Journal of Enterprise Information Management

CNC router selection for SMEs in woodwork manufacturing using hesitant fuzzy AHP method
Alper Camci, Gül Tekin Temur, Ahmet Beskese,
Article information:
To cite this document:
Alper Camci, Gül Tekin Temur, Ahmet Beskese, "CNC router selection for SMEs in woodwork manufacturing using hesitant
fuzzy AHP method", Journal of Enterprise Information Management, https://doi.org/10.1108/JEIM-01-2018-0017
Permanent link to this document:
https://doi.org/10.1108/JEIM-01-2018-0017
Downloaded on: 17 June 2018, At: 17:32 (PT)
References: this document contains references to 0 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1 times since 2018*
Access to this document was granted through an Emerald subscription provided by emerald-srm:277069 []
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

 CNC router selection for SMEs in woodwork manufacturing
using hesitant fuzzy AHP method

ABSTRACT

Purpose: Despite being a low-tech industry, woodwork manufacturing industry that includes
furniture and cabinet making, witnessed technological leaps in production technologies due to
technical developments in computer numerical control (CNC) machining processes. The
managers of this industry have attached high importance to the selection of efficient machines
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

as their decisions directly affects the quality and performance of products produced by the
firms. Improper selection process can result in a significant decrease in productivity and
flexibility. Therefore, a systematic decision making procedure is needed to prevent inaccurate
investments on machines. In this study, a hesitant fuzzy analytic hierarchy process (HFAHP)
based multi criteria decision making (MCDM) system is proposed for CNC router selection in
Small and Medium Sized Enterprises (SMEs) in woodwork manufacturing.
Design/methodology/approach: The study proposes a hierarchical model consisting of 4
main criteria and 11 sub criteria for woodwork manufacturing. Technical, personnel,
economic and vendor aspects constitute the main criteria. Because of the hierarchical structure
of the model, HFAHP is utilised to define the importance weights of the criteria, and to select
the most appropriate CNC alternative for a manufacturing company under focus. In a
selection procedure, the judgments of decision makers may have vagueness to specify the
importance of criteria affecting the decision process. In the literature, the fuzzy set theory has
been utilised to deal with such uncertainties. However, when the ideas of the managers have
high potential to fall into contradiction in pairwise comparisons, a novel approach is needed to
overcome the obstacles. HFAHP allows the membership degree having a set of possible
values. It is specifically useful in compromised decisions where experts cannot agree on a
single value and prefer to come up with an interval of linguistic variables.
Findings: It is revealed that for SMEs in woodwork manufacturing, the most important
criterion in selecting the CNC routers is the technical aspects. It may seem counter intuitive
that they do not refrain finding the technical criteria superior to the economic aspects, even
though they have limited budgets compared to large scale firms. This demonstrates that in
current competitive environment, SMEs understand the need for high-quality production
strategy. The weights of the remaining two criteria (personnel and vendor aspects) are
relatively low because they expect that they can easily overcome the problem of adapting the
 workers by training, and all vendors have quality standard qualifications so they can offer a
satisfactory service and supplementary systems.
Practical implications: The ready-to-use model proposed, is specialized for SMEs in
woodwork manufacturing. However, to make it an easily adaptable model for every company
in the woodwork industry regardless of its size, the calculation process of the priority weights
is illustrated in detail with a numerical example. Any company can follow the process using
their own preferences to end up with a specific model that will perfectly reflect their own
specific priorities. For demonstrating the application of the model, a case study is conducted
in a woodwork manufacturing SME to select the best CNC router among three alternatives.
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Originality/value: The originality and value of the paper is two-fold. First, to the best of our
knowledge, this is the first study that proposes a woodworking-specific CNC router selection
for SMEs. Second, to handle the high uncertainty in the judgements, and to facilitate
consensus among the experts during face to face meetings to develop compromised matrices,
a very recently developed method, HFAHP is used.
Keywords: CNC Router/Machine Selection, Small and Medium Size Enterprises (SME),
Woodwork Manufacturing, Hesitant Fuzzy Analytic Hierarchy Process (HFAHP), Multi
Criteria Decision Making.

1. INTRODUCTION

Machining is the most flexible and precise manufacturing process during which designed part
is created by cutting away material sequentially from an unfinished piece (Groover, 2010). It
is usually the final process to achieve design tolerances impossible for other manufacturing
processes. Even though the largest application of machining is in the metal working industry,
the historical origins of almost every type of machining equipment and tool is in the
woodworking and carpentry. The woodwork manufacturing is an important industry group in
which the primary and secondary raw materials obtained from forestry are processed by
enterprises of various sizes with different production technologies to manufacture items such
as furniture, cabinets, carpentry and other wood products.

According to the United Nations Industrial Development Organization (UNIDO, 2017), in

2015 combined volume of manufacture of wood and of products of wood and cork, and
manufacture of furniture was 4.8% of total world manufacturing value added, with 1.5%
for the former and 3.3% for the latter. As a comparison, manufacture of motor vehicles,
 trailers and semi-trailers constitutes 8.1% and manufacture of textiles constitutes 2.4% share
of the world manufacturing value added (UNIDO, 2017). Italian research and consulting
company, the Centre for Industrial Studies (CSIL) reports that in 2017 the annual output of
furniture industries of 100 countries is US$ 420 billion (CSIL, 2017).

For millennia, woodworking machines are controlled manually or mechanically, but as the
advent of CNC (computer numerical control), woodworking machines proliferated across the
globe for their capabilities, speed, precision and cost and labor advantages. One of the most
common types of CNC woodworking machines is the flat table or nesting table CNC routers
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

used in the processing of wood products for furniture production and cabinetmaking. CNC
flat table routers are of a relatively simple design consisting of a base, a flat cutting table, a
gantry that carry the cutting spindle, a linear drive system that drives the gantry along the flat
table (x-axis), a linear drive system that drives the cutting along the gantry (y-axis), linear
drive that drives the cutting tool in the spindle up and down (z-axis) and finally computer
numerical control (CNC) system as shown in Figure 1.

Figure 1. Schematic of a flat table CNC router.

The routing process begins with the attachment of the unfinished sheet product (wood,
plywood or fibreboard) on the cutting table. This sheet material is fixated to the cutting table
through vacuum holes on the table. The finished products are designed using CAD/CAM
software then transferred to CNC router. The CNC router cuts the unfinished pieces according
to these CAD/CAM drawings using its high speed spindle.

In the contemporary business life, SMEs are considered major economic players making a
significant contribution to gross domestic product, and they constitute a potential source of
 economic growth (Taylor and Murphy, 2004; Love et al., 2005). Some 80 percent of world
economic growth comes from activities of SME's (Sadeghi, Azar and Rad, 2012). Regardless
of industry, establishing a successful technology-based SME is a very challenging and
complex issue (Sadeghi et al., 2012). This issue becomes more important for woodwork and
furniture industries considering the vast majority of the companies in those industries are
small and medium sized (Susnjara, 2006).

According to industry expert Ken Susnjara, even though lacking some of the flexibility of
CNC machining centers, due its cost, space and labor savings, a CNC flat table router is a
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

better fit for small and medium sized furniture and woodworking enterprises (Susnjara, 2006).
However, selecting the best CNC machine for SMEs is a very complex decision having both
quantitative and qualitative decision criteria. Decision makers’ inadequacy to evaluate the
pros and cons of adopting new technology in a holistic manner risks the competitive
advantage of many companies (Irani et al., 2002). Thus, selecting the right CNC router is a
critical decision for a woodworking SME, and should be handled properly.

