Received 21 February 2023, accepted 1 March 2023, date of publication 6 March 2023, date of current version 6 April 2023.

Digital Object Identifier 10.1109/ACCESS.2023.3252566

The Rising Trends of Smart E-Commerce Logistics

HICHAM KALKHA 1, AZEDDINE KHIAT 1, AYOUB BAHNASSE 2, AND HASSAN OUAJJI 1
1 Laboratory Computing, Artificial Intelligence and Cyber Security (2IACS), ENSET of Mohammedia, University Hassan II of Casablanca,
Casablanca 20000, Morocco
2 Laboratory Engineering of Structures, Processes, Intelligent Systems, and Computer Science (ISPS2I), ENSAM of Casablanca, Casablanca 20670, Morocco

Corresponding author: Hicham Kalkha ([email protected])

ABSTRACT Smart Logistics (SL) offers a competitive advantage for e-commerce by utilizing Information
and Communication Technologies (ICT) such as IoT, AI, Blockchain, Cloud computing, 5G, etc. This
technology automates, optimizes, and enables real-time tracking and monitoring of shipments, predicts, and
prevents delays, and optimizes delivery routes and schedules. It also provides greater visibility and control,
allowing e-commerce businesses to react quickly and efficiently to changes in demand or supply. The purpose
of this study is to investigate the impact of digitalization on trade logistics in e-commerce, emphasizing
the significance of smart logistics for the e-commerce industry. We reviewed 288 articles published in the
last decade in the Scopus database to assess the maturity of research in this area. For researchers, this
study provides a better understanding of smart e-commerce logistics and identifies research gaps in the
literature. For e-commerce professionals, it can help them adopt the latest technological trends in their
logistics. Through a systematic literature review and network analysis approach, the study has contributed
by identifying 5 clusters related to ICT application fields in e-commerce and 5 clusters related to important
ICT enablers in smart logistics. We also identified several research gaps and areas for future study, including
the underutilization of computer vision technology and the need for further research on product quality
inspection and accessibility for people with disabilities. Additionally, we suggest exploring the power of
deep learning to solve Vehicle Routing Problems (VRP) and optimizing sensing data volume for minimizing
costs associated with data storage and transfer. This study provides a comprehensive overview of the state
of the art in smart logistics for e-commerce and serves as a guide for future research in this field.

INDEX TERMS Smart logistics, SL, smart e-commerce logistics, ICT, trends, supply chain, information and
communication technologies, artificial intelligence, e-commerce logistics, cloud computing, 5G, blockchain.

I. INTRODUCTION chain management [4]. An intelligent supply chain provides

In recent years, e-commerce retail sales have grown rapidly, e-commerce companies with a competitive advantage by pri-
making it difficult for professionals to keep up with cus- oritizing customer expectations, such as security, personal-
tomer demands and stay competitive in the global market [1] ization, reasonable costs, and on-time delivery (OTD) [5].
Traditional logistics methods have hindered the develop- However, there are currently no systematic reviews that focus
ment of e-businesses, as costs and delays have become on intelligent e-commerce logistics, with most researchers
major issues that customers cannot afford [2] To address only exploring the application of ICT in supply chain man-
this, e-businesses need to enhance their logistics by adopt- agement [1], [4], [6] cloud computing [7], big data and
ing sustainability enablers to increase their agility [3] The artificial intelligence [8], or IoT technology [9]. Recently
emergence of information and communication technology in 2021, Issaoui et al performed a literature analysis of
(ICT) has attracted significant attention from industry and 108 articles in the field of Intelligent Logistics, including
academia worldwide, leading to a major shift towards dig- articles from Springer, IEEE, Science Direct, and Google
ital transformation in many areas, including global supply Scholar databases in this study, authors extensively discussed
the contribution of information technology in logistics, how-
The associate editor coordinating the review of this manuscript and ever, their research does not focus on e-commerce logistics
approving it for publication was Justin Zhang . context [10].
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
VOLUME 11, 2023 For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/ 33839
 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

In this paper, we present a comprehensive exploration of benefits of delivering high-quality results, optimizing analy-
the latest trends in the application of digital technologies sis objectives, and ensuring consistent results. However, both
in e-commerce logistics. Our research aimed to answer two qualitative and quantitative features are combined to evaluate
primary research questions: existing theories.
1) What are the hottest trends in smart logistics (SL) The SLR method is a valuable tool that allows for anal-
literature in e-commerce, and what are the different ysis with minimal bias and summarizes the state-of-the-art
Information and Communication Technologies (ICT) of a specific subject by systematically evaluating relevant
used in e-commerce logistics? findings from recent research studies [17]. Furthermore, SLR
2) How has smart logistics (SL) evolved within enables the selective collection of databases by including the
e-commerce literature since its emergence, and what sequential identification, screening, clustering, and evalua-
are the main countries and international collaborations tion of the dataset [18]. Its replicability as a method-driven
contributing to its growth? approach has motivated the authors to adopt it as a pillar of
this study, in addition to bibliometric analysis. SLR provides
This paper presents a comprehensive analysis of the latest a reliable and robust method that can be readily applied
trends and insights in smart logistics within e-commerce, to a broad field of research to select the most appropriate
aiming to encourage new researchers to contribute to the field contributions.
and to aid practitioners in developing effective e-commerce To answer the second research question, authorship, key-
logistics strategies. Our approach involves a combination word, citation, and co-citation analyses were conducted.
of systematic literature review and bibliometric analysis to The frequency of citations reflects the quality or academic
identify influential researchers, articles, and countries, cooc- influence of articles and authors [19]. The citation net-
currence keywords, and high-impact journals in the field, work analysis enables the identification of the journals and
also to give a full vision of these technologies contribution. papers that have made the most significant contributions to
To achieve these goals, we used the PRISMA framework theory-building in a specific research field.
for paper selection and the VOSviewer software package
for bibliometric and network analysis. The methodology is
detailed in the second section of the paper, and the results A. SCOPE OF THE STUDY
of the SLR and network analysis are presented in the third This study provides literature analysis and insights on smart
section. Finally, the paper concludes with a discussion of the e-commerce logistics and answers the two RQs in the pre-
limitations of the study and suggests future research initia- vious section. There are numerous literature reviews on the
tives in the fourth section. smart logistics context. However, all of them sidelined the
hottest trends of smart logistics for e-commerce. Beyond
II. SYSTEMATIC LITERATURE REVIEW the pandemic covid 19 and the Russian-Ukrainian Crisis in
AND NETWORK ANALYSIS 2022; Cyber-attacks, frauds, and the difficulty of managing
Literature reviews play a crucial role in advancing research in a very large number of orders become a concern for the
positivist approaches [11]. To achieve this, it is necessary to e-retailer who needs to take an urgent decision at every
critically reflect on current positivist practices and carefully stage of the process [20]. Fortunately, technological progress
consider interactions within the supply chain domain [12]. is moving forward at a fast pace, reflected by scientific
One effective method for studying a topic under investigation research in this field. Since our study focuses on the lat-
is systematic literature and network analysis (SLNA) [13]. est trends in information and communication technologies
By combining the methods of systematic literature research (ICT) solutions for E-commerce. Taking into consideration
(SLR) and bibliographic network analysis, as proposed by the accelerated technological revolution in many fields of
Colicchia and Strozzi [14], this methodology is well-suited ICT within the last decade. The rise of artificial intelligence
to the strategy developed for answering research questions in (AI) is notable, with the creation of the chatbot ‘‘Eugene
the previous section. Goostman,’’ which passed the Turing test in 2014 [21], and
The SLNA methodology consists of two steps. In the the release of Microsoft Tay, an AI-powered chatbot that can
first step, visualization and analysis of the bibliographic net- mimic human-like conversations on Twitter in 2016 [22].
work are performed, and the PRISMA framework for paper Deep neural networks have also shown their capabilities,
selection is used to enable readers to assess the quality of as demonstrated by the creation of Alpha-Go, which defeated
methods and validity of findings, as shown in Figure 2 [15]. humans in the game of GO due to increased computational
Instead of the CIMO method used in the SLNA demonstra- power, enabling machine learning algorithms to handle a
tion by Colicchia and Strozzi [14], the PRISMA framework vast number of input variables and make fast and precise
is employed. In the second step, visualization and analy- decisions [23]. The period has also seen significant advance-
sis of the bibliographic network are performed using the ments on the Internet of Things (IoT) technologies, such as
VOSviewer software package, which facilitates bibliometric the invention of the smartwatch with embedded sensors that
analysis and provides clear visualization [16]. By combin- can recognize objects in 2012 [24]. Blockchain-as-a-service
ing these two methodologies, the aim is to maximize the and cloud-based IoT were launched respectively in 2015 and

33840 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 1. Characteristics of the search string.

FIGURE 1. Timeline of ICT innovations (2012-2022).

