5. The Bartlett test was also
determined to have a p value of <0.001 for the study. Varimax (with Kaiser Normalization) rotation
method was used in the research. The returned component matrix is shown in Table 5. Variables with
factor loadings less than 0.5 as a result of EFA were excluded from the analysis. It was determined that the
highest loading was 'SLP2' with 1.063 and the lowest loading was 'SLP5' with 0.598. No cross loading was
observed. Therefore, the results of EFA are sufficient for DFA analysis.
Table 5. Rotated Component Matrix
Cronbachs Eigenval Percent Cumulative
Items Items & Component Loading AVE
Alpha ues VA Percent VA
Digitalization Level/Maturity 3,91
In our business, basic digitalization 0.638 3,98 0,917 3,844 18,070 18,070
processes are applied.
Vertical digitalization processes are 0.927 3.88
Level/Maturity
Digitalization

applied in our business.

Horizontal digitalization processes are 0.820 3.96
applied in our business.
In our enterprise, integrated 1.016 3,84
digitalization processes are applied at
horizontal and vertical levels. (End to
End)
Smart Logistics Practises 3,63
Electronic document flow is used in 0.641 3,53 0,843 3,811 23,486 41.556
Smart Logistics Practises

our business.
Automatic placement and collection 1.063 3,53
systems are used in warehouse
processes.
Automated systems and vehicles are 0.759 3,48
used in internal transport processes.
Warehouse Management System is 0.620 4,01
used in our business.
Transportation processes are planned 0.598 3.45

461
 automatically (Route optimization,
delivery planning etc.,)
Logistics Performance 2,23
Transport costs have decreased. 0.937 2,29 0,965 4,769 31,401 72,957
Performance

Storage costs are reduced. 0.943 2,19

Logistics

Transfer costs are reduced. 0.962 2,19

Inventory costs have decreased. 0.899 2,31
Packaging costs are reduced. 0.872 2,17
KMO = 0,822; Bartlett Sphericity Test =5201,047, p<0,001
Digitalization Level, Smart Logistics Practices questions are rated on a 1–5 scale.Here the indicators are “Strongly Disagree” = 1;
“Disagree” = 2; “Neutral” = 3; “Agree” = 4; “Strongly Agree” = 5.
Logistics Performance questions are rated on a 1–5 scale. Here the indicators are “Not at all” = 1; “Some” = 2; “Some Degree” =
3; “Significantly” = 4; “Very Significantly” = 5.

Correlations
Table 6 gives details of correlation analysis, which gives significant correlations at 95% (2-tailed) and
99% (2-tailed).
Table 6. Correlations Table
DL SLP LP
DL 1
SLP 0.445** 1
LP 0.202** 0.322** 1
*Correlation is significant at the 0.05 level (2-tailed). **Correlation is significant at the 0.01 level (2-tailed).

CFA
CFA was carried out on the Level of Digitalization, Smart Logistics Applications and Logistics
Performance structure. All of these 3 structures were allowed to freely interact with each other. The structure
consists of five items on "Logistics Performance", five items on "Smart Logistics Applications", and four
items on "Digitalization Level". CFA results are shown in Table 6. As a result of the analysis, it can be
concluded that the goodness statistics for the collected data set have acceptable values. CFA model estimates
are shown in Table 7. Loads between factors and measured variables are greater than 0.6. Therefore,
acceptable level of convergent validity is indicated by loading the indicators in different structures (Barki &
Hartwick, 2001). The measurement model loads are mostly greater than 0.6, except for a few loadings as
below. This indicates acceptable convergent validity. The CFA path diagram is shown in Figure 2, drawn
using AMOS software. The final version of the scales used in the research, digitalization level/maturity 4
items, smart logistics applications 5 and logistics performance 5 items were evaluated in the model.
Tablo 6. CFA Fit Index
χ2/df GFI RMSEA CFI TLI IFI RMR
Standard <3 >0.90 <0.08 >0.90 >0.90 >0.90 <0.08
Model 2.920 0.942 0,064 0,921 0,915 0,942 0,041
χ2: Ki kare (CMIN) df: Degrees of Freedom; RMSEA: Root Mean Square Error of Approximation; CFI: Comparative Fit Index;
GFI: Goodness of Fit Index; RMR: Root Mean Square Residual; IFI: Incremental Fit lndex (Bagozzi ve Yi, 1988)

Tablo 7. Estimates for CFA Model

Item Construct Standardized Unstandardized Standard Critical P-Value
Estimate Estimate Error(S.E.) Ratio(C.R.)
DL1 ← DL 0.667 1,000 ***
DL2 ← DL 0.839 1.202 0.068 17.736 ***
DL3 ← DL 0.922 1.223 0.075 16.354 ***
DL4 ← DL 0.967 1.398 0.083 16.762 ***
LP1 ← LP 0.911 1.000
LP2 ← LP 0.980 0.989 0.026 38.385
LP3 ← LP 0.959 1.022 0.029 35.579 ***