This study aims to propose a dedicated hierarchical model and an HFAHP based methodology
to help the decision makers in woodwork manufacturing SMEs in their CNC selection
process. The hierarchical model consists of four main criteria, related with technical,
personnel, economic and vendor aspects, and 11 sub criteria. Those criteria are evaluated by
three experts using pairwise comparisons. The experts are required to discuss their
preferences to come up with a compromised decision. The importance weights of the criteria
to be used in the selection process are then obtained by using HFAHP.

The ready-to-use model proposed reflects the preferences of experts having intense
experience in such decisions in woodwork manufacturing SMEs. However, any company can
follow the process using their own preferences to modify the model addressing their own
priorities. To make it easily adaptable, the calculation process of the priority weights is
illustrated in detail with a numerical example. For demonstrating the application of the model,
a case study is conducted in a woodwork manufacturing SME to select the best CNC router
among three alternatives.

The rest of the study is organized as follows: Section 2 gives a literature review for CNC
machine selection. Section 3 includes the methodology focussing on HFAHP. Section 4
 presents the hierarchical decision model for the problem at hand, and the calculation of the
importance weights of the criteria of the model. Section 5 illustrates the usage of the model in
a real-world case from a furniture manufacturing SME and presents the sensitivity analysis.
Finally, Section 6 concludes the current research results and gives brief information about
possible directions for future research related to this study.

2. LITERATURE REVIEW ON CNC MACHINE SELECTION

The proliferation of CNC production technologies has a disruptive effect on the woodwork
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

manufacturing companies. The adoption of CNC machines has been increasing ever since
they were first introduced in mid 1980s (Susnjara, 2006). Using survey data from 174
furniture, cabinet and architectural millwork companies, Wiedenbeck and Parsons (2010)
report that 93% of large firms, 73% of medium firms and 17% of small firms adopted CNC
machines in their production lines. Moreover, they also report that the companies adopting
these technologies report more significant revenue increases compared to those who did not
adopt (Wiedenbeck and Parsons, 2010). Technological advances and the threat of global trade
force companies in furniture and cabinet woodworking industry to implement CNC type
technologies to stay competitive and produce high quality products. For these reasons,
selecting a CNC machine is a crucial decision for a manufacturer, as wrong decisions might
affect the competitiveness of the company significantly.

To our knowledge, there is a lack of studies directly related to woodworking-specific CNC

machine or router selection problem for SMEs. However, CNC machine selection problem in
metal working has been the subject of numerous studies in the literature over the years. Most
common approach to tackle CNC selection problem has been the multi criteria decision
making (MCDM), where mutually exclusive alternatives are evaluated based on the
predetermined criteria which can be both quantitative and qualitative. The Analytic Hierarchy
Process (AHP) method, developed by Saaty (1981) has been the most widely used method
applied in CNC machine selection problems. Lin and Yang (1996) developed a model based
on AHP for the machine selection among the alternatives of CNC, NC, or conventional
machines. Oeltjenbruns et al. (1995) and Çimren et al. (2007) utilized AHP to evaluate and
rank CNC machine tools. In combination with AHP, Myint and Tabucanon (1994) used goal
programming (GP), Yurdakul (2004) applied Analytic Network Process (ANP) and
Dağdeviren (2008) utilized PROMETHEE methods for machine selection problems.
 Every study mentioned so far assumed that the data sets in the decision making process
consist of only crisp numbers representing exact values. In real life, it is not easy to give exact
valuation in complex decision situations. So, another popular approach to decision making
problems is the incorporation of fuzzy set theory developed by Zadeh (1965) which takes into
consideration vagueness and ambiguity associated with the decision making by using
approximate information and uncertainty in the decision process instead of crisp values used
by methods like classical AHP, thus simplifying and improving many engineering decision
problems (Öztayşi, 2015). For CNC machine selection, Wang et al. (2000) used a fuzzy
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

multiple attribute decision-making model, Ayağ and Özdemir (2006) proposed a fuzzy AHP
approach, and expanding on the selection criteria from this study utilised a fuzzy ANP
method (Ayağ and Özdemir, 2011). Önüt et al. (2008) applied fuzzy TOPSIS approach in
CNC selection problem combined with the fuzzy AHP which was used to determine the
criteria weights. And also it is common for same researchers to publish studies on the same
CNC selection problem using different approaches. Taha and Rostam (2011) developed an
artificial neural network (ANN) to verify the results of fuzzy AHP solution, and then in their
next study (Taha and Rostam, 2012), using a reduced number of criteria, utilized a hybrid
fuzzy AHP and PROMETHEE method where the fuzzy AHP was used to determine the
criteria weights again.

Even though fuzzy methods proved to be more successful in handling the imprecise and
vague information compared to crisp methods, when two or more sources of vagueness
appear simultaneously, the modelling tools of ordinary fuzzy sets are limited (Rodriguez et
al., 2012). In order to accommodate extensive vagueness, fuzzy set theory has been evolving
into type-2 fuzzy sets, non-stationary fuzzy sets, fuzzy multi sets, intuitionistic fuzzy (IF) sets
and hesitant fuzzy sets (Rodriguez et al., 2012). For CNC selection problem, some of these
new evolved approaches used in the literature are, type 2 fuzzy AHP (Kahraman and Öztayşi,
2013), interval-valued intuitionistic fuzzy present worth analysis (Kahraman et al., 2016) and
IF entropy and TOPSIS (Tho et al., 2013).

The literature review of the MCDM approaches used in CNC machine selection problems
reveals several practices; the first one is that increasingly fuzzy and evolved fuzzy methods
are being used in such problems, and the second one is that, even though been around more
than 3 decades and numerous evolutions, AHP still proves to be a robust and popular method
 for multi criteria decision making. Recognizing the results of the literature review, this study
contributes to the literature by (1) extending the applications of evolved fuzzy decision tools
in CNC machine selection by utilizing hesitant fuzzy approach, (2) applying the proposed
MCDM tool in woodworking CNC router selection by considering woodworking specific
criteria, and (3) incorporating a modified evaluation process that includes consensus among
the experts rather than individual evaluations of experts.

In the literature the selection criteria for CNC machines differ from study to study. Some
researchers limited their main criteria set to technical and process aspects (Ayağ and Özdemir,
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

2006; Ayağ and Özdemir, 2011; Çimren et al., 2006; Taha and Rostam, 2011; Taha and
Rostam, 2012; Tho et al., 2013), whereas some others include economic aspects into the list
(Myint and Tabucanon, 1994; Oeltjenbruns et al., 1995; Wang et al., 2000; Yurdakul, 2004;
Dağdeviren, 2008; Önüt et al., 2008). One interesting findings is that the computer system
related criteria did not appear in any other research studies except for the study by Önüt et al.
(2008). Similarly, personnel aspects are also only mentioned in Önüt et al. (2008).

Considering these studies, it can be stated that there is a need for further research to determine
the specific selection criteria for CNC routers in woodworking industry, and propose a novel
approach that has not been utilized before.

3. METHODOLOGY

3.1. Hesitant Fuzzy Sets

Fuzzy set theory was specifically designed to deal with uncertainty, vagueness and
imprecision intrinsic to many real world problems (Beskese et al., 2004). However, for some
multi criteria decision making problems, it is very hard to tackle with high uncertainty even
by using traditional fuzzy sets. Thus, in search for reaching better definitions of membership
values or functions of the parameters in decision making problems, ordinary fuzzy sets have
been recently extended to intuitionistic and hesitant fuzzy sets (HFS) (Öztayşi et al., 2015).
HFS (Torra and Narukawa, 2009; Torra, 2010) can handle problems in which experts do not
have a certain idea to define their choices over alternatives (Öztayşi et al., 2015). As Sun et
al. (2018) and Li et al. (2015) mention, HFSs are useful to handle uncertainty that occurs if an
 expert has no exact idea on selection of a membership degree. To do so, it allows the
membership degree having a set of possible values. This is especially useful when a
compromised decision is preferred over an aggregated one in group decision making.