2016. The appearance of Google Home in 2017 as an ecosys-

tem of sensors and AI to serve humans is an excellent example
of collaboration between AI and IoT [25]. The development
of deep learning frameworks such as Google TensorFlow
in late 2015 and Meta PyTorch in late 2016 has made it
effortless to develop deep learning models, without start-
ing from scratch. Google Tensor Processing Units (TPUs)
have been available since 2018 [26]. The emergence of 5G,
which officially launched in 2019, is expected to activate
the full potential of IoT and cloud computing, with extended LIMIT-TO (PUBYEAR, 2020) OR LIMIT-TO (PUBYEAR,
coverage, lower latency, and connection density of massive 2019) OR LIMIT-TO (PUBYEAR, 2018) OR LIMIT-TO
bandwidth. The ICT technology revolution of the last decade, (PUBYEAR, 2017) OR LIMIT-TO (PUBYEAR, 2016) OR
as shown in Figure 1, motivated our team to investigate LIMIT-TO (PUBYEAR, 2015) OR LIMIT-TO (PUBYEAR,
literature on this topic exclusively during this period, and this 2014) OR LIMIT-TO (PUBYEAR, 2013) OR LIMIµT-TO
paper focuses on literature from 2012 to 2022. (PUBYEAR, 2012)) AND (LIMIT-TO (LANGUAGE,
Our systematic literature review (SLR) focused on collect- ‘‘English’’))
ing datasets from the Scopus database, which provides almost To ensure the accuracy of our dataset, we categorized
60% more coverage than the Web of Science database [27]. the identified records into three groups based on their rel-
Scopus is known for its extensive and authoritative con- evance to the topic, following the approach suggested by
tent, which includes records from reputable publishers such Woschank et al. [18]. To obtain clear research results, we car-
as Elsevier, Springer, and IEEE, instead of mass publi- ried out a two-step screening process. In the first step,
cations [28]. Given the comprehensive nature of Scopus, we examined the title and abstract of the studies to eliminate
we determined that an additional review of other databases the least relevant records. In the second step, we conducted
would not significantly contribute to our study. To ensure that a thorough reading of the full text to identify articles that
our SLR was valuable to users, we followed the PRISMA provide significant contributions to ICT in smart e-commerce
method 2020 proposed by Matthew et al. to capture all rel- logistics, while excluding those that only discuss general
evant information [29] Figure 2. To identify relevant litera- topics or inspire literature reviews.
ture, we used a keyword approach that crisscrossed keywords
related to the five emerging ICT areas (AI, IoT, Cloud Com- B. LITERATURE SURVEY APPROACH
puting, Blockchain, 5G) and Smart Logistics with keywords This study used a systematic literature review and
related to e-commerce. To further limit our search to this bibliometric analysis to investigate the application of digital
field, we screened article titles, abstracts, and keywords in technologies in e-commerce logistics from 2012 to 2022. The
the subject areas of ‘‘engineering,’’ ‘‘business, management, systematic literature review analyzed 48 papers to identify the
and accounting.’’ We limited our search to conference papers, latest trends in smart logistics literature, specifically explor-
conference reviews, articles, or reviews to ensure that only ing the most interesting uses of cloud computing, blockchain,
high-quality studies were considered. The final meta-search AI, IoT, and 5G in smart logistics and measuring their
query was formulated as follows: TABLE 1: TITLE-ABS- impact on e-commerce businesses. The bibliometric analysis
KEY (IoT OR AI OR ‘‘Artificial Intelligence’’ OR ‘‘Internet examined 288 selected papers to investigate the growth and
of Things’’ OR Blockchain OR ‘‘Cloud Computing’’ OR 5G evolution of smart logistics in e-commerce literature, with
OR ‘‘Smart Logistics’’ AND e-Commerce) AND (LIMIT- the aim of quantifying the research in this field. It analyzed
TO ( DOCTYPE, ‘‘cp’’) OR LIMIT-TO (DOCTYPE, ‘‘ar’’) how smart logistics has evolved within e-commerce literature
OR LIMIT-TO (DOCTYPE, ‘‘cr’’) OR LIMIT-TO (DOC- since its emergence, and identified the main countries, influ-
TYPE, ‘‘re’’)) AND (LIMIT-TO (SUBJAREA, ‘‘ENGI’’) ential researchers, and articles in this field, also cooccurrence
OR LIMIT-TO ( SUBJAREA, ‘‘BUSI’’)) AND (LIMIT-TO network analysis to identify potential collaboration between
(PUBYEAR, 2022) OR LIMIT-TO (PUBYEAR, 2021) OR these technologies to empower the logistics capabilities.

VOLUME 11, 2023 33841

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

FIGURE 2. Dataset collection map (PRISMA).

III. FINDINGS AND DISCUSSION

A. RQ1: WHAT ARE THE HOTTEST TRENDS OF SMART
E-COMMERCE LOGISTICS IN THE LITERATURE
To advance research in the field, the authors focused on
48 articles that made significant contributions and conducted
a thorough analysis of each application area along the value
stream. Most of these contributions were in the form of case
studies, as shown in Figure 3. Through this analysis, the
authors identified five clusters related to the use of digi-
tal technologies in e-commerce logistics. These clusters are
described in the following paragraphs and summarized in FIGURE 3. Classification of the dataset by type of study.
Table 2.

1) PROCUREMENT promoting factor is forecast reliability and digitalization of

Online retail requires special agility toward demand varia- the procurement process [78]. Only 13% of the identified
tions, seasonal fluctuations, the bull-wipe effect, and stock- articles were assigned to the cluster ‘‘Procurement’’ appli-
less policy. To increase profitability and avoid shortfall or cation field. 2 papers are targeting demand forecasting, and
over-storage by ensuring an optimal inventory turnover. The the others are targeting Negotiation Competency. Although

33842 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 2. Clustering of the identified studies.

some works dealt with this topic, this application area is still packaging. . . ) [79]. Operational efficiency and proactivity are
an emerging research field with much need for further inves- the key factors for this important link in the e-commerce value
tigation and case study applications especially in demand stream. The warehouse is a large field of automation, route
forecasting, as it is one of the most important pillars in e- optimization, and digitalization. A total of 23% of the identi-
commerce businesses. e.g., to deal with the bull-wipe effect fied articles were assigned to the cluster ‘‘Warehousing’’ most
to ensure tangible benefits across the whole supply chain. of these articles dealt with reliability because reliable and
Demand forecasting (2 contributions): Terrada et al. pro- real-time data is a key competitive advantage for e-commerce
posed a demand forecasting system based on deep learning businesses.
to deal with regression problems. Long Short-Term Memory Inventory reliability (5 contributions): In e-commerce
(LSTM) shows a better result compared to ARIMA in this logistics, hazards impact the schedule smoothness, and the
case study [30]. The use of neural networks for this purpose risk of human errors in manual tasks increases consequently,
is also confirmed by Aci et al. in their study, who proposed reliable inventory and real-time data are the keys to business
LSTM as an effective model for demand forecasting among success. In this context, Kalkha et al. proposed a new concept
26 algorithms of DL and ML [31]. of picking areas, equipped with load cells to communicate
Negotiation Competency (4 contributions): More et al. was real-time inventory to reinforce RFID solutions for small
one of the very first researchers to propose an agent-based items (SKUs) in picking areas enhancing inventory reliability
negotiation system using cloud computing for an effective and procurement visibility to avoid inventory shortage [43].
e-negotiation with better security and low resource con- Shouborno et al. suggest a new concept of moving conveyors
sumption [34]. Vij et al. Propose rule-based and case-based that sort e-commerce items with the help of sensors to gain
reasoning algorithms for bilateral and multilateral accurate more precision for small e-commerce systems to eliminate
E-Negotiation with win-win strategy [33] in the same field as the risk of placing a wrong entry in packaging and pre-
other contexts, Simkova and Smutny proposed a digital form dict inventory shortage [40] The Warehouse environment is
of negotiation (E-Negotiation) based on Genetic Algorithm exposed to a set of constraints (luminosity variation, dust. . . )
(GA) method for online dispute resolution more practical and the activity in many of e-commerce warehouses and
solution to replace traditional e-mail negotiation [32]. Jiang Hubs is extended to night shift. Chen et al. worked on the
et al. Proposed a privacy-preserving business protocol based problem of QR codes in uneven illumination in warehouse
on private smart contracts (Blockchain) in the negotiation automatic sorting systems. They proposed Otsu’s method to
phase [35]. binarize the QR code. The detailed process of binarization
The Procurement cluster in online retail requires agility is to use the window selected adaptively to divide the QR
towards demand variations and stockless policies. It is an code image. Then, each piece of image is binarized by the
emerging research field with a need for investigation, espe- Otsu method and combined sequentially to reconstruct the
cially in demand forecasting. Two articles propose demand QR code. The method can eliminate the impact of the uneven
forecasting systems based on deep learning and neural net- illumination of the QR code effectively [39] The ensure
works, respectively. The other articles focus on Negotiation full-time data transfer under adverse conditions, Emil et al.
Competency, with four contributions proposing agent-based, 2021 proposed a management system based on 5G networks
rule-based and case-based reasoning algorithms, digital that can be used more efficiently for an IoT-enabled supply
forms of negotiation, and privacy-preserving business proto- chain, this paper proposed a system for proactively using NS
cols based on private smart contracts. to adapt the 5G network elements along a Smart Logistics
distribution chain. 5G communication technology features
2) WAREHOUSING ultralow latency, high-speed broadband, and massive access,
Inventory management has got a strategic place in the enabler of cloud computing and big data solutions for logis-
e-commerce value stream. It has a strong link with pro- tics [58]. Guo et al. 2021 studied the acquisition, storage,
curement and demand forecasting. The main warehousing and transfer of sensory data from sensors in e-commerce,
activities are receiving goods, storage, and orders prepa- and they propose a storage method for massive sensing
ration including some value-added tasks (picking, kitting, data processing, which makes full use of the IoT storage