462
 LP4 ← LP 0.895 0.998 0.035 28.694 ***
LP5 ← LP 0.843 0.824 0.033 24.725 ***
SL12 ← SLP 0.362 1.000
SL11 ← SLP 0.723 1.932 0.303 6.376 ***
SL10 ← SLP 0.520 1.404 0.171 8.219 ***
SL9 ← SLP 0.809 2.285 0.303 7.531 ***
SL8 ← SLP 0.859 2.615 0.402 6.514 ***

Figure 2. CFA Path Diagram

SEM
SEM was carried out in two stages: i) verification of hidden structures and ii) judgment of tting model
based on structural model. SEM examines non-causal and causal relationships between variables. Indicates
the corresponding priority when the bidirectional arrows in the CFA model should be replaced with single
headed arrows. The developed path diagram is shown in Figure 3. SEM results are shown in Table 8. As a
result of the analysis, it can be concluded that the goodness of t statistics for the collected data set has
acceptable values. Estimates for the structural model are shown in Table 9.
Table 8. Model Fit Index
χ2/df GFI RMSEA CFI TLI IFI RMR
Standard <3 >0.90 <0.08 >0.90 >0.90 >0.90 <0.08
Model 2.669 0.945 0.052 0.922 0.925 0.944 0.040
χ2: Ki kare (CMIN) df: Degrees of Freedom; RMSEA: Root Mean Square Error of Approximation; CFI: Comparative Fit Index;
GFI: Goodness of Fit Index; RMR: Root Mean Square Residual; IFI: Incremental Fit lndex (Bagozzi ve Yi, 1988)

Table 9. Estimates for The Structural Model

463
 Item Construct Standardized Unstandardized Standard Critical P-Value
Estimate Estimate Error (S.E.) Ratio(C.R.)
SLP ← DL 0.408 0.258 0.052 4.972 ***
LP ← SLP 0.345 1.071 0,227 4.719 ***
DL1 ← DL 0.677 1,000 ***
DL2 ← DL 0.839 1.202 0.068 17.731 ***
DL3 ← DL 0.922 1.224 0.075 16.348 ***
DL4 ← DL 0.967 1.399 0.083 16.754 ***
LP1 ← LP 0.911 1.000
LP2 ← LP 0.980 0.989 0.026 38.368
LP3 ← LP 0.959 1.022 0.029 35.575 ***
LP4 ← LP 0.895 0.998 0.035 28.685 ***
LP5 ← LP 0.843 0.824 0.033 24.715 ***
SL12 ← SLP 0.363 1.000 ***
SL11 ← SLP 0.722 1.924 0.301 6.403 ***
SL10 ← SLP 0.520 1.400 0.170 8.256
SL9 ← SLP 0.807 2.272 0.300 7.565 ***
SL8 ← SLP 0.861 2.612 0.399 6.545 ***
*** Significant at p < 0.001.

Figure 3. SEM Path Diagram

Based on the SEM analysis, a positive relationship was determined between the Level of Digitalization
and Smart Logistics Applications (0.529), and between Smart Logistics Applications and Logistics
Performance (0.439), and H1 and H2 hypotheses were supported. In addition, it was seen that the Level of
Digitalization indirectly affected the Logistics Performance (0.217) and H3 hypothesis was supported.
Table 10 gives the regression weights, details of direct and indirect effects. The structural model of the
research is shown in Figure 4.
Table 10. Regression Weights, Direct And Indirect Effect

464
 Item Construct Standardized Estimate P-Value Result
(Path coefficient)

(Beta) (Path S.E T value R2 P-Value Result

coefficient)

DLà SLP 0.529 0.052 4.972 0.550 0.008 H1 Supported

SLP à LP 0.439 0,227 4.719 0.624 0.012 H2 Supported
DLàSLP àLP 0.217 0.105 3.237 0.581 0.004 H3 Supported
Significant at p < 0.05.

0.217

0.529 Smart
Digitalization 0.439 Logistics
Logistics
Level Performance
Practices

Figure 4. Structural Model

DISCUSSION
The article focuses on the effect of the level of digitalization and the application level of smart logistics
activities on logistics performance. Table 10 shows the path analysis results supporting all three
hypotheses. It shows the results of an effect decomposition based on the standardized coefficients presented
in Table 10. DL has a positive direct impact on SLP as assumed. Also, DL indirectly affects LP via SLP.
Again, SLP has positive direct effects on LP as assumed.
Figure 4 shows the working pattern with beta and R2 values. The increase in the DL level supports the
SLP application of the enterprises. Again, the increase in DL level indirectly causes an increase in LP. The
increase in the level of SLP implementation of the enterprises will also increase the LP. DL is positively
associated with both SLP and LP. It is also positively associated with SLP and LP. The results show that
businesses can improve LP by raising DL levels by co-executing SLP activities. As a hypothesis, it was
found that DL affects SLP (H1 Supported), SLP affects LP (H2 Supported), and DL affects LP via SLP
(H3 Supported) (Table 10).