As a novel approach, hesitant fuzzy sets have been utilised by many researchers for many
kinds of problems theoretically and practically. Öztayşi et al. (2015) developed the HFAHP
for a supplier selection process in which it is hard to define a certain membership degree to a
value. Başar (2017) proposed a HFAHP based model to forecast the cost of software which
carries high uncertainty because of vagueness in development method, data, design and
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

forecasting. Liang and Xu (2017) extended a traditional TOPSIS approach with Pythagorean
fuzzy sets and conducted a case study for energy project selection to validate the applicability
of the new method. Liu et al. (2017) determined novel operational rules to propose a
comparative method to hesitant intuitionistic fuzzy linguistic set (HIFLS) and generated two
prioritized aggregation operators which are utilised under the conditions of HIFLS. Sun et al.
(2018) claimed that there are some limitations in classical hesitant fuzzy sets belonging to
measures of distance, correlation coefficient and similarity, and therefore they proposed a
hesitant fuzzy set grey relational degree base method.

3.2. Hesitant Fuzzy Analytic Hierarchy Process (HFAHP)

Analytic Hierarchy Process (AHP) can be fuzzified under the title of “Fuzzy AHP” if pairwise
comparisons have linguistic terms which have vagueness. However, if there is high
uncertainty on the judgments of experts, or they cannot decide on a single value for their
pairwise comparisons, Hesitant Fuzzy AHP (HFAHP) can be used. HFAHP allows the
membership degree having a set of possible values. It is specifically useful in compromised
decisions where experts cannot agree on a single value and prefer to come up with an interval
of linguistic variables.

HFAHP procedure which is depending on Buckley’s theory can be summarized in ten steps as
adapted from Öztayşi et al. (2015).

Step 1) A hierarchical model is built consisting of main and sub criteria which have direct
impact on decision making problem, and alternatives.
Step 2) Similar to a traditional AHP procedure; criteria, sub criteria and alternatives are
 evaluated using pair wise comparisons. The evaluations are made by using linguistic terms in
Table 1. The experts are not forced to select only one value whenever they could not agree on
a single linguistic term. They are let to keep those answers as a range as long as the distance
between two edges of this range is no more than “2”. If the range is larger, the evaluations
should be renewed with negotiations of the experts.

INSERT TABLE 1

Step 3) The fuzzy envelopes approach is utilised (Liu and Rodriguez, 2014). The first thing to
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

do is constructing the fuzzy envelopes for the compromised decisions of the experts. Then,
the scales are ordered from lowest to highest value (from
 to
 ). Also, the evaluations are
varying between two terms:
 to
 . Therefore, we can see that
 ≤
 <
 ≤
 ). In fuzzy
envelope approach, OWA operator of dimension  is used which can be calculated by using
Eq. (1).

OWA ( , ,…. ) = ∑   (1)

where  is the largest of the aggregated arguments , , … . and W =

( ,  , … .  ) is the associated weighting vector where ∑  = 1.
On the other hand, a and d parameters of a trapezoidal fuzzy membership function which is
shown by  = (a, b, c, d), can be found by using Equations (2-5):

a=min , , ,…, !
     
 , 
= (2)

d=max , , ,…, !
     
 , = (3)

%  , &' & + 1 = ) @

# #
(4)

b= *+, - ./0,…,/0
1 (231)/-
5 ,26 231 27 898:
$*+ ?
# 5,26 231 27 <== #
" >
1 (2313;)/-
- ./0, …,/0
,

%

 , &' & + 1 = ) @
# #
(5)

c= *+,- ./10,/1A;
0, …,/0
(231)/-
5 ,26 231 27 898:
$*+ ?
# 5,26 231 27 <== #
" >
1 1A; (2313;)/-
- ./0, /0, …,/0
,
 Assume that α is in the unit interval B0,1D. First and second types of weights are specified by

using α (Filev and Yager, 1998) as in Eq. (6) and Eq. (7). Where α = and α =
F(F) (F)F
F F

(G is the top rank number of evaluation scores (for instance, in Table 1 the top rank number is
10), ) is the rank of highest evaluation and & is the rank of lowest evaluation.), first type of
weights (+ = ( ,  ,…,  )) can be calculated by using Eq. (6).

 = α ,  = α (1 − α ),…,  = α (1 − α ) F (6)
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Second type of weights (+ = ( ,  ,…,  )) can be calculated by using Eq. (7).

 =α F
,  = (1 − α )α F
,…,  = 1 − α , (7)

Step 4) IJ that shows the collaborative pair wise comparison matrix is prepared where K̃ =
(KM , K0; , K0- , KN ) as shown in Eq. (8):

1 ⋯ K̃Q
IJ = O ⋮ ⋱ ⋮ T
(8)

K̃ Q ⋯ 1

The reciprocal values are calculated as shown in Eq. (9).

K̃ = . , , , 5
(9)
U21N U210- U210; U21M

Step 5) In order to check the consistency, the values of the matrices are defuzzified (Erdoğan
and Kaya, 2016; Otay et al., 2017). Considering the trapezoidal fuzzy number dW =
(X, Y , Y , Z), the crisp number (μ\ ) can be calculated by using Eq. (10).

X + 2Y + 2Y + Z
μ\ =
6 (10)

Consistency ratio (CR) is then calculated by using Eq. (11), and (12);
 `Y −
I_ =
a
−1 (11)

I_
Ib =
b_
(12)

where I_ refers to the consistency index, `cde symbolizes the largest eigenvector of the
matrix,  is the number of criteria, and b_ is the random index that differs for each . If the
Ib value is lower than 0.1, it can be accepted as satisfactory and the result is found as
consistent.
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Step 6) Each row’s geometric mean (f̃ ) is calculated by using Eq. (13).

f̃ = (K̃ ⊗ K̃ … ⊗ K̃ ) / (13)

Step 7) By assuming that f̃Q ⊕ f̃ … ⊕ f̃ is the maximum parameter of ALI (shown in Table
1, it is 9) to decrease the deviation in the weights (Öztayşi et al., 2015), each criterion’s and
each alternative’s fuzzy weights (
i  ) are calculated by using Eq. (14).


i  =f̃ ⊗ (f̃ ⊕ f̃ … ⊕ f̃ ) F (14)

Step 8) Defuzzification of the fuzzy numbers is performed by using Eq. (15).

KQ + 2Kc + 2Kc + Kk
j=
6
(15)

Step 9) Normalization is applied to defuzzified weights to obtain the local weights of the
criteria. To calculate the global weights of the sub-criteria, their local weights are multiplied
with the weight of the main-criteria they belong.

Step 10) To calculate the score of each alternative i (S ), previous steps are rerun for each
matrix formed for each alternative. The final fuzzy scores can be retrieved by using Eq. (16).