VOLUME 11, 2023 33843

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

resources. The authors Compare the localization effect using optimization and text clustering, a model based on LSTM,
6 algorithms. AHSL0 has a better result among the 5 algo- and a conceptual framework that integrates IoT in WMS.
rithms tested in the experience of trajectory tracing and
forecasting [37]. 3) MARKETING
Order Picking Route Optimization (3 contributions): To sell via e-commerce, there must be customers and there-
Xu et al. proposed an algorithm based on Ant Colony fore bring them to your store website. The goal is to increase
Optimization (ACO) to improve the data picking system; the traffic to your e-commerce (number of visits), encour-
their experiment shows that the optimized picking path is age the visitor to navigate your site (bounce rate), and con-
8,34 % shorter than the WMS route [42]. In the same context vert the visitor into a customer (conversion rate). ICT had
Xin et al. 2019 proposed an algorithm based on text cluster- manifested in selecting niches, identifying needs, and pro-
ing and correlation analysis called BTC for storage location- viding customized content to visitors privately. Marketing is a
allocation, this algorithm determines the storage location of major strength of e-commerce compared to brick-and-mortar.
SKUs in the warehouse, and the positions of related prod- Many studies on ICT have been conducted to unlock the full
ucts (items usually shipped together) are close to each other. potential of digital marketing. A total of 23% of the identified
Which reduces the total order picking distance and improves articles were assigned to the cluster ‘‘Marketing’’ Our sample
the efficiency of warehouse operation. The algorithm clusters covered several digital marketing types. Researchers should
SKUs using text analysis, and the clusters undergo a correla- focus more on product marketing and pricing accuracy to get
tion leading to SKU location-allocation. BTC shows a better exceptional performance.
result compared to (COL) policy and the class-based storage Strategic marketing (2 contributions): Zhao et al. pro-
location (CBSL) policy to allocate the storage location [41]. posed a model based on blockchain and LSTM to analyze
Issaoui et al. proposed (LSTM) model to optimize resource and treat customer feedback transparently. Using blockchain
allocation and enhance task scheduling in a smart logistics for data storage, and long short-term memory (LSTM) net-
framework. The approach based on the LSTM model is more work to mine reviews data for emotional tendencies analysis.
convenient for real-time resource allocation (response time LSTM better understands text data such as reviews containing
50s) and the proposed approach is suitable for real-time idioms. Regulators handle problematic users based on the
scheduling [38]. information recorded in the blockchain. The user receives
Minimize inventory cost (3 contributions): To minimize feedback through smart contact [47] Minjing et al. Proposed
the overall inventory cost. Preil et al. proposed The Monte Ant Colony Algorithm to predict customer intentions; Ants
Carlo tree search (MCTS) to tradeoff between backorders represent users, and pheromones represent user intent. User
(inventory shortage) and extra storage cost approach by elim- intents of browsing, collecting, cart shopping, and purchasing
inating the bullwhip effect. The objective function focuses behavior are obtained from ant responses to pheromones.
on minimizing overall supply chain costs, by absorbing the The algorithm shows a better accuracy compared to Neural
bullwhip effect. MCTS Showed a better result compared to networks (NN) [54].
reinforcement learning (RL) and Genetic Algorithm (GA) Operational Marketing (7 contributions): Zhang et al.
by analyzing lead times and customer demand dataset [36]. proposed a probabilistic generative model that recognizes
Wanganoo to Proposed a conceptual framework to integrate customers’ intentions from time and geographical data using
IoT in WMS in the context of 3PL e-commerce to reduce a machine learning algorithm. This solution is appropriate
inventory cost and time in the return management process for recommendation to present the right product/service cat-
by enhancing the real-time visibility of Warehouse oper- egories to the right people at the right time and right loca-
ations in Reverse logistics, using cloud IoT (Internet of tions [56]. Hanafi et al. Built a recommendation model based
Things) and RFID, the system connects WMS and TMS for on the classification method using CNN for classification
reversed logistics [44]. Zhang et al. proposed a new concept of issues, combined with a probabilistic method using Matrix
smart durable packaging management. Using sustainable and Factorization (PMF) for regression to ensure prediction by
cost-effective packages for logistics tasks distribution in e- the proposed recommendation model [52]. Lou proposed an
commerce. Clustering customers’ orders and optimizing the accurate and faster recommendation system in E-commerce
box fulfillment. In the context of physical internet PI and 3PL based on collaborative filtering and cloud computing [48]
e-commerce increase logistics efficiency [45]. Bosri et al. Proposed a recommendation system through the
Operational efficiency and proactivity are vital in this integration of AI and blockchain. artificial intelligence (AI)
area, and there is a growing trend towards automation, processes user data without taking his consent. Blockchain
route optimization, and digitalization. Reliable and real-time allows AI to perform permissionless data storing and rec-
data is essential for successful e-commerce businesses. The ommendation sending is done cryptographically to ensure
most significant focus in warehousing is inventory reliability, privacy [50]. Duong et al. used a machine learning approach
where several proposals aim to minimize human error and (SVM and Random Forest) to predict the gender of viewers
prevent inventory shortages. Additionally, there are efforts to by investigating categories/product features, they hypothesize
optimize order picking routes and minimize inventory costs. that the relation between categories/products viewed during
These proposals include algorithms based on ant colony a single session may reflect the gender of the viewer [55].

33844 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

Koehnet al. proposed an RNN-based sequence classifiers to Guo et al. proposed corresponding techniques and meth-
predict user conversions with better accuracy. They have ods to achieve product and transaction traceability based on
developed a framework to measure the profitability of a blockchain to Combat counterfeit e-commerce, by scanning a
conversion classifier. GRU performs better than LSTM to code to verify the origin of the product, companies can build
predict customer conversion [51]. Nursetyo et al. propose its traceability system or can chose to cooperate with third
an intelligent chatbot system based on Artificial Intelligence parties [62]. Vanneschi et al. proposed an algorithm based
Markup Language (AIML) smarter, faster, and more accurate on genetic programming (GP) to prevent customers’ defaults.
chatbot. Input questions from users will be processed through The application of credit scoring (CS) is employed to identify
three stages, namely, parsing, pattern matching, and crawling customers’ default probability. the GP’s performance is com-
data using AIML. It can answer all user requests well, with an pared to that of three state-of-the-art). The GP model shows
average response time of 3.4 seconds which is significantly a higher discriminatory power compared to machine learning
fast [53]. methods, specifically logistic regression (LR), support vector
Product marketing (1 contribution): Rangu et al. identify machines (SVMs), and boosted trees (BTs) [67] Mishra et al.
issues with a product through reviews and comments and Use compared extant Fraud detection techniques and their mer-
text mining SVM to detect a quality problem with a product its and demerits and proposed LR-k-fold machine learning
throw customer review [46]. technique to obtaining data from various sources on the cloud
Pricing accuracy (1 contribution): Gao et al. proposed and IoT are preceptors. ML trains and infer to decide whether
KNN-BPNN algorithm optimized with (PSO) to predict cus- fraud or a safe transaction [60]. Ali et al. Compared 3 Deep
tomer purchases to optimize pricing and increase profitabil- learning models ‘‘NN, NB, and GAN’’ for Fraud detection.
ity. the swarm optimization algorithm is used to optimize the The overall performance of the neural network (NN)is better
binomial function coefficient and the K-value of the nearest than GAN and NB for Fraud detection. Considering GAN to
neighbor algorithm, to predict customer purchase intention over-sample minority class in the data by generating example
with the highest accuracy of 94,2% [49]. instances from noise value and combining these examples
Researchers have proposed various solutions for opera- with the training set to train a binary classifier, and train NN
tional marketing, including recommendation systems based using the backpropagation algorithm [66]. Chen et al. pro-
on machine learning, classification methods, and chatbots. posed an encryption algorithm for the e-commerce payment
Strategic marketing solutions have also been proposed, such method to increase security using blockchain [61].
as using blockchain and LSTM for customer feedback analy- Security and Accessibility (4 contributions): Tiwari and
sis and predicting customer intentions using the Ant Colony Sharma proposed an e-commerce framework based on the
Algorithm. One study used text mining SVM to detect quality cloud model, offering many advantages such as Scalability,
issues with products through customer reviews. Another pro- Interoperability, Security and Trust, Redundancy in Cloud
posed an algorithm optimized with PSO to predict customer Services, Cost Reduction, and Speed [57] Narayan et al. Pro-
purchases for pricing optimization. Overall, ICT has provided posed a new concept of smart cart with malpractice detection
various tools for e-commerce businesses to improve their features and other miscellaneous auxiliary functions such as
marketing strategies and increase profitability. admin database, customer following mechanism, and a rec-
ommendation and payment system. They implement this cart
4) SALES AND AFTER-SALES SERVICE using Arduino, a smart cart with three verification methods
Nowadays, the success of e-commerce sales depends mostly of the product: Pi-cam, load cell, and RFID System. A pay-
on IT advancement to enhance and support downstream logis- ment link/dynamic QR code with the total price is created
tics as an important part of the value stream. That good which can be used for payment [80]. Sohaib et al. propose a
feedback matters. Many studies have discussed the benefits framework for the integration of IoT and cloud computing to
of adopting ICT to improve customer experience by adapt- integrate the potential of disabled people in the context of e-
ing the platform content to their needs; safety, accessibility, commerce. The integration of IoT and cloud computing can
interactivity. . . etc. A total of 27% of the identified articles enable smart devices of sensing data stream to support many
were assigned to the cluster ‘‘Sales and after-sales service’’ disabled people for online shopping in a reliable manner with
dealing with traceability, security, and sales forecasting. thus, the help of interceptors such as RFID, VR, and other smart
an important focus on treatability and security as they are the devices connected to cloud e-commerce platforms [68] Cloud
major brake on the development of e-commerce. computing demonstrated a large field of application not only
Traceability and Anti-counterfeit (6 contributions): Gai in security but also in investigations. Guimaraes et al. pro-
et al. propose a novel scheme named Anti-Counterfeit Deter- posed a model to investigate user satisfaction and its impact
ministic Prediction Model based on MCM Monte Carlo on loyalty with the test of user satisfaction and conversion
Model to address counterfeit problems in e-commerce. the into customer loyalty using SaaS applications [70].
proposed algorithm can examine the correlations between Sales forecasting (3 contributions): Chen et al. propose
two features of the products to find out the products that the Neural Network forecasting model, using TensorFlow
do not match the benchmark and detect malicious infor- and Kera’s frameworks to train their model and experiments
mation [69]. With the advancement of blockchain research, show the proposed NN model performs better than the Linear