CONCLUSION
Adopting a general systems perspective, we explore enterprises in the logistics sector related to SLP
applications on LP by DL level. DL is positively associated with SLP and indirectly affects LP via SLP.
The implementation of the level/maturity of digitalization as a strategy by the businesses will provide an
infrastructure for smart logistics activities. Firms with higher DL levels were found to perform higher
SLPs. Digitalization is a pioneering step necessary for SLP. Logistics businesses aiming to better satisfy
customers through enhanced LP should focus on digitalization and smart implementation strategies as a
focus strategy that acts as a conduit for other strategies. Ideally, both a high level of digitalization/maturity
and the proliferation of smart logistics practices would lead to better logistics performance.
While ı believe the study was successful in achieving its purpose, there are some limitations that provide
opportunities for future research. Data may have suffered from common method variance as part of non-
response bias, as the answers of a single participant from each enterprise were relied upon. In addition, not

465
 all complex relationships between variables were evaluated in the analysis. The research focused on
logistics businesses in the service sector, and therefore the results cannot be generalized for businesses in
the manufacturing sector.
Future research may collect data from multiple experts and managers in the logistics and supply chain
units of each enterprise. Again, future studies can explore using more variables and their effects on
different performance dimensions. Businesses in the manufacturing sector may also be included in future
research.

REFERENCE
Albukhitan, S. (2020). Developing digital transformation strategy for manufacturing. Procedia computer
science, 170, 664-671.
Bagozzi, R., ve Yi, Y., (1988). On the evaluation of structural equation model. Journal Academic Market
Science, 16 (1), 74-94.
Barki, H., Hartwick, J., (2001). Interpersonal con ict and its management in information system
development. Mis Quart., 195–228.
Barreto, L., Amaral, A., & Pereira, T. (2017). Industry 4.0 implications in logistics: an overview. Procedia
Manufacturing, 13, 1245-1252.
Bechtsis, D., Tsolakis, N., Vlachos, D., & Iakovou, E. (2017). Sustainable supply chain management in
the digitalisation era: The impact of Automated Guided Vehicles. Journal of Cleaner Production, 142, 3970-
3984.
Bickauske, D., Simanaviciene, Z., Jakubavicius, A., Vilys, M., & Mykhalchyshyna, L. (2020). Analysis
and perspectives of the level of enterprises digitalization (Lithuanian manufacturing sector case). Independent
Journal of Management & Production, 11(9), 2291-2307.
Bigliardi, B., Bottani, E., & Casella, G. (2020). Enabling technologies, application areas and impact of
industry 4.0: A bibliographic analysis. Procedia manufacturing, 42, 322-326.
Billon, M., Lera-Lopez, F., & Marco, R. (2010). Differences in digitalization levels: a multivariate
analysis studying the global digital divide. Review of World Economics, 146(1), 39-73.
Brozzi, R., Riedl, M., & Matta, D. (2021). Key Readiness Indicators To Assess The Digital Level of
Manufacturing SMEs. Procedia CIRP, 96, 201-206.
Cimini, C., Lagorio, A., Pirola, F., & Pinto, R. (2019). Exploring human factors in Logistics 4.0:
empirical evidence from a case study. IFAC-PapersOnLine, 52(13), 2183-2188.
Coelho, F., Relvas, S., & Barbosa-Póvoa, A. P. (2021). Simulation-based decision support tool for in-
house logistics: the basis for a digital twin. Computers & Industrial Engineering, 153, 107094.
Cyplik, P., Oleskow-Szlapka, J., Tobola, A., & Adamczak, M. (2019). Building a model for assessing the
maturity of polish enterprises in terms of logistics 4.0 assumptions. Business Logistics in Modern
Management, 105-120.
Douaioui, K., Fri, M., & Mabroukki, C. (2018). The interaction between industry 4.0 and smart logistics:
concepts and perspectives. In 2018 international colloquium on logistics and supply chain management,. 128-
132.
Edirisuriya, A., Weerabahu, S., & Wickramarachchi, R. (2018). Applicability of lean and green concepts
in Logistics 4.0: a systematic review of literature. In 2018 International Conference on Production and
Operations Management Society, 1-8.
Facchini, F., Oleśków-Szłapka, J., Ranieri, L., & Urbinati, A. (2020). A maturity model for logistics 4.0:
An empirical analysis and a roadmap for future research. Sustainability, 12(1), 86.
Gierlich, M., Schüritz, R., Volkwein, M., Hess, T., Xu, D., Jiang, J., & Kim, H. W. (2019). SMEs'
Approaches for Digitalization in Platform Ecosystems. In PACIS, 190.
Gregor, T., Krajčovič, M., & Więcek, D. (2017). Smart connected logistics. Procedia Engineering, 192,
265-270.
Heuser, L., Nochta, Z., & Trunk, N. C. (2008). ICT shaping the world: A scientific view.
Ismail, M. H., Khater, M., & Zaki, M. (2017). Digital business transformation and strategy: What do we
know so far. Cambridge Service Alliance, 10, 1-35.
Issaoui, Y., Khiat, A., Bahnasse, A., & Ouajji, H. (2019). Smart logistics: Study of the application of
blockchain technology. Procedia Computer Science, 160, 266-271.