S = ∑ 
 , ∀& (16)

where  is the global weight of criterion ) and
 is the score of alternative criterion & with
respect to ).
 By ranking the defuzzified values, the best alternative is obtained.
4. PROPOSED MODEL

4.1. Structure of the Proposed Model

The main criteria that affect selecting the best CNC router are obtained from the relevant
literature focusing on CNC routers and face to face interviews which are conducted with
highly experienced experts from woodwork manufacturing industry. The hierarchical model
for CNC router selection can be seen in Figure 2.
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

The Goal:
CNC Router Selection for Woodwork Manufacturing

Technical Aspects (C1) Personnel Aspects (C2) Economic Aspects (C3) Vendor Aspects (C4)

T1 T2 T3 T4 P1 P2 E1 E2 V1 V2 V3

Aternative 1 Alternative 2 Alternative 3

Figure 2. The Proposed Model for CNC Router Selection

Technical aspects involve the overall design, process, performance and technology
characteristics that affect the efficient operation of the CNC router. It consists of four main
sub criteria:
- CNC router construction (T1) is associated with cutting table material, table
construction and stiffness of the router. Due to much higher spindle rotation speeds
compared to those of metal cutting CNC machines, woodworking CNC routers require
a more stiff structure with superior vibration dampening characteristics (Annamalai,
2003).
 - Process design (T2) is associated with how effectively the CNC router works as a
whole with its main and supplementary systems, including dust collection and cooling
systems, power usage and compatibility to worker health and safety.
- Design performance criteria (T3) is associated with the performance of CNC router
design during cutting operations with cutting spindle performance, tool change
flexibility, vacuuming flexibility and automatic part flow.
- CNC computer system criteria (T4) is associated with the level of computer
technology of CNC routers with characteristics of 3D programming level and level of
advancement of the CNC computer unit including Wi-Fi capability.
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Personnel aspects are about the ease with which the workers of the company adapt to the new
machine and their capability levels required to achieve the highest performance. It consists of
two main sub criteria: adaptation capability of workers (P1) and worker technical capability
needed (P2).

Economic aspects are related with fixed and variable costs associated with the purchase and
operation of the CNC router. It can be categorized into two parts: initial costs including
purchase cost and set up cost (E1) and running cost which is directly resulted from power
consumption of the machines (E2).

Vendor aspects are associated with qualitative characteristics of the suppliers and the
manufacturers of the routers, including vendor CNC knowhow (V1), vendor service response
capability (V2) and reputation of the vendor/brand (V3).

4.2. Criteria Weights

For determination of general weights of the model, three experts were requested to come
together, and evaluate the criteria and sub criteria using pairwise comparisons. All those
experts had different industry backgrounds and considerable experiences in selection and
procurement of CNC type machinery. To utilize the synergy of the different perspectives, they
were not asked to give their individual preferences. Rather, they discussed several facets of
each comparison, and tried to come up with a compromised decision. Originally, classical
Buckley’s approach was intended to use. However, as the experts started their discussions, it
was witnessed that they were unable to agree on one single linguistic term for many of the
 comparisons. There was not a long distance between the experts’ final ideas, but a
compromise couldn’t be reached. Rather than assigning a specific value such as 3, for
example, they preferred to use the statement “Between 2 and 4”. This led the authors to use a
novel approach called HFAHP which looks like a very good solution to the problem at hand.

In order to perform HFAHP, the evaluation results retrieved from the consensus, have to be
close to each other (maximum difference between ranks have to be “2”). If they are not, the
evaluation process should be renewed until consensus that gives the satisfactory results is
reached. After achieving the closeness, it becomes possible to establish compromised pairwise
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

comparison matrices. The experts were encouraged to further negotiate and finally they
reached to a consensus satisfying the aforementioned rule.

To make the evaluation process more understandable, two examples are given in this section.
One of them includes a pairwise comparison in which the distance between two edges of the
range equals to 2. Second one consists of a pairwise comparison in which the distance
between two edges of the range equals to 1.

First of all, four main criteria are evaluated using pairwise comparison. The consistency ratio
is found as 0.045 which is highly satisfactory. Table 2 indicates the fuzzy envelops of the
main criteria evaluations.

INSERT TABLE 2

Then the trapezoidal fuzzy sets are calculated by applying OWA operator to the fuzzy
envelops. For instance, for fuzzy envelope of criterion 1 and criterion 2 which is specified as
“Between WHI and VHI”, the trapezoidal fuzzy set which is shown as (a, b, c, d) is found as (1,
4.77, 5.22, 9) as shown in Table 3. First, α and α are found. For this case, G which is the top
rank number equals to 10; & which is the rank of lowest evaluation (WHI) equals to 7 and )
which is the rank of highest evaluation (VHI) equals to 9. Then;

α = = 0.88
F(nFo)
F

α = = 0.11
(nFo)F
F
  
Since a=  and d= ; a=1 (minimum value of WHI shown as 1, 3, 5) and d=9 (maximum
value of VHI shown as 5, 7, 9).
If & + ) is even, then  = *+ 1 (231)/-
- ./0, …,/0 5.
In our case, & + ) = 16; therefore
,

 = α ∗ 3 + α ∗ 5 = 0.88 ∗ 3 + 0.11 ∗ 5 = 4.778

If & + ) is even, then K = *+ 1 1A; (231)/-

- ./0, /0, …,/0 5.
,

K = 2 ∗ 5 − 4.778 = 5.22
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

So, (a, b, c, d) equals to (1, 4.77, 5.22, 9).

INSERT TABLE 3

As the next step, geometric means are calculated for the trapezoidal fuzzy sets. For instance,
the calculation of the first row is calculated as following:

 = (1 ∗ 1 ∗ 1 ∗ 1) /v
= 1
 = (1 ∗ 4.77 ∗ 1 ∗ 3) /v
= 1.94
K = (1 ∗ 5.22 ∗ 1 ∗ 5) /v
= 2.26
x = (1 ∗ 9 ∗ 5 ∗ 7) /v
= 4.21

For normalization, all values are divided by 9 which is the highest score in the linguistic scale
table.
, = 1/9 = 0.11
, = 1.94/9 = 0.21
K, = 2.26/9 = 0.25
x, = 4.2/9 = 0.46
Table 4 shows the trapezoidal fuzzy weights.

INSERT TABLE 4

Previous steps are performed for other main criteria and the crisp weights are obtained by
defuzzifying the fuzzy numbers by using Eq (2). The crisp weights of all main criteria can be
seen in Table 5.
 INSERT TABLE 5

From these results, we can clearly see that the overriding criterion is the technical aspects
which basically ensure the quality production using high speed CNC machining. At this level,
even the small variations can lead to large undesired effects in woodwork manufacturing. The
second most important criterion is the economic aspects, because of the sensitivity of the
SMEs to capital. The weights of the remaining two criteria (personnel and vendor aspects) are
relatively low due to fact that the worker and vendor characteristics in the industry are not
very different from each other. On the other hand, the literature review revealed that there is
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

no consensus on the most important criteria in selecting the CNC machines. For some the
economic criteria rank highest (Oeltjenbruns et al.1995; Wang et al., 2000; Önüt et al. 2008),
and yet some others rank technical criteria as the most important (Dağdeviren, 2008; Taha and
Rostam, 2011; Taha and Rostam, 2012).

The same steps are performed for sub criteria. First of all, technical aspects are taken into
consideration. The consensus results of the evaluation of the sub criteria of technical aspects
(T1, T2, T3, T4) are transformed to fuzzy envelops in pairwise comparison matrices, as
shown in Table 6. Also, the consistency is found as 0.002 which is highly satisfactory.

INSERT TABLE 6

Then the trapezoidal fuzzy sets are calculated by applying OWA operator to the fuzzy
envelops. For instance, for fuzzy envelope of T1 (CNC router construction) and T4 (process
design) which is specified as “Between EHI and WHI”, the trapezoidal fuzzy set which is
shown as (a, b, c, d) is found as (1, 1, 3, 5) as shown in Table 7. Initially, α and α are found.
For this case, G which is the top rank number equals to 10; & which is the rank of lowest
evaluation (EHI) equals to 6 and ) which is the rank of highest evaluation (WHI) equals to 7.
Then;

α =
F(oFy)
F
=1

α =
(oFy)F
F
=0

 
Since a=  and d= ; a=1 (minimum value of EHI shown as 1, 1, 3) and d=5 (maximum
 value of WHI shown as 1, 3, 5).