VOLUME 11, 2023 33845

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

Regression algorithm and SVM algorithm. They utilized VRDP for different size instances and different customer
shape to interpret and evaluate its performance [64] Pacella location distributions, for large instances they’ve got an aver-
et al. Proposed a demand forecast model based on bidirec- age of 30% cost-savings on VRPD compared to VRP with the
tional LSTM (BLSTM), which can process data in both direc- truck only [77]. Gu et al. proposed depot clustering and com-
tions with two separate hidden layers. Augments the accuracy bination with MDVRP Multi-Depot-VRP based on artificial
of the baseline LSTM [63]. Vallés-Pérez et al. develop three bee colony algorithm (ABC) algorithm for route optimiza-
alternatives to tackle the problem of forecasting sales with tion. By degrading MDVRP to multiple single VRP problems
precision (day/store/item) using deep learning techniques. with depot clustering and combination. The proposed algo-
Seq-to-Seq encoder, decoder, and transformer architecture rithm achieves up to 70% advantage over the Greedy Algo-
capable of solving the problem of sales forecasting. they have rithm and a 3% advantage over the Genetic Algorithm [76].
also introduced random max time steps that allowed to train Liu et al. proposed VRP based on Ant Colony Optimization
the model to adapt to different time steps, to avoid training with improved default parameters. ACO was improved to suit
the model in every use [59]. last-mile distribution by modifying the heuristic information
The use of ICT in e-commerce has greatly improved down- [74]. Wanganoo et al. improve visibility and optimize last
stream logistics and customer experience. In the ‘‘Sales and mile delivery parcels in e-commerce by integrating technolo-
after-sales service’’ cluster, 27% of the identified articles gies such as GPS, IoT, and TMS which will provide the deliv-
focused on traceability, security, and sales forecasting. Sev- ery operator with real-time visibility. The server will notify
eral contributions have been made in these areas. In terms of the customer of containing GPS location of the parcel and
traceability and anti-counterfeit, researchers have proposed the QR code to unlock the parcel automatically [73]. Issaoui
algorithms based on Monte Carlo models, blockchain, and et al. Proposed a prediction model-based LSTM followed by
genetic programming to detect malicious information and an optimization model based ACO-PSO to Reduce delivery
prevent fraud. In terms of security and accessibility, cloud time and costs in 2 phases: 1st Predict the delivery time using
computing has been used to provide advantages such as scal- LSTM in 4 steps collection, update, processing, and predic-
ability, interoperability, and redundancy, and to investigate tion of routes or real-time anomalies. 2nd phase they optimize
user satisfaction and its impact on loyalty. In terms of sales the model a combination of Ant Colony Optimization (ACO)
forecasting, researchers have proposed neural network fore- and Particle Swarm Optimization (PSO) is proposed instead
casting models and demand forecast models based on LSTM of Dijkstra for more accurate optimization [75]. Information
and transformer architectures, to forecast sales with preci- and Communication Technology (ICT) has made significant
sion. Overall, ICT has contributed significantly to improving contributions to shipping and reverse logistics. In the cluster
sales and after-sales services in e-commerce. of ‘‘Shipping and Reverse logistics,’’ 15% of identified arti-
cles were focused on security and route optimization. In terms
5) SHIPPING AND REVERSE LOGISTICS of security enhancement, blockchain to enhance drone secu-
Shipping orders from a warehouse to a set of customers and rity and smart contracts to provide transparency in the return
taking decisions with multiple constraints require a certain process have been proposed. In terms of route optimization,
level of accuracy and real-time data sharing. The promot- various algorithms such as ACO, MDVRP, and LSTM-ACO-
ing factors in this physical link with a customer are time, PSO have been proposed to solve the VRPD and reduce deliv-
quality, and cost. A total of 15% of the identified articles ery time and costs for last mile delivery. These algorithms
were assigned to the cluster ‘‘Shipping and Reverse logistics’’ have shown significant performance improvements over tra-
dealing with security, and Route optimization. Transport is ditional approaches, resulting in up to 70% cost savings for
highly affecting price, and it is an important element to be large VRPD instances. Technologies like GPS, IoT, and TMS
optimized. have also been integrated to improve visibility and optimize
Security enhancement (2 contributions): Ossamah et al the last-mile delivery process for e-commerce. Overall, ICT
proposed enhanced security with blockchain for Drones. has played a crucial role in optimizing shipping and reverse
Blockchain uses the peer-to-peer (P2P), Which is solid and logistics by providing real-time data sharing, enhancing secu-
secure. Each peer in the P2P network avails part of its com- rity, and improving route optimization.
puting resources such as disk storage and processing power, In addition to upstream, downstream, and internal logistics,
to other participants in the network. This decentralized infor- we also examine the role of marketing and sales which are
mation element is used to secure drones, Digital Fingerprint, closely linked to logistics and benefit from the advantages
Consensus Mechanisms, and access Control [72] Shih et al. of information and communication technology (ICT). Our
Proposed 8 algorithms for every step in reverse logistics table presents an overview of the different technologies used
based on ETH smart contracts to provide a certain degree of in smart logistics, revealing a surprising lack of research on
transparency in the return process [71]. product quality with AI, despite the vast potential of IoT in
Route optimization (5 contributions): Huang et al. pro- this area. We also note a low focus on accessibility, with
posed an ACO Ant colony Optimization algorithm to solve only one article proposing a solution for people with dis-
the VRPD showing better performance compared to tradi- abilities, despite the significant market share they represent.
tional VRP for trucks, ACO is used to effectively solve the AI emerges as the most used technology in smart logistics,

33846 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 3. Clustering of the identified studies.

FIGURE 4. Number of papers published in Scopus database ‘before 2022.’

with 32 articles in our dataset highlighting its potential to the last four years. As of the time of writing (June 2022),
enhance data reliability and operational efficiency. Collabo- We have eliminated the data of 2022 in order, not to fake
ration between AI and other ICTs, such as IoT, blockchain, results. The number of publications for both e-commerce
and cloud computing, is also explored to enhance efficiency, logistics and ICT was the highest. Demonstrating strong indi-
security, and accessibility, respectively. While 5G technology cations of scholar production and interest in ICT and logistics.
is still in its early stages, its potential impact on smart logistics Smart logistics (SL) evolved slowly within ICT and logistics
connectivity is already being studied. Operational efficiency literature in the first decade. The topic of SL exhibited a
and security are the primary areas of focus in current ICT gradual increase with ups and down in the second decade.
research, as shown in Table 3, detailed in Appendix B. We suppose that the significant growth of smart logistics pub-
lications (+795%) is due to the need for powerful logistics
B. HOW HAS SMART LOGISTICS EVOLVED WITHIN to satisfy customer demand in a globally competitive market
E-COMMERCE LITERATURE SINCE ITS EMERGENCE? with internet facilities TABLE 5 Requiring smart logistics
We are primarily interested in utilizing statistics to provide for e-commerce with the technological background to tackle
an answer to this question by analyzing the development of the weaknesses of brick-and-mortar logistics. The number of
research in the areas of ICT, e-commerce, logistics, and smart publications has doubled in the last 3 years.
logistics applications for e-commerce using descriptive statis- TABLE 5 displays the 6 journals with the greatest impact
tics. Our analysis covered the periods 2012-2022, and we in publishing smart logistics articles. Combined, these 6 jour-
found that there has been a significant research lag in the field nals account for 172 articles, making up 60% of the total
of Smart Logistics for e-commerce, as shown in Figure 4. 288 articles in the study, and receiving 81% of the total
This discovery has motivated our team to explore this topic citations. Elsevier Ltd ranks first in the list, with 15 published
further. We observed that there were only 2565 articles related papers and a total of 705 citations, while the Institute of
to this field in the Scopus database since 1998, whereas there Electrical and Electronics Engineers (IEEE) Inc. is the most
were much larger numbers of articles for ICT (238282 articles prolific, having published 78 articles. Springer published
since 1934) and e-commerce research (89566 articles since 65 articles and Elsevier published 4 high-quality articles.
1929) in the same database. Most of the papers are in the form of articles and conference
In the first decade (2000-2010), the research in E- papers, as detailed in TABLE 7.
commerce and logistics has been growing insignificantly in In TABLE 8, the top 10 countries in terms of research
terms of quantity, overpassing the ICT. However, in the sec- productivity in Smart Logistics (SL) for e-commerce are
ond decade (2010–2020) The number of publications in ICT presented, with China having the highest number of published
followed an exponential trend, reflecting the global interest in papers, followed by India and the USA. These three countries
this world, whereas the E-commerce and logistics one keeps not only have high productivity, but they are also among
growing with ups and downs, with an apparent increase in the most collaborated co-author countries, leading research

VOLUME 11, 2023 33847

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 4. The evolution of research in smart logistics ‘2012-2022’.

TABLE 5. Most frequently cited journals ‘2012-2022’.

TABLE 6. Classification of papers by type of studies. TABLE 8. Top 10 countries working in this research field ‘2012-2022’.

TABLE 7. Top 10 influencing scholars in this research field ‘2012-2022’.

along with ‘‘supply chain’’ ranking second. The keywords

‘‘blockchain’’ and ‘‘IoT’’ follow closely, highlighting the
significance of these technologies in the field of smart logis-
tics. Scholars consider IoT a groundbreaking technology for
automating physical processes [13], [81]. In e-commerce,
trust is a critical issue, and researchers have proposed inte-
grating blockchain technologies such as ‘‘smart contracts’’ in
in Asia, Australia, Europe, and America. This indicates that the fintech context to ensure safety and privacy [35], [82].
SL research for e-commerce is being conducted worldwide, Other keywords ranked in order are artificial intelligence,
with a concentration of research in developing countries like big data, cloud computing, machine learning, data mining,
India and China, reflecting their focus on ICT and global recommender systems, advanced analytics, information man-
e-commerce. It should be noted that only papers published agement, learning systems, and decision making. These tech-
in English were included in this study, and no African or nologies underscore the importance of data science in the
Arabian affiliations were represented. The most influential logistics context of e-commerce, where data accuracy is a key
researchers are also listed in TABLE 6. success factor. In the highly competitive market, proactive
TABLE displays the most frequently used keywords in matching of customer needs with the right product at the right
smart logistics research, with ‘‘Electronic commerce’’ being time and place at a reasonable cost is critical to achieving high
the most used due to its relevance to the study’s framework, sales volume and benefits.

33848 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 9. Top 20 frequent keywords in the SL research. • The storage layer is responsible for data security storage.
Guo et al. proposed a storage method for processing
massive sensing data that optimizes storage and packet
size during transfer [37].
• The application layer serves as the interface between
the IoT device and the network, processing data glob-
ally or partially using APIs in cloud computing cases.
By integrating a machine learning algorithm, an intelli-
gent sorting conveyor was proposed to avoid the risk of
placing wrong entries in packaging and predict inventory
shortages [40].