466
 Jabeur, N., Al-Belushi, T., Mbarki, M., & Gharrad, H. (2017). Toward leveraging smart logistics
collaboration with a multi-agent system based solution. Procedia Computer Science, 109, 672-679.
Kauf, S. (2019). Smart logistics as a basis for the development of the smart city. Transportation Research
Procedia, 39, 143-149.
Korczak, J., & Kijewska, K. (2019). Smart Logistics in the development of Smart Cities. Transportation
Research Procedia, 39, 201-211.
Kostrzewski, M., Kosacka-Olejnik, M., & Werner-Lewandowska, K. (2019). Assessment of
innovativeness level for chosen solutions related to Logistics 4.0. Procedia Manufacturing, 38, 621-628.
McFarlane, D., Giannikas, V., & Lu, W. (2016). Intelligent logistics: Involving the customer. Computers
in Industry, 81, 105-115.
Oleśków-Szłapka, J., Wojciechowski, H., Domański, R., & Pawłowski, G. (2019). Logistics 4.0 maturity
levels assessed based on GDM (grey decision model) and artificial intelligence in logistics 4.0-trends and
future perspective. Procedia Manufacturing, 39, 1734-1742.
Pan, S., Zhong, R. Y., & Qu, T. (2019). Smart product-service systems in interoperable logistics: Design
and implementation prospects. Advanced engineering informatics, 42, 100996.
Parviainen, P., Kääriäinen, J., Tihinen, M., and Teppola, S. (2017). Tackling the digitalization challenge:
How to benefit from digitalization in practice, International Journal of Information Systems and Project
Management, 5(1), 63–77.
Romero-Martínez, A. M., & García-Muiña, F. E. (2021). Digitalization level, corruptive practices, and
location choice in the hotel industry. Journal of Business Research, 136, 176-185.
Ross, J. W., Beath, C. M., & Sebastian, I. M. (2017). How to develop a great digital strategy. MIT Sloan
Management Review, 58(2), 7.
Schuh, G., Scheuer, T., Nick, G., Szaller, Á., & Várgedő, T. (2021). A two-step digitalization level
assessment approach for manufacturing companies. Procedia Manufacturing, 54, 25-30.
Sternad, M., Lerher, T., & Gajšek, B. (2018). Maturity levels for logistics 4.0 based on NRW's Industry
4.0 maturity model. Business Logistics in Modern Management, 695-708.
Stoldt, J., Trapp, T. U., Toussaint, S., Süße, M., Schlegel, A., & Putz, M. (2018). Planning for
digitalisation in SMEs using tools of the digital factory. Procedia CIRP, 72, 179-184.
Uckelmann, D. (2008). A definition approach to smart logistics. In International Conference on Next
Generation Wired/Wireless Networking, 273-284.
Wang, M., Asian, S., Wood, L. C., & Wang, B. (2020). Logistics innovation capability and its impacts on
the supply chain risks in the Industry 4.0 era. Modern Supply Chain Research and Applications.
Wang, M., Asian, S., Wood, L.C. and Wang, B. (2020), Logistics innovation capability and its impacts on
the supply chain risks in the Industry 4.0 era. Modern Supply Chain Research and Applications, 2 (2), 83-98.
Wang, X. (2019). Study on relationship between green logistics activity and logistics
performance. Cluster Computing, 22(3), 6579-6588.
Yin, Y., Stecke, K. E., & Li, D. (2018). The evolution of production systems from Industry 2.0 through
Industry 4.0. International Journal of Production Research, 56(1-2), 848-861.
Zelbst, P. J., Green Jr, K. W., Abshire, R. D. and Sower, V. E. (2010). Relationships among market
orientation, JIT, TQM, and agility. Industrial Management & Data Systems, 110 (5), 637-658.

467