If &' & + 1 = ); then  = .


In our case, & + 1 = ); therefore b = 1.

If &' & + 1 = ); then K = 

 . In our case, & + 1 = ); therefore c = 3.

So, (a, b, c, d) equals to (1, 1, 3, 5).

INSERT TABLE 7
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

As the next step, geometric means are calculated and the results are divided to 9. Table 8
shows the trapezoidal fuzzy weights.

INSERT TABLE 8

Previous steps are performed for other sub criteria and the crisp weights are obtained by
defuzzifying the fuzzy numbers by using Eq. (15). The weights of main criteria and global
weights of all sub criteria can be seen at Table 9.

INSERT TABLE 9

As previously stated, the most important criterion in this decision model is the technical
aspects. And the further evaluation of the criterion shows us that first of all, the construction
characteristics of the CNC routers like structural stiffness, table construction and material play
critical role in CNC router selection. Following closely the construction characteristic, design
performance criteria which include the cutting spindle performance, tool change flexibility,
vacuuming flexibility and automatic part flow, emerges as the second important factor. Both
these results conform to the critical factors identified by Annamalai (2003) for high speed
CNC machines for woodwork and furniture manufacturing.

In the second most important criterion, the economic aspects, the initial cost of the machine
has more importance than running cost. According to Annamalai (2003), for lower production
volumes decrease, the initial cost of the CNC router has much larger impact on the unit cost
 of the parts produced. This is the main reason why for an SME initial cost of the CNC router
is a very important factor.

One interesting finding is that the adaptability of the worker to a new CNC router determined
as more important than the technical capability of the workers. This shows that the place a
bigger importance on seamlessly adapting the new machines to their existing operations than
worrying about whether their workers have technical capabilities required for new machinery.

5. CASE STUDY IN WOODWORK MANUFACTURING

Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

According to the Union Chambers and Commodity Exchanges of Turkey (TOBB, 2013), in
2012 there are over 31.000 enterprises with 151.000 employees working for the Turkish
furniture products assembly sector with annual exports exceeding 1.9 Billion USD and
comprising 1% of the world total furniture production. As these figures suggest, with the
exception of few large manufacturers, majority of the companies in Turkish furniture
manufacture and assembly industry (similar to other countries) are small and medium sized
companies.

On the raw material side, Turkey has a competitive wood industry, especially in supplying
semi-processed wood to many developed countries. Recently, Turkey also has enlarged its
capacity in forest based products and has the advantage of taking position for supplying
products to Middle East (Ministry of Economy, 2016).

The three alternative CNC routers are compared by adopting the scoring data obtained from
the case study in (Aybar, 2017). In this case study, CNC selection decision of an SME
company that serves in the furniture sector has been discussed (Aybar, 2017). In the
company, desk and cupboard-oriented office furniture are produced with around 150 tables
and cabinets per month with 12 employees. The company wants to buy a CNC router in order
to increase the output, prevent the accidents, decrease the scraps and increase the motivation
of the employees.

For determination of the weights of CNC alternatives, three experts give the scores to three
alternative CNC routers. One of them is a 25 years experienced foreman who has his own
wood processing shop. He also has 10 year experience in CNC machining. Second one is an
 experienced sales technician in a leading European CNC router manufacturing company.
Third one is a highly experienced CNC operator who has experience in numerous different
CNC machinery during his career. At the end of the scoring process, three experts are asked
to evaluate three alternative CNC routers with respect to all sub criteria.

The fuzzy envelops for evaluation of three alternatives (A1, A2, A3) in terms of sub criteria are
given at Table 10, 11, 12 and 13.

INSERT TABLE 10
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

INSERT TABLE 11

INSERT TABLE 12

INSERT TABLE 13

Then the trapezoidal fuzzy sets are calculated by applying OWA operator to the fuzzy
envelops. For fuzzy envelopes for alternatives with respect to CNC router construction (T1),
comparison of alternative 1 and alternative 2 which is specified as “Between ELI and EE”, the
trapezoidal fuzzy set which is shown as (a, b, c, d) is found as (0.33, 1, 1, 1) as shown in
Table 14. Initially, α and α are found. For this case, G which is the top rank number equals
to 10; & which is the rank of lowest evaluation (ELI) equals to 4 and ) which is the rank of
highest evaluation (EE) equals to 5. Then;

α =
F(|Fv)
F
=1

α =
(|Fv)F
F
=0

 
Since a=  and d= ; a=0.33 (minimum value of ELI shown as 0.33, 1, 1) and d=1
(maximum value of EE shown as 1, 1, 1).

If &' & + 1 = ); then  = .


In our case, & + 1 = ); therefore b = 1.
If &' & + 1 = ); then K = 
 . In our case, & + 1 = ); therefore c = 1.
So, (a, b, c, d) equals to (0.33, 1, 1, 1).
 INSERT TABLE 14

As the next step, geometric means are calculated and the results are divided by 9. After
normalization, the defuzzification is performed and alternative CNC routers are ranked in
terms of their importance on each sub criterion. The global weights of alternative CNC routers
in terms of sub criteria can be seen at Table 15.

INSERT TABLE 15
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

At the end, all obtained weights are shown to the experts to check whether the weights fit the
ideas of them. The experts become agree that the weights reflect their judgments correctly.
After that, the crisp values of weights of main criteria, sub criteria and alternatives in terms of
criteria are obtained. To conclude the HFAHP process, it is necessary to multiply the sub
criteria weights of alternatives with obtained criteria weights. Then, the obtained values of
each alternative are multiplied with the weights of criteria. The final weights of alternatives
are given at Table 16. It is seen that third alternative is the best that satisfy the needs of the
manufacturer.

INSERT TABLE 16

In order to explore how changes in weights of main criteria affect the preference ranking of
alternatives, sensitivity analysis is conducted. To do so, the criteria weights are changed one
at a time from 0.05 to 0.95 while the weights of the rest of the criteria are kept proportionally
the same. Figure 3 parts a, b, c, and d illustrate the sensitivity analysis results for technical,
personnel, cost, and vendor aspects, respectively.
 (a)SENSITIVITY OF TECHNICAL ASPECTS (b) SENSITIVITY OF PERSONNEL ASPECTS

A1 A2 A3 A1 A2 A3

0.45 0.4
0.4 0.35
0.35
0.3
0.3
0.25
0.25
0.2
0.2
0.15
0.15
0.1 0.1

0.05 0.05
0 0
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

(c) SENSITIVITY OF ECONOMıC ASPECTS (d) SENSITIVITY OF VENDOR ASPECTS

A1 A2 A3 A1 A2 A3

0.45 0.37
0.4 0.36
0.35 0.35
0.3 0.34
0.25 0.33
0.2 0.32
0.15 0.31
0.1 0.3
0.05 0.29
0 0.28

Figure 3. Sensitivity Analysis of Main Criteria with respect to Different Weights

As can be seen from Figure 3 (b) and (d), the results are not sensitive to changes in personnel
and vendor aspects. It means that, the preference rankings of alternatives remain the same
regardless of the importance weight changes of those criteria. Therefore, no matter how the
individual criterion weight changes for personnel aspects, and vendor aspects, A3 remains the
best alternative, and A2 is the runner-up.

In part (a) of Figure 3, it can be seen that an increase in the criterion weight of technical
aspects does not affect the ranking of the alternatives. However, as the criterion weight of the
technical aspects is decreased below 0.25, it is observed that the ranking of A1 starts to
increase, and the ranking of A3 starts to decrease. This finding is in line with the initial
expectations of the experts, because A1 has the least technical advantages whereas A3 has the
most technically advanced properties.