Researchers propose collaboration between ICT technolo-

gies, such as machine learning, IoT, and 5G, to enhance
management systems and ensure precision in both physical
and information flow. A new concept revealed in this cluster
is the ‘‘Internet of Robotic Things’’ (IoRT), Esenogho et al.
believe that the Massive Internet of Things (MIoT) will play
a significant role in the development of the 5G/6G network
infrastructure, integrating communication infrastructure, arti-
ficial intelligence (AI), and Internet of Things (IoT) support.
In other words, MIoT will enable the connection of a massive
number of devices, allowing for real-time data collection,
analysis, and decision-making [83].
In this cluster, one possible area of improvement could
be the development of more advanced and efficient per-
ception technologies for data collection, such as using
advanced sensors and computer vision techniques. Addi-
The authors utilized network analysis with keyword co-
tionally, there could be research on developing more
occurrences to identify research foci by generating clus-
robust and scalable storage and data processing meth-
ters [1]. To conduct the analysis, all keywords were extracted
ods to handle the large amount of data generated by
from each paper and preprocessed, replacing redundancies
IoT systems. Another area of improvement could be
with their conventional semantics. The minimum number of
exploring the use of blockchain technology to enhance
keyword occurrences was set at 5 using Vosviewer. The anal-
data security and prevent fraudulent activities in logistics
ysis yielded five distinct clusters, as presented in FIGURE 5.
automation.
To represent each cluster, the most frequent keywords were
Cluster 2 ‘Personalization This cluster focuses on the
retained and summarized in TABLE 10.
application of Artificial Intelligence to Customer Relation-
Cluster 1 ‘Automation’ Many researchers have proposed
ship Management (CRM). The primary keywords in this clus-
systems based on Internet of Things (IoT) technologies to
ter are electronic commerce, artificial intelligence, and sales.
enhance logistics automation and enable logisticians to make
Researchers have proposed using artificial intelligence and
proactive decisions. The basic architecture of an IoT system
data mining to create intelligent systems for marketing, such
includes four layers [10]:
as recommenders’ systems that use collaborative filtering to
• The perception layer is responsible for collecting data, increase sales [48]. Other proposed applications of AI in this
such as through RFID technology. Various systems have field include predicting customer behavior to optimize Qual-
been developed using this technology, such as intelligent ity, Cost, and Time and improve profitability [84], as well as
shelves, which can obtain real-time data in warehouses using deep learning algorithms to predict demand with high
and ensure inventory reliability [43]. Narayan et al. pro- accuracy [30]. Additionally, AI can be used to reduce storage
posed an intelligent cart that utilizes load cells, pi-cam, costs [36]. Overall, the goal of this cluster is to enhance oper-
and other sensors to reduce payment malpractice and ational efficiency and increase customer satisfaction. there
improve the billing process [65]. could be research on developing more advanced AI algo-
• The network layer is responsible for forwarding the rithms for customer behavior prediction and personalization,
packets obtained from the perception layer to the inter- as well as investigating the potential of using natural language
net. This layer includes keywords such as 5G mobile processing and chatbots to improve customer interaction and
communication systems, which improve the speed and engagement. Additionally, there could be research on devel-
reliability of data transfer and enhance automated oping more efficient and accurate recommender systems for
control [58]. product recommendations.

VOLUME 11, 2023 33849

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

FIGURE 5. Keyword co-occurrences map with Vosviewer.

TABLE 10. keyword co-occurrences clusters.

Cluster 3 ‘Efficiency’ This cluster is closely related to or discussion of its potential to replace cloud-based data
Cluster 2, and it highlights the potential for a robust and storage. Similarly, there is a lack of discussion about the
intelligent CRM system to reduce various costs. Moreover, role of AI in fraud detection, which could enhance infor-
cloud computing is regarded as a more efficient means of mation transparency, secure payment, and prevent customer
processing large amounts of data generated and collected defaults. However, this cluster does contain keywords related
through e-commerce supply chain networks, utilizing cloud to crime and counterfeit, which highlights the importance of
resources [57]. The cluster emphasizes the need for organi- customers’ trust in e-commerce. One potential research gap
zations to explore opportunities for technological innovation in this cluster is the lack of discussion on the integration
to optimize transportation costs in supply and shipping by of blockchain technology for data storage and its potential
integrating intelligent VRP solutions based on deep learning. to enhance information transparency, secure payment, and
Additionally, efficient resource allocation and task schedul- prevent fraud in e-commerce. Additionally, there is a focus
ing in e-commerce are important considerations. The cluster on crime and counterfeit, but more research could be done on
also notes research on economic order quantity by AI and the specific types of fraud that are prevalent in e-commerce
optimizing routes using deep learning for VRP. one possible and how AI can be utilized to detect and prevent them.
area of improvement could be developing more advanced Furthermore, while there is a mention of customers’ trust in
optimization algorithms for logistics and transportation, such e-commerce, more research could be done on how trust can
as using genetic algorithms or swarm intelligence to improve be built and maintained in the digital marketplace.
resource allocation and task scheduling. Additionally, there Cluster 5 ‘Digitalization’ The digitalization of financial
could be research on developing more efficient and scalable transactions in e-commerce holds the promise of a better
cloud computing methods for processing large amounts of future for trading, and this cluster focuses on some keywords
data in e-commerce supply chain networks. that reflect this trend. These include sustainable logistics and
Cluster 4 ‘Privacy and safety’ It is noteworthy that in the semantic web, which may be explained by the fact that the
this cluster, there is no mention of the keyword ‘‘blockchain’’ next generation of the web, based on blockchain technology,

33850 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

FIGURE 6. Intelligent e-commerce scenario.

TABLE 11. Keyword co-occurrences clusters.

is expected to revolutionize e-commerce by solving issues smart logistics research are ‘‘Electronic commerce,’’ ‘‘supply
related to trust and traceability. while there is mention of chain,’’ ‘‘blockchain,’’ and ‘‘IoT,’’ highlighting the signifi-
the semantic web and sustainable logistics, more research cance of these technologies in the field of smart logistics.
could be done on the specific ways in which blockchain The most cited journal in smart logistics is Elsevier Ltd, with
technology and digitalization can revolutionize e-commerce, 15 published papers and a total of 705 citations. The study
especially in terms of enhancing traceability and supply chain also lists the most influential researchers in smart logistics.
transparency. Additionally, research could be conducted on Some of the most frequently cited influencers include Haiyan
the potential drawbacks and challenges associated with the Wang, Chunguang Li, and Xiaohua Liu.
increased digitalization of financial transactions, such as the Our network analysis identifies five clusters of key-
risk of cyber-attacks and the potential for increased financial words cooccurrence related to the application of technology
exclusion. in e-commerce logistics, marketing, operations, and secu-
In conclusion, our bibliometric analysis shows a significant rity. Overall, the study demonstrates the potential for tech-
growth of smart logistics publications (+795%) is due to the nological innovations to enhance efficiency and customer
need for powerful logistics to satisfy customer demand in a satisfaction in e-commerce, while also highlighting the
globally competitive market with internet facilities, requir- importance of trust and security in these systems. Through
ing smart logistics for e-commerce with the technological this network analysis we suggest several areas of improve-
background to tackle the weaknesses of brick-and-mortar ment and research for each cluster, such as developing
logistics. China has the highest number of published papers more advanced perception technologies, improving recom-
in smart logistics, followed by India and the USA, indicating mender systems and customer behavior prediction algo-
that SL research for e-commerce is being conducted world- rithms, exploring the use of blockchain technology to
wide, with a concentration of research in developing coun- enhance data security, developing more advanced opti-
tries like India and China, reflecting their focus on ICT and mization algorithms for logistics and transportation, inves-
global e-commerce. The most frequently used keywords in tigating how trust can be built and maintained in the

VOLUME 11, 2023 33851

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 12. Keyword co-occurrences clusters.

33852 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

TABLE 12. (Continued.) Keyword co-occurrences clusters.

digital marketplace, and researching the potential draw- contributions, with no consideration given to existing liter-
backs and challenges associated with increased digitaliza- ature reviews. Furthermore, our dataset was limited to the
tion. Figure 6 présents an intelligent e-commerce logistics Scopus database, and we only analyzed data from 2012-2022,
scenario. in order to identify trends. It is important to note that some
papers were excluded from our analysis due to copyright
IV. LIMITATION AND FUTURE RESEARCH restrictions or unavailability.
This study is subject to certain limitations. Firstly, the scope Our discussion highlighted several areas for future
of our systematic literature review was limited to significant research. We found that blockchain technology not only

VOLUME 11, 2023 33853

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

improves transaction security, but can also enhance physi- APPENDIX A

cal security, such as drone security. We suggest extending See Table 11.
this solution to other logistics nodes, such as parcel lock-
ers. Although AI has successfully solved many optimization APPENDIX B
problems, deep learning techniques, such as LSTM, have See Table 12.
not been fully utilized in optimization, and are still mostly
limited to forecasting. To address this, we suggest exploring REFERENCES
the power of deep learning to solve Vehicle Routing Prob-
[1] A. Rejeb, S. Simske, K. Rejeb, H. Treiblmaier, and S. Zailani, ‘‘Internet of
lems (VRP). In addition, we found that computer vision Things research in supply chain management and logistics: A bibliomet-
technology is underuitilized, and the quality of product veri- ric analysis,’’ Internet Things, vol. 12, Dec. 2020, Art. no. 100318, doi:
fication and inspection needs further attention from scholars, 10.1016/j.iot.2020.100318.
[2] B. Lin, J. Collins Cole, and R. K. Su, Supply Chain Costing: An Activity-
as it has implications for customer trust. Finally, we recom- Based Perspective. Bingley, U.K.: MCB Univ. Press, 2001. [Online]. Avail-
mend optimizing sensing data volume and relying on deep able: http://www.emerald-library.com/ft
learning for interpretation and completion of missing data [3] H. Balouei Jamkhaneh, R. Shahin, and G. L. Tortorella, ‘‘Analysis of
logistics 4.0 service quality and its sustainability enabler scenarios in
with the help of time series forecasting, in order to minimize emerging economy,’’ Cleaner Logistics Supply Chain, vol. 4, Jul. 2022,
costs associated with data storage and transfer. Art. no. 100053, doi: 10.1016/J.CLSCN.2022.100053.
[4] Z. Inamdar, R. Raut, V. S. Narwane, B. Gardas, B. Narkhede, and
M. Sagnak, ‘‘A systematic literature review with bibliometric analysis of
V. CONCLUSION big data analytics adoption from period 2014 to 2018,’’ J. Enterprise Inf.
Our systematic literature review and network analysis of Manage., vol. 34, no. 1, pp. 101–139, Mar. 2020, doi: 10.1108/JEIM-09-
288 research articles from the Scopus database published 2019-0267.
[5] R. Raam and R. Bhandari, ‘‘Impact of technology on logistics and supply
between 2012 and 2022 reveal that the field of smart logistics chain management,’’ IOSR J. Bus. Manag., vol. 2, no. 17, pp. 19–24, 2014.
for e-commerce is still in its early stages and has significant [Online]. Available: http://www.iosrjournals.org
potential for growth. Our analysis highlights the increased [6] Y. Riahi, T. Saikouk, A. Gunasekaran, and I. Badraoui, ‘‘Artificial intel-
ligence applications in supply chain: A descriptive bibliometric analysis
interest in this field of study in recent years, which corre- and future research directions,’’ Expert Syst. Appl., vol. 173, Jul. 2021,
sponds to the growth of e-commerce demand. Art. no. 114702, doi: 10.1016/j.eswa.2021.114702.
In summary, our study focused on the trends and research [7] S. Singh, S. Chaudhary, V. Sharma, A. Gupta, and M. Vatsa, ‘‘Cloud
computing and e-commerce based entrepreneurship,’’ in Proc. 2nd Int.
gaps in the field of smart logistics for e-commerce. Conf. Smart Electron. Commun. (ICOSEC), Oct. 2021, pp. 1468–1474,
Our systematic literature review identified a lack of doi: 10.1109/ICOSEC51865.2021.9591756.
research on product quality inspection with AI and low focus [8] L. Li, ‘‘E-commerce data analysis based on big data and arti-
ficial intelligence,’’ in Proc. Int. Conf. Comput. Netw., Electron.
on accessibility for people with disabilities, despite their Autom. (ICCNEA), Sep. 2019, pp. 133–138, doi: 10.1109/ICCNEA.2019.
significant market share. We found that AI is the most used 00034.
technology in smart logistics, with many articles highlight- [9] S. Mohamed and K. Sethom, ‘‘Incorporating RFID technology and
IOT into a e-commerce: A step toward intelligent purchasing pro-
ing its potential to enhance data reliability and operational cesses,’’ in Proc. 20th Int. Conf. Sci. Techn. Autom. Control Com-
efficiency. Our discussion identified several areas for future put. Eng. (STA), Dec. 2020, pp. 328–332, doi: 10.1109/STA50679.2020.
research, including exploring the power of deep learning in 9329337.
[10] Y. Issaoui, A. Khiat, A. Bahnasse, and H. Ouajji, ‘‘Toward smart logis-
solving Vehicle Routing Problems, improving product veri- tics: Engineering insights and emerging trends,’’ Arch. Comput. Methods
fication and inspection with computer vision technology, the Eng., vol. 28, no. 4, pp. 3183–3210, Jun. 2021, doi: 10.1007/s11831-020-
use of blockchain to enhance physical security and optimiz- 09494-2.
[11] J. T. Mentzer and K. B. Kahn, ‘‘A framework of logistics research,’’ J. Bus.
ing sensing data volume and transfer. Logistics, vol. 16, no. 1, p. 231, 1995.
Additionally, our bibliometric analysis provided insights [12] H. Davis, WebSci ’13: Proceedings of the 3rd Annual ACM Web Science
into the research trends and hot topics in this field, as well Conference, 2013: Paris, France. 2013.
[13] F. Strozzi, C. Colicchia, A. Creazza, and C. Noè, ‘‘Literature review on the
as the most influential researchers and journals. We found ‘smart factory’ concept using bibliometric tools,’’ Int. J. Prod. Res., vol. 55,
that the number of publications in this area is increasing, no. 22, pp. 6572–6591, 2017, doi: 10.1080/00207543.2017.1326643.
particularly in the last four years. AI applications in supply [14] C. Colicchia and F. Strozzi, ‘‘Supply chain risk management:
chains and smart logistics are receiving significant attention, A new methodology for a systematic literature review,’’ Supply
Chain Manag., Int. J., vol. 17, no. 4, pp. 403–418, Jun. 2012, doi:
and new themes such as the emergence of WEB 3.0 and 10.1108/13598541211246558.
the Physical Internet are being explored. This study provides [15] M. D. F. McInnes, D. Moher, B. D. Thombs, and T. A. McGrath, ‘‘Preferred
a roadmap for future researchers to identify research gaps reporting items for a systematic review and meta-analysis of diagnostic test
accuracy studies: The PRISMA-DTA statement,’’ Jama, vol. 319, no. 4,
and contribute to the development of smart logistics for pp. 388–396, Jan. 2018, doi: 10.1001/jama.2017.19163.
e-commerce. [16] N. J. van Eck and L. Waltman, ‘‘Software survey: VOSviewer, a com-
Although our study is subject to certain limitations, such as puter program for bibliometric mapping,’’ Scientometrics, vol. 84, no. 2,
pp. 523–538, Aug. 2010, doi: 10.1007/s11192-009-0146-3.
the exclusion of existing literature reviews and the limitation [17] M. Petticrew and H. Roberts, Systematic Reviews in the Social Sciences: A
of our dataset to the Scopus database, we hope our findings Practical Guide. USA: Wiley, 2008.
will encourage new researchers to contribute to this growing [18] M. Woschank, E. Rauch, and H. Zsifkovits, ‘‘A review of further direc-
tions for artificial intelligence, machine learning, and deep learning in
field of study and to focus on research gaps to further advance smart logistics,’’ Sustainability, vol. 12, no. 9, p. 3760, May 2020, doi:
smart logistics for e-commerce. 10.3390/su12093760.