In part (c) of Figure 3, it can be concluded that the decrease in the criterion weight of
economic aspects does not change the ranking of the alternatives. However, as the weight is
 increased above 0.40, the rank of A3 starts to decrease, and after the weight 0.6 on, the overall
ranking of the alternatives becomes A1, A2, and A3. This result also agrees with the initial
expectations of the experts, because A3 has the least economic advantages whereas A1 has the
most.

6. CONCLUSION

Machinery selection involves a high number of considerations including quantitative and

qualitative factors. In this study, a systematic multi criteria decision making approach is
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

proposed for woodworking-specific CNC router selection under the uncertainties of the
judgments of decision makers. The reasons for uncertainty can stem from insufficiency of
extensive and systematic knowledge. For example, an expert can score the relative importance
of one factor as “Between 3 and 4” rather than using an exact value. As a novel fuzzy
approach, HFSs are utilised to tackle vagueness occur in assignment of scores for pairwise
comparison of criteria and alternatives.

In the literature, there is a lack of studies tackling uncertainties in woodworking-specific CNC

router selection process. This study contributes to the relevant literature by proposing a
woodworking-specific CNC router selection for SMEs that handles the high uncertainty in the
judgements, and by facilitating consensus among the experts during face to face meetings to
develop compromised matrices. This generic model can be adopted by every company within
the industry by performing the steps clarified in the case study which is conducted to
demonstrate the applicability of the proposed model. Four main criteria including technical,
personnel, economic and vendor aspects are taken into consideration and three CNC routers
are evaluated within pairwise comparisons that carry hesitancy, and the best alternative is
chosen among three alternative CNC routers by checking the ranking scores. As a result, it is
revealed that the most important criterion for SME’s in woodwork manufacturing is the
technical aspects. The evaluation of sub criteria of the technical aspects indicate that first of
all, the construction characteristics of the CNC routers like structural stiffness, table
construction and material play critical role in CNC router selection. The second most
important criterion is the economic aspects in which the initial cost of the machine has higher
effect on CNC router selection compared to the running cost. It is also noticeable to mention
that the decision makers attach higher importance to the adaptability of the worker compared
to the technical capability of the workers. According to the results, the company decided to
 buy the third CNC router. Furthermore, a sensitivity analysis is conducted in order to reveal
how changes in importance of main criteria affect the results. It is observed that, as it is
expected, (1) the rank of A3 which has the highest technical advantage decreases when the
importance of technical aspects decreases, and (2) the rank of A3 which has least economic
advantage decreases when the importance of economic aspects increases. In order to confirm
whether the company is satisfied with this decision, we interviewed with the owner of the
company, and he stated that they are pleased with this decision because their production
volume is increased, the number of accidents are decreased and the motivation of the
employees is increased.
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

In the future, other extensions of fuzzy sets such as intuitionistic fuzzy sets or Pythagorean
fuzzy sets can be utilized to overcome uncertainty. The model also can be extended for large-
scale firms by arranging the criteria.

REFERENCES

Annamalai, S. (2003). “An investigation of high speed machining on CNC routers used for
upholstered furniture manufacturing”, Master’s Thesis, NC State University, available at:
/ http://www.lib.ncsu.edu/resolver/1840.16/2225. (Accessed 12 January 2018).

Aybar, M. (2017). CNC Machine Selection with Fuzzy TOPSIS Approach, Beykent
Üniversitesi / Fen Bilimleri Enstitüsü / Endüstri Mühendisliği Anabilim Dalı.

Ayağ, Z. and Özdemir, R.G. (2006), “A fuzzy AHP approach to evaluating machine tool
alternatives”, Journal of Intelligent Manufacturing, Vol. 17, pp. 179-190.

Ayağ, Z. and Özdemir, R.G. (2011), “An intelligent approach to machine tool selection
through fuzzy analytic network process”, Journal of Intelligent Manufacturing, Vol. 22, No.
2, pp. 163-177.

Başar, A. (2017). “Hesitant fuzzy pairwise comparison for software cost estimation: a case
study in Turkey”, Turk J Elec Eng & Comp Sci, Vol. 25, pp. 2897-2909.

Beskese, A., Kahraman, C. and Irani, Z. (2004) “Quantification of flexibility in advanced

manufacturing systems using fuzzy concept”, International Journal of Production Economics,

 Vol. 89, No. 1, pp. 45–56.

Çimren, E., Catay, B. and Budak, E. (2007), “Development of a machine tool selection
system using AHP”, International Journal of Advanced Manufacturing Technology, Vol. 35,
pp. 363-376.

Centre for Industrial Studies (CSIL) (2017). World Furniture Outlook 2017 / 2018. Milano,
Italy, available at: https://www.worldfurnitureonline.com/PDF/press-
release/W0_July17_PR.pdf/ (accessed 12 January 2018).
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Erdoğan, M. and Kaya, Đ. (2016), “A combined fuzzy approach to determine the best region
for a nuclearpower plant in Turkey”, Applied Soft Computing, Vol. 39, pp. 84-93.

Dağdeviren, M. (2008), “Decision making in equipment selection: an integrated approach

with AHP and PROMETHEE”, Journal of Intelligent Manufacturing, Vol. 19, pp. 397–406.

Filev, D. and Yager, R.R. (1998), “On the issue of obtaining OWA operator weights”, Fuzzy
Sets and Systems, Vol. 94, Issue 2, pp.157-169.

Groover, M. (2010), Fundamentals of Modern Manufacturing, Mikell P. Groover, John Wiley

& Sons, Inc, New Jersey.

Hsieh, C.H. and Chen, S.H. (1999), “A model and algorithm of fuzzy product positioning”,
Information Sciences, Vol. 121, pp. 61–82.

Irani, Z., Sharif, A., Love, P. E. D. and Kahraman, C. (2002), "Applying concepts of fuzzy
cognitive mapping to model: The IT/IS investment evaluation process", International Journal
of Production Economics, Vol. 75, No. 1–2, pp. 199–211.

Kahraman, C., Onar. S.Ç., and Öztayşi B. (2016), “Interval valued intuitionistic fuzzy
investment analysis: application to CNC lathe selection”, IFAC-PapersOnLine, Vol. 49-12,
pp. 1323–1328.

Kahraman, C. and Öztayşi, B. (2013), “Multicriteria CNC machine selection using type2
fuzzy AHP”, 7. International Academic Conference, September 1-4, 2013 Prague, Czech
Republic.

Li, D., Zeng, W. and Li, J. (2015), “New distance and similarity measures on hesitant fuzzy
sets and their applications in multiple criteria decision making”, Engineering Applications of
 Artificial Intelligence, Vol. 40, pp. 11–16.

Liang, D. and Xu, Z. (2017), “The new extension of TOPSIS method for multiple criteria
decision making with hesitant Pythagorean fuzzy sets”, Applied Soft Computing, Vol. 60, pp.
167–179.

Lin, Z. C. and Yang, C. B. (1996), “Evaluation of machine selection by the AHP method”,
Journal of Materials Processing Technology, Vol. 57, pp. 253-258.

Liu, H. and Rodriguez, R.M. (2014), “A fuzzy envelope for hesitant fuzzy linguistic term set
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

and its application to multi criteria decision making”, Information Sciences, Vol.258, pp. 220–
238.

Liu, P., Mahmood, T. and Khan, Q. (2017), “Multi-attribute decision-making based on

prioritized aggregation operator under hesitant intuitionistic fuzzy linguistic environment”,
Symmetry, Vol. 9, doi:10.3390/sym9110270.

Love, P. E. D., Irani, Z., Standing, C., Lin, C. and Burn, J. M. (2005), "The enigma of
evaluation: benefits, costs and risks of IT in Australian small–medium-sized enterprises",
Information & Management. Vol. 42, No.7, pp. 947–964.

Myint, S. and Tabucanon, M.T. (1994), “A multiple-criteria approach to machine selection

for flexible manufacturing systems”, International Journal of Production Economics, Vol. 33,
pp. 121-131

Oeltjenbruns, H., Kolarik,W.J. and Schnadt-Kirschner, R. (1995), “Strategic planning in

manufacturing systems—AHP application to an equipment replacement decision,”
International Journal of Production Economics, Vol. 38, pp. 189–197.

Otay, Đ., Oztaysi, B., Cevik Onar, S. and Kahraman, C. (2017), “Multi-expert performance
evaluation of healthcare institutions using an integrated intuitionistic fuzzy AHP&DEA
methodology”, Knowledge-Based Systems, Vol. 133, No. 90-106.

Önüt, S., Kara, S. and Efendigil, T. (2008), “A hybrid fuzzy MCDM approach to machine tool
selection”, Journal of Intelligent Manufacturing, Vol. 19, pp. 443-453.

Öztayşi, B. (2015), “A group decision making approach using interval type-2 fuzzy AHP for
enterprise information systems project selection”, Journal of Multiple-Valued Logic and Soft
 Computing, Vol. 24, pp. 475-500.

Öztayşi, B., Çevik Onar, S., Boltürk, E. and Kahraman, C. (2015), “Hesitant fuzzy analytic
hierarchy process”, IEEE Int Conf Fuzzy, pp. 1-7.

Republic of Turkey – Ministry of Economy, Furniture Industry Report, 2016. Available

at: https://www.economy.gov.tr/portal/content/conn/UCM/uuid/dDocName:EK-
021146/ (accessed 12 January 2018).

Rodriguez, R. M., Martinez, L. and Herrera, F. (2012), “Hesitant fuzzy linguistic term sets or
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

decision making”, IEEE Transactions on Fuzzy Systems, Vol. 20, No. 1, pp. 109-119.

Sadeghi, A., Azar, A. and Rad, R. S. (2012), "Developing a fuzzy group AHP model for
prioritizing the factors affecting success of high-tech SME’s in Iran: a case study", Procedia -
Social and Behavioral Sciences. Vol. 62, pp. 957–961.

Saaty, T. L. (1981), The Analytical Hierarchy Process. New York: Mcgraw Hill.

Sun, G., Guan, X., Yi, X. and Zhou, Z. (2018), “Grey relational analysis between hesitant
fuzzy sets with applications to pattern recognition”, Expert Systems with Applications, Vol.
92, pp. 521–532

Susnjara, K. (2006), The New Furniture: How Modern Technology is Changing the Furniture
and Cabinet Industry. Thermwood Corporation, Dale, Indiana.

Taha, Z. and Rostam, S. (2011), “A fuzzy AHP–ANN-based decision support system for
machine tool selection in a flexible manufacturing cell”, The International Journal of
AdvancedManufacturing Technology, Vol. 57, pp. 719–733.

Taha, Z. and Rostam,S. (2012). “A hybrid fuzzy AHP-PROMETHEE decision support system
for machine tool selection in flexible manufacturing cell”, Journal of Intelligent
Manufacturing, Vol. 23, pp. 2137-2149

Taylor, M. and Murphy, A. (2004), “SMEs and e‐business”, Journal of Small Business and
Enterprise Development, Vol. 11, No. 3, pp. 280–289.

The Union Chambers and Commodity Exchanges of Turkey - TOBB (2013), The Turkish
Furniture Products Assembly Sector Report 2013, TOBB Publication No: 2014/212

Tho, N. H., Dawal, S. Z. M., Yusoff, N., Tahriri, F., and Aoyama, H. (2013), “Selecting a
 CNC machine tool using the intuitionistic fuzzy TOPSIS approach for FMC”, Applied
Mechanics and Materials, Vol. 315, pp. 196-205.
Torra, V. (2010), “Hesitant fuzzy sets”, International Journal of Intelligence Systems, Vol.
25, pp. 529–539.

Torra, V. and Narukawa, Y. (2009), “On hesitant fuzzy sets and decision”, In: The 18th IEEE
International Conference on Fuzzy Systems, Jeju Island, Korea, pp. 1378–1382.

United Nations Industrial Development Organization (UNIDO) (2017). “Industrial

Development Report 2018. Demand for Manufacturing: Driving Inclusive and Sustainable
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Industrial Development”. Vienna, available at:

https://www.unido.org/sites/default/files/files/2017-11/IDR2018_FULL%20REPORT.pdf.
(accessed 12 January 2018).

Wang, T. Y., Shaw, C. -F. and Chen, Y. L. (2000), “Machine selection in flexible
manufacturing cell: A fuzzy multiple attribute decision making approach”, International
Journal of Production Research, Vol. 38, No. 9, pp. 2079–209.

Wiedenbeck J. and Parsons, J. (2010), “Digital technology use by companies in the furniture,
cabinet, architectural millwork, and related industries”, Forest Products Journal, Vol. 60, No.
1, pp. 78-85.

Yurdakul, M. (2004), “AHP as a strategic decision-making tool to justify machine tool

selection” Journal of Materials Processing Technology, Vol. 146, pp. 365–37.

Zadeh, L. A. (1965), “Fuzzy sets”, Information and Control, Vol. 8, No. 3, pp. 338-353.

Biographies

Dr. Alper Camci serves as a faculty member in the Engineering Management department of
Bahcesehir University in Istanbul. Prior to joining Bahcesehir University, Dr. Camci served as
an academic director in Arcadia University and faculty member for Okan University
and Franklin University. Dr. Camci completed his Ph.D. at University of Central Florida in
industrial engineering and management systems, MS at Wayne State University and
undergraduate studies at Istanbul Technical University in mechanical engineering. Dr Camci’s
academic interests include project and program management, operations management,
complexity in organizations, technology and innovation management, supply chain
management, organizational transformation and change management.
 Dr. Gül T. Temur graduated from management engineering department of Istanbul
Technical University in 2006 and she completed her Doctor of Philosophy at the same
department. Gül T. Temur has worked as a Research Assistant between 2006-2013 in the
Faculty of Management at Istanbul Technical University (ITU). She has been still working in
Bahcesehir University in the Engineering Management department as assoc. professor. Her
main research interests are “Supply Chain Management”, “Reverse Logistics”, “Decision
Making” and “Artificial Intelligence”.

Dr. Ahmet Beskese is the Chair, and an Associate Professor of Department of Industrial
Engineering, Bahcesehir University (Turkey). He is an industrial engineer and holds higher
degrees in Engineering Management. Having participated in several projects both as a
manager and a team member, and given trainings in various companies, he retains close links
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

with industry, He is a founding member of National Institution of Ergonomics, and

Intellectual Capital Improvement Association. Dr Beskese’s research interests include
strategic management, fuzzy multi attribute decision making, quality management, and
intellectual capital.
 Rank Linguistic Term Abbreviation Triangular Fuzzy Number

10 Absolutely High Importance AHI 7,9,9

9 Very High Importance VHI 5,7,9

8 Essentially High Importance ESHI 3,5,7

7 Weakly High Importance WHI 1,3,5

Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

6 Equally High Importance EHI 1,1,3

5 Exactly Equal EE 1,1,1

4 Equally Low Importance ELI 0.33,1,1

3 Weakly Low Importance WLI 0.2,0.33,1

2 Essentially Low Importance ESLI 0.14,0.2,0.33

1 Very Low Importance VLI 0.11,0.14,0.2

0 Absolutely Low Importance ALI 0.11,0.11,0.14

Table 1. Linguistic Scales for HFAHP (adapted from Öztayşi et al., 2015)