33854 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

[19] W. Shan and J. Wang, ‘‘Mapping the landscape and evolutions of green sup- [39] R. Chen, Y. Yu, X. Xu, L. Wang, H. Zhao, and H.-Z. Tan, ‘‘Adaptive
ply chain management,’’ Sustainability, vol. 10, no. 3, p. 597, Feb. 2018, binarization of QR code images for fast automatic sorting in warehouse
doi: 10.3390/su10030597. systems,’’ Sensors, vol. 19, no. 24, p. 5466, Dec. 2019, doi: 10.3390/
[20] A. U. Din, H. Han, A. Ariza-Montes, A. Vega-Muñoz, A. Raposo, and s19245466.
S. Mohapatra, ‘‘The impact of COVID-19 on the food supply chain and [40] S. A. I. Shouborno, T. I. Mahmud, N. Ishraq, R. Ali, and
the role of e-commerce for food purchasing,’’ Sustainability, vol. 14, no. 5, T. H. Joy, ‘‘Complete automation of an e-commerce system with
p. 3074, Mar. 2022, doi: 10.3390/su14053074. Internet of Things,’’ in Proc. IEEE Int. Conf. Robot., Automat., Artif.-
[21] D. Aamoth. (2014). Interview With Eugene Goostman, the Fake Intell. Internet–Things (RAAICON), Nov./Dec. 2019, pp. 81–86, doi:
Kid Who Passed the Turing Test. TIME.com. [Online]. Available: 10.1109/RAAICON48939.2019.39.
http://time.com/2847900/eugene-goostman-turing-test/ [41] C. Xin, X. Liu, Y. Deng, and Q. Lang, ‘‘An optimization algorithm
[22] G. Neff and P. Nagy, ‘‘Talking to bots: Symbiotic agency and the case of based on text clustering for warehouse storage location allocation,’’ in
Tay,’’ Int. J. Commun., vol. 10, pp. 4915–4931, Oct. 2016. Proc. 1st Int. Conf. Ind. Artif. Intell. (IAI), Jul. 2019, pp. 1–6, doi:
[23] S. R. Granter, A. H. Beck, and D. J. Papke Jr., ‘‘AlphaGo, deep 10.1109/ICIAI.2019.8850832.
learning, and the future of the human microscopist,’’ Arch. Pathol. [42] W. Xu, C. Y. Shi, H. T. Song, and Y. X. Chen, ‘‘Applied tech-
Lab. Med., vol. 141, no. 5, pp. 619–621, 2017, doi: 10.5858/arpa. nology on improving the order picking efficiency in the area of
2016-0471-ED. EC of China post logistics based on ACO,’’ in Advanced Materials
[24] E. Morganti, L. Angelini, A. Adami, D. Lalanne, L. Lorenzelli, and Research, vol. 859. Switzerland: Trans Tech Publications, 2014, doi:
E. Mugellini, ‘‘A smart watch with embedded sensors to recognize 10.4028/www.scientific.net/AMR.859.486.
objects, grasps and forearm gestures,’’ Proc. Eng., vol. 41, pp. 1169–1175, [43] H. Kalkha, A. Khiat, A. Bahnasse, and H. Ouajji, ‘‘Toward a
Jan. 2012, doi: 10.1016/j.proeng.2012.07.297. reliable and responsive e-commerce with IoT,’’ Proc. Comput.
[25] B. Li, T. N. Sainath, A. Narayanan, and J. Caroselli, ‘‘Acoustic modeling Sci., vol. 198, pp. 614–619, Jan. 2022, doi: 10.1016/J.PROCS.2021.
for Google home,’’ in Proc. Interspeech, 2017, pp. 399–403. 12.295.
[26] B. Varghese, P. Leitner, S. Ray, K. Chard, A. Barker, Y. Elkhatib, [44] L. Wanganoo, ‘‘Streamlining reverse logistics through IoT driven ware-
H. Herry, C.-H. Hong, J. Singer, F. P. Tso, E. Yoneki, and M.-F. Zhani, house management system,’’ in Proc. 8th Int. Conf. Rel., Infocom Technol.
‘‘Cloud futurology,’’ Computer, vol. 52, no. 9, pp. 68–77, Sep. 2019, doi: Optim. (Trends Future Directions) (ICRITO), Jun. 2020, pp. 854–858, doi:
10.1109/MC.2019.2895307. 10.1109/ICRITO48877.2020.9197929.
[27] D. Zhao and A. Strotmann, ‘‘Analysis and visualization of citation net- [45] Y. Zhang, S. Liu, Y. Liu, and R. Li, ‘‘Smart box-enabled product–
works,’’ Synth. Lectures Inf. Concepts, Retr., Services, vol. 7, no. 1, service system for cloud logistics,’’ Int. J. Prod. Res., vol. 54, no. 22,
pp. 1–207, Feb. 2015, doi: 10.2200/s00624ed1v01y201501icr039. pp. 6693–6706, 2016, doi: 10.1080/00207543.2015.1134840.
[28] A. Foncubierta-Rodríguez, H. Müller, and A. Depeursinge, ‘‘Retrieval of [46] C. Rangu, S. Chatterjee, and S. R. Valluru, ‘‘Text mining approach
high-dimensional visual data: Current state, trends and challenges ahead,’’ for product quality enhancement: (Improving product quality through
Multimedia Tools Appl., vol. 69, pp. 539–567, Jan. 2014. [Online]. Avail- machine learning),’’ in Proc. IEEE 7th Int. Advance Comput. Conf. (IACC),
able: http://www.youtube.com/ Jan. 2017, pp. 456–460, doi: 10.1109/IACC.2017.0100.
[29] M. J. Page, J. E. McKenzie, P. M. Bossuyt, I. Boutron, T. C. Hoffmann, [47] Z. Zhao, Z. Hao, G. Wang, D. Mao, B. Zhang, M. Zuo, J. Yen, and G. Tu,
C. D. Mulrow, and L. Shamseer, ‘‘The PRISMA 2020 statement: ‘‘Sentiment analysis of review data using blockchain and LSTM to improve
An updated guideline for reporting systematic reviews,’’ Int. J. Surg., regulation for a sustainable market,’’ J. Theor. Appl. Electron. Commerce
pp. vol. 88, Apr. 2021, Art. no. 105906. Res., vol. 17, no. 1, pp. 1–19, Dec. 2021, doi: 10.3390/jtaer17010001.
[30] L. Terrada, M. E. Khaili, and H. Ouajji, ‘‘Demand forecasting model [48] F. Lou, ‘‘E-commerce recommendation technology based on collabora-
using deep learning methods for supply chain management 4.0,’’ Int. tive filtering algorithm and mobile cloud computing,’’ Wireless Com-
J. Adv. Comput. Sci. Appl., vol. 13, no. 5, pp. 704–711, 2022, doi: mun. Mobile Comput., vol. 2022, Mar. 2022, Art. no. 7321021, doi:
10.14569/IJACSA.2022.0130581. 10.1155/2022/7321021.
[31] M. Aci and G. A. Dogansoy, ‘‘Demand forecasting for e-retail sec- [49] Z. Gao, ‘‘Precision marketing mode of agricultural products e-commerce
tor using machine learning and deep learning methods,’’ J. Fac. Eng. based on KNN algorithm,’’ in Innovative Computing (Lecture Notes in
Archit. Gazi Univ., vol. 37, no. 3, pp. 1325–1339, 2022, doi: 10.17341/ Electrical Engineering), vol. 791. U.K.: Hindawi, 2022, pp. 1151–1158,
gazimmfd.944081. doi: 10.1007/978-981-16-4258-6_141.
[32] N. Simkova and Z. Smutny, ‘‘Business E-NeGotiAtion: A method using [50] R. Bosri, M. S. Rahman, M. Z. A. Bhuiyan, and A. Al Omar, ‘‘Integrating
a genetic algorithm for online dispute resolution in B2B relationships,’’ blockchain with artificial intelligence for privacy-preserving recommender
J. Theor. Appl. Electron. Commerce Res., vol. 16, no. 5, pp. 1186–1216, systems,’’ IEEE Trans. Netw. Sci. Eng., vol. 8, no. 2, pp. 1009–1018,
Mar. 2021, doi: 10.3390/jtaer16050067. Apr. 2021, doi: 10.1109/TNSE.2020.3031179.
[33] S. R. Vij, D. Mukhopadhyay, and D. A. J. Agrawal, ‘‘Automated negoti- [51] D. Koehn, S. Lessmann, and M. Schaal, ‘‘Predicting online shopping
ation in E commerce: Protocol relevance and improvement techniques,’’ behaviour from clickstream data using deep learning,’’ Expert Syst.
Comput., Mater. Continua, vol. 61, no. 3, pp. 1009–1024, 2019, doi: Appl., vol. 150, Jul. 2020, Art. no. 113342, doi: 10.1016/j.eswa.2020.
10.32604/cmc.2019.08417. 113342.
[34] A. More, S. Vij, and D. Mukhopadhyay, ‘‘Agent based negotiation [52] N. S. Hanafi and A. Basari, ‘‘Convolutional-NN and word embedding for
using cloud—An approach in e-commerce,’’ Proc. ICT Crit. Infrastruct., making an effective product recommendation based on enhanced contex-
48th Annu. Conv. Comput. Soc. India (Advances in Intelligent Sys- tual understanding of a product review,’’ Int. J. Adv. Sci., Eng. Inf. Technol.,
tems and Computing), vol. 248. USA: Springer, 2014, pp. 489–496, doi: vol. 9, no. 3, pp. 1063–1070, 2019, doi: 10.18517/ijaseit.9.3.8843.
10.1007/978-3-319-03107-1_54. [53] A. Nursetyo, D. R. I. M. Setiadi, and E. R. Subhiyakto, ‘‘Smart chatbot
[35] Y. Jiang, C. Wang, Y. Wang, and L. Gao, ‘‘A privacy-preserving e- system for e-commerce assitance based on AIML,’’ in Proc. Int. Seminar
commerce system based on the blockchain technology,’’ in Proc. IEEE Res. Inf. Technol. Intell. Syst. (ISRITI), Nov. 2018, pp. 641–645, doi:
Int. Workshop Blockchain Oriented Softw. Eng. (IWBOSE), Feb. 2019, 10.1109/ISRITI.2018.8864349.
pp. 50–55, doi: 10.1109/IWBOSE.2019.8666470. [54] P. Minjing, L. Xinglin, L. Ximing, Z. Mingliang, Z. Xianyong,
[36] D. Preil and M. Krapp, ‘‘Artificial intelligence-based inventory manage- D. Xiangming, and W. Mingfen, ‘‘Recognizing intentions of e-commerce
ment: A Monte Carlo tree search approach,’’ Ann. Oper. Res., vol. 308, consumers based on ant colony optimization simulation,’’ J. Intell. Fuzzy
nos. 1–2, pp. 415–439, Jan. 2022, doi: 10.1007/s10479-021-03935-2. Syst., vol. 33, no. 5, pp. 2687–2697, Oct. 2017, doi: 10.3233/JIFS-169318.
[37] Y. Guo and G. Liang, ‘‘Perceptual feedback mechanism sensor technology [55] D. Duong, H. Tan, and S. Pham, ‘‘Customer gender prediction based
in e-commerce IoT application research,’’ J. Sensors, vol. 2021, pp. 1–12, on e-commerce data,’’ in Proc. 8th Int. Conf. Knowl. Syst. Eng. (KSE),
Sep. 2021, doi: 10.1155/2021/3840103. Oct. 2016, pp. 91–95, doi: 10.1109/KSE.2016.7758035.
[38] Y. Issaoui, A. Khiat, A. Bahnasse, and H. Ouajji, ‘‘An advanced [56] M. Zhang, G. Chen, and Q. Wei, ‘‘Discovering consumers’ purchase inten-
LSTM model for optimal scheduling in smart logistic environment: tions based on mobile search behaviors,’’ in Flexible Query Answering Sys-
E-commerce case,’’ IEEE Access, vol. 9, pp. 126337–126356, 2021, doi: tems 2015, vol. 400. USA: Springer, 2016, pp. 15–28, doi: 10.1007/978-3-
10.1109/ACCESS.2021.3111306. 319-26154-6_2.

VOLUME 11, 2023 33855

 H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

[57] N. Tiwari and N. Sharma, ‘‘A proposed structure, primarily based totally [74] W. Liu, ‘‘Route optimization for last-mile distribution of rural
on cloud computing for enhancing electronic commerce applications,’’ in e-commerce logistics based on ant colony optimization,’’ IEEE
Rising Threats in Expert Applications and Solutions (Lecture Notes in Access, vol. 8, pp. 12179–12187, 2020, doi: 10.1109/ACCESS.2020.
Networks and Systems), vol. 434. USA: Springer, 2022, pp. 307–318, doi: 2964328.
10.1007/978-981-19-1122-4_34. [75] Y. Issaoui, A. Khiat, K. Haricha, A. Bahnasse, and H. Ouajji,
[58] E. J. Khatib and R. Barco, ‘‘Optimization of 5G networks for ‘‘An advanced system to enhance and optimize delivery operations in a
smart logistics,’’ Energies, vol. 14, no. 6, p. 1758, Mar. 2021, doi: smart logistics environment,’’ IEEE Access, vol. 10, pp. 6175–6193, 2022,
10.3390/en14061758. doi: 10.1109/ACCESS.2022.3141311.
[59] I. Vallés-Pérez, E. Soria-Olivas, M. Martínez-Sober, A. J. Serrano- [76] Z. Gu, Y. Zhu, Y. Wang, X. Du, M. Guizani, and Z. Tian, ‘‘Applying
López, J. Gómez-Sanchís, and F. Mateo, ‘‘Approaching sales forecast- artificial bee colony algorithm to the multidepot vehicle routing prob-
ing using recurrent neural networks and transformers,’’ Expert Syst. lem,’’ Softw., Pract. Exper., vol. 52, no. 3, pp. 756–771, Mar. 2022, doi:
Appl., vol. 201, Sep. 2022, Art. no. 116993, doi: 10.1016/j.eswa.2022. 10.1002/spe.2838.
116993. [77] S.-H. Huang, Y.-H. Huang, C. A. Blazquez, and C.-Y. Chen, ‘‘Solving
[60] K. N. Mishra and S. C. Pandey, ‘‘Fraud prediction in smart societies using the vehicle routing problem with drone for delivery services using an ant
logistic regression and k-fold machine learning techniques,’’ Wireless Pers. colony optimization algorithm,’’ Adv. Eng. Informat., vol. 51, Jan. 2022,
Commun., vol. 119, pp. 1341–1367, Feb. 2021, doi: 10.1007/s11277-021- Art. no. 101536, doi: 10.1016/j.aei.2022.101536.
08283-9. [78] H. Patil and B. R. Divekar, ‘‘Inventory management challenges for B2C
[61] T.-C. Chen, Y.-S. Liang, P.-S. Ko, and J.-C. Huang, ‘‘Optimization model e-commerce retailers,’’ Proc. Econ. Finance, vol. 11, no. 14, pp. 561–571,
of cross-border e-commerce payment security by blockchain finance,’’ 2014, doi: 10.1016/s2212-5671(14)00221-4.
Wireless Commun. Mobile Comput., vol. 2021, pp. 1–9, Dec. 2021, doi: [79] K. B. Praveen, P. Kumar, J. Prateek, G. Pragathi, and J. Madhuri, ‘‘Inven-
10.1155/2021/9192219. tory management using machine learning,’’ Int. J. Eng. Res. Technol.,
[62] F. Guo, D. Ma, J. Hu, and L. Zhang, ‘‘Optimized combination of e- vol. 9, no. 6, pp. 866–869, Jul. 2020.
commerce platform sales model and blockchain anti-counterfeit traceabil- [80] A. Narayan, A. Krishnan, and A. S. Ponraj, ‘‘IoT based comprehensive
ity service strategy,’’ IEEE Access, vol. 9, pp. 138082–138105, 2021, doi: retail malpractice detection and payment system,’’ in Proc. IEEE Int. Conf.
10.1109/ACCESS.2021.3117906. Electron., Comput. Commun. Technol. (CONECCT), Jul. 2020, pp. 1–8,
[63] M. Pacella and G. Papadia, ‘‘Evaluation of deep learning with long doi: 10.1109/CONECCT50063.2020.9198519.
short-term memory networks for time series forecasting in supply [81] Y. Yamamoto, T. Kawabe, S. Tsuruta, E. Damiani, A. Yoshitaka,
chain management,’’ Proc. CIRP, vol. 99, pp. 604–609, Jan. 2021, doi: Y. Mizuno, and R. Knauf, ‘‘IoT-aware online shopping system enhanced
10.1016/j.procir.2021.03.081. with gaze analysis,’’ in Proc. World Autom. Congr. (WAC), Jul. 2016,
[64] J. Chen, W. Koju, S. Xu, and Z. Liu, ‘‘Sales forecasting using deep neural pp. 1–6, doi: 10.1109/WAC.2016.7583028.
network and SHAP techniques,’’ in Proc. IEEE 2nd Int. Conf. Big Data, [82] J. Tang, F. Liu, S. Lu, S. Qin, and S. Shao, ‘‘Research on the new gener-
Artif. Intell. Internet Things Eng. (ICBAIE), Mar. 2021, pp. 135–138, doi: ation electronic payment system applied in ‘the belt and road,’’’ in Proc.
10.1109/ICBAIE52039.2021.9389930. Int. Conf. e-commerce Internet Technol. (ECIT), Apr. 2020, pp. 5–9, doi:
[65] A. Narayan, A. Krishnan, and A. S. Ponraj, ‘‘IOT based comprehensive 10.1109/ECIT50008.2020.00009.
retail malpractice detection and payment system,’’ in Proc. IEEE Int. Conf. [83] E. Esenogho, K. Djouani, and A. M. Kurien, ‘‘Integrating artificial intelli-
Electron., Comput. Commun. Technol. (CONECCT), Jul. 2020, pp. 1–8, gence Internet of Things and 5G for next-generation smartgrid: A survey
doi: 10.1109/CONECCT50063.2020.9198519. of trends challenges and prospect,’’ IEEE Access, vol. 10, pp. 4794–4831,
[66] I. Ali, K. Aurangzeb, M. Awais, R. J. Hussen Khan, and S. Aslam, 2022, doi: 10.1109/ACCESS.2022.3140595.
‘‘An efficient credit card fraud detection system using deep-learning [84] Y. Huang and X. Wang, ‘‘Research on the influence of intelligent
based approaches,’’ in Proc. IEEE 23rd Int. Multitopic Conf. recommendation function on user behavior in highly digitized net-
(INMIC), Nov. 2020, pp. 1–6, doi: 10.1109/INMIC50486.2020. work platforms,’’ in Proc. IEEE Int. Conf. Electron. Technol., Com-
9318202. mun. Inf. (ICETCI), Aug. 2021, pp. 250–254, doi: 10.1109/ICETCI53161.
[67] L. Vanneschi, D. M. Horn, M. Castelli, and A. Popovič, ‘‘An 2021.9563603.
artificial intelligence system for predicting customer default in e-
commerce,’’ Expert Syst. Appl., vol. 104, pp. 1–21, Aug. 2018, doi:
10.1016/j.eswa.2018.03.025.
[68] O. Sohaib and K. Kang, ‘‘Internet of Things (IoT) in e-commerce: For
people with disabilities,’’ in Proc. Int. Conf. Electron. Bus. (ICEB),
Jun. 2015, pp. 308–311. [Online]. Available: https://www.scopus.
com/inward/record.uri?eid=2-s2.0-84959216167andpartnerID=
40andmd5=9f59a98364b5ca6dfce8b2d15ec6b912
[69] K. Gai, M. Qiu, H. Zhao, and W. Dai, ‘‘Anti-counterfeit scheme using
Monte Carlo simulation for e-commerce in cloud systems,’’ in Proc. IEEE
2nd Int. Conf. Cyber Secur. Cloud Comput., Nov. 2015, pp. 74–79, doi:
10.1109/CSCloud.2015.75.
[70] T. Guimaraes and K. Paranjape, ‘‘Testing cloud computing for cus-
tomer satisfaction and loyalty,’’ Int. J. Electron. Customer Relationship
Manag., vol. 8, nos. 1–3, pp. 72–86, 2014, doi: 10.1504/IJECRM.2014.
066885.
[71] D.-H. Shih, F.-C. Huang, C.-Y. Chieh, M.-H. Shih, and T.-W. Wu, ‘‘Pre-
HICHAM KALKHA received the degree in engi-
venting return fraud in reverse logistics—A case study of ESPRES solution
neering from ESITH Casablanca, with versatile
by Ethereum,’’ J. Theor. Appl. Electron. Commerce Res., vol. 16, no. 6,
pp. 2170–2191, 2021, doi: 10.3390/jtaer16060121. training in industrial management and logistics.
[72] A. Ossamah, ‘‘Blockchain as a solution to drone cybersecurity,’’ in Proc. He is currently pursuing the Ph.D. degree with
IEEE 6th World Forum Internet Things (WF-IoT), Jun. 2020, pp. 1–9, doi: the Computer Science, Artificial Intelligence, and
10.1109/WF-IoT48130.2020.9221466. Cyber Security Laboratory (L2IACS), Hassan II
[73] L. Wanganoo and A. Patil, ‘‘Preparing for the smart cities: IoT University of Casablanca. He is a Logistics Engi-
enabled last-mile delivery,’’ in Proc. Adv. Sci. Eng. Technol. Int. neer. His research interests include information
Conf. (ASET), Feb./Apr. 2020, pp. 1–6, doi: 10.1109/ASET48392.2020. and communications technology (ICT) and its
9118197. applications in e-commerce logistics.

33856 VOLUME 11, 2023

H. Kalkha et al.: Rising Trends of Smart E-Commerce Logistics

AZEDDINE KHIAT is a Professor Researcher HASSAN OUAJJI received the degree (higher
in the Department Mathematics and Com- education) from the Department of Electronic
puter Science, H.D.R. and Ph.D. in computer and Electrical Engineering, ENSET Mohamme-
science, networks and telecommunications at dia. He is currently a Full Professor with the
ENSET Mohammedia, University Hassan II of Department of Electronic and Electrical Engineer-
Casablanca, Morocco. A researcher member on ing, ENSET Mohammedia. He is also a Research
the Computing, Artificial Intelligence and Cyber Member with the SSDIA Laboratory and the Head
Security laboratory (2IACS), Artificial Intelli- of the Network and Telecommunications Team,
gence, Cybersecurity and Digital Trust team Hassan II University of Casablanca, Morocco.
(IACSCN) at ENSET Mohammedia, University He has published tens of refereed journal and con-
Hassan II of Casablanca, Morocco. Outstanding Reviewer on various ference papers. He is a member of several scientific projects with Hassan II
indexed journals. organizing committee and Technical Program Committee University of Casablanca. His research interests include acoustic electrical
of tens of international conferences. Published tens of refereed journal engineering, signal processing, electronic, networks, and telecommunica-
and conference papers. Her research interests are: Wireless Networks, tions. He is an organizing committee member and a technical program
QoS on mobile networks, Handover on wireless networks, networks, and committee member of tens of international conferences. He is an outstanding
telecommunications, SDN, new generation networks, security, cybersecurity, reviewer of various indexed journals.
mobile learning, IoT, smart city, smart Grid, and industry 4.0. certified
on several products and solutions among them: Fortinet NSE4, Microsoft
Azure Administrator Associate, Juniper Associate, Juniper Cloud Associate,
Huawei RS Associate, Oracle Associate and Red Hat Associate.

AYOUB BAHNASSE received the Ph.D. degree

in networks and telecommunication from Chouaib
Doukkali University, Morocco. He is currently
a Professor and a Researcher with ENSAM
Casablanca, Hassan II University of Casablanca,
Morocco, where he is also a Research Associate
with the Information Treatment Laboratory, Fac-
ulty of Sciences Ben M’sik. He is a member of
the ‘‘Big Data and Internet of Things Project for
Urban Service’’ Project, Hassan II University of
Casablanca. He has published tens of refereed journal and conference papers.
His research interests include new-generation networks, security, mobile
learning, wireless networks, quality of service, multimedia systems, the IoT,
smart city, and MPLS. He is an organizing committee member and a technical
program committee member of tens of international conferences in around
20 countries. He is an Outstanding Reviewer of various indexed journals,
such as Computer Networks (Elsevier), Computers and Electrical Engineer-
ing (Elsevier), Computers and Security (Elsevier), IEEE TRANSACTIONS ON
COMMUNICATIONS, Telecommunication Systems (Springer), Wireless Personal
Communications (Springer), Mobile Networks and Applications (Springer),
and tens of indexed journals. He has been certified in several products and
solutions, among them are Fortinet NSE4, Microsoft Azure Administrator
Associate, Juniper DevOps Associate, Juniper Cloud Associate, Juniper
Design Associate, and Huawei RS Associate.

VOLUME 11, 2023 33857