C1 C2 C3 C4
C1 EE Between WHI and VHI Between EE and WHI Between WHI and ESHI
C2 EE Between WLI and EE Between ELI and EHI
C3 EE Between EHI and ESHI
C4 EE
Table 2. Fuzzy Envelops of the Main Criteria Evaluations
 C1 C2 C3 C4
C1 (1,1,1,1) (1,4.77,5.22,9) (1,1,1,5) (1,3,5,7)
C2 (0.11,0.19,0.20,1) (1,1,1,1) (0.2,0.92,1.07,1) (0.33,1,1,3)
C3 (0.2,1,1,1) (1,0.93,1.08,5) (1,1,1,1) (1,2.77,3.22,7)
C4 (0.14,0.2,0.33,1) (0.33,1,1,3) (0.14,0.31,0.36,1) (1,1,1,1)
Table 3. Trapezoidal Fuzzy Sets of Main Criteria
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Main Criterion Weight

C1 (0.11,0.22,0.25,0.47)
C2 (0.03,0.07,0.08,0.15)
C3 (0.07,0.14,0.15,0.27)
C4 (0.03,0.06,0.07,0.15)
Table 4. Trapezoidal Fuzzy Criteria Weights

Main Criterion Crisp Weights

Technical Aspects (C1) 0.452
Personnel Aspects (C2) 0.142
Economic Aspects (C3) 0.273
Vendor Aspects (C4) 0.130
Table 5. The Crisp Values of the Weights of Main Criteria

T1 T2 T3 T4
T1 EE Between EE and WHI Between ELI and EHI Between EHI and WHI
T2 EE Between WLI and EE Between ELI and EHI
T3 EE Between EE and WHI
T4 EE
Table 6. Fuzzy Envelops of the Evaluations of Technical Aspects
 T1 T2 T3 T4
T1 (1,1,1,1) (1,1,1,5) (0.33,1,1,3) (1,1,3,5)
T2 (0.2,1,1,1) (1,1,1,1) (0.2,0.93,1.07,1) (0.33,1,1,3)
T3 (0.33,1,1,3) (1,0.93,1.08,5) (1,1,1,1) (1,1,1,5)
T4 (0.2,0.33,1,1) (0.33,1,1,3) (0.2,1,1,1) (1,1,1,1)
Table 7. Trapezoidal Fuzzy Sets of Technical Aspects
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Sub Criterion Weight

T1 (0.08,0.11,0.15,0.33)
T2 (0.04,0.11,0.11,0.15)
T3 (0.08,0.11,0.11,0.33)
T4 (0.04,0.08,0.11,0.15)
Table 8. Trapezoidal Fuzzy Criteria Weights of Technical Criteria

Weights of Sub Global Weights

Main Criteria
Main Criteria Criteria of Sub Criteria

CNC router construction (T1) 0.141

Technical Process design (T2) 0.096
0.452
Aspects (C1) Design performance criteria (T3) 0.129
CNC computer system criteria (T4) 0.087
Personnel Adaptation of workers (P1) 0.124
0.142
Aspects (C2) Worker technical capability needed (P2) 0.019
Initial cost (E1) 0.180
Economic
0.273
Aspects (C3) Running cost (E2) 0.094

Vendor's CNC experience (V1) 0.019

Vendor
0.130 Vendor's service response capability (V2) 0.056
Aspects (C4)
Reputation of vendor/brand (V3) 0.056
Table 9. The Weights of Main and Sub Criteria
 CNC router construction (T1)
A1 A2 A3
A1
EE Between ELI and EE Between WLI and ELI
A2
EE ELI
A3
EE
Process design (T2)
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

A1 A2 A3
A1
EE Between ELI and EE ELI
A2
EE Between ELI and EE
A3
EE
Design performance criteria (T3)
T3 A1 A2 A3
A1
EE ELI Between WLI and ELI
A2
EE ELI
A3
EE
CNC computer system criteria (T4)
A1 A2 A3
A1
EE EE Between ELI and EE
A2
EE Between ELI and EE
A3
EE
Table 10. Fuzzy Envelopes for Alternatives With Respect to Sub Criteria of Technical Aspects
 Adaptation of workers (P1)
A1 A2 A3
A1
EE Between ELI and EE ELI
A2
EE Between ELI and EE
A3
EE
Worker technical capability needed (P2)
A1 A2 A3
A1
EE Between ELI and EE EE
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

A2
EE Between EE and EHI
A3
EE
Table 11. Fuzzy Envelopes for Alternatives With Respect to Sub Criteria of Personnel Aspects

Initial cost (E1)

A1 A2 A3
A1
EE Between ELI and EE ELI
A2
EE Between ELI and EE
A3
EE
Running cost (E2)
A1 A2 A3
A1
EE Between EE and EHI Between WHI and ESHI
A2
EE WHI
A3
EE
Table 12. Fuzzy Envelopes for Alternatives With Respect to Sub Criteria of Economic Aspects
 Vendor's CNC experience (V1)
A1 A2 A3
A1
EE Between EE and EHI EHI
A2
EE Between EE and EHI
A3
EE
Vendor's service response capability (V2)
A1 A2 A3
A1
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

EE Between ELI and EE Between ELI and EE

A2
EE EE
A3
EE
Reputation of vendor/brand (V3)
A1 A2 A3
A1
EE Between ELI and EE ELI
A2
EE ELI
A3
EE

Table 13. Fuzzy Envelopes for Alternatives With Respect to Sub Criteria of Vendor Aspects

A1 A2 A3
A1 (1,1,1,1) (0.33,1,1,1) (0.2,0.33,1,1)
A2 (1,1,1,3) (1,1,1,1) (0.33,1,1,1)
A3 (1,1,3,5) (1,1,1,3) (1,1,1,1)
Table 14. Trapezoidal Fuzzy Sets for Alternatives In Terms of Technical Aspects
 Global
Weights Crisp Local
Main Criteria Sub Crisp Global Weights
of Weights of
(Weights) Criteria of Alternatives
Sub Alternatives
Criteria
A1 A2 A3 A1 A2 A3
CNC router construction (T1) 0.141 0.249 0.318 0.433 0.035 0.045 0.061
C1) Technical
Process design (T2) 0.096 0.293 0.328 0.379 0.028 0.031 0.036
Downloaded by UNIVERSITY OF TOLEDO LIBRARIES At 17:32 17 June 2018 (PT)

Aspects
Design performance criteria (T3) 0.129 0.249 0.318 0.433 0.032 0.041 0.056
(0.452)
CNC computer system criteria (T4) 0.087 0.308 0.308 0.383 0.027 0.027 0.033
C2) Personnel 0.041 0.047
Adaptation of workers (P1) 0.124 0.293 0.328 0.379 0.036
Aspects
(0.143) Worker technical capability needed (P2) 0.019 0.308 0.383 0.308 0.006 0.007 0.006

C3) Economic Initial cost (E1) 0.180 0.351 0.298 0.351 0.063 0.054 0.063
Aspects
(0.274) Running cost (E2) 0.094 0.472 0.383 0.144 0.044 0.036 0.014

Vendor's CNC experience (V1) 0.019 0.379 0.328 0.293 0.007 0.006 0.005
C4) Vendor
Aspects Vendor's service response capability (V2) 0.056 0.298 0.351 0.351 0.017 0.020 0.020
(0.131)
Reputation of vendor/brand (V3) 0.056 0.293 0.328 0.379 0.016 0.018 0.021
Global Alternative 0.312 0.326 0.362
Weights
Table 15. Global Weights of Alternative CNC Routers In Terms of Sub Criteria

Alternatives Weights
A1 0.311
A2 0.325
A3 0.362
Table 16. Final Scores of Alternatives