Paper 1 - Digital Twins

Computers & Industrial Engineering 187 (2024) 109768
Contents lists available at ScienceDirect
Computers & Industrial Engineering

journal homepage: www.elsevier.com/locate/caie
Digital twins for logistics and supply chain systems: Literature review,
conceptual framework, research potential, and practical challenges
Tho V. Le ∗, Ruoling Fan
School of Engineering Technology, Purdue University, West Lafayette, IN 47907, USA
ARTICLE INFO ABSTRACT
Keywords: To facilitate an effective, efficient, transparent, and timely decision-making process as well as to provide
Digital twins guidelines for industry planning and public policy development, a conceptual framework of digital twins (DTs)
Logistics for logistics and supply chain systems (LSCS) is needed. This paper first introduces the background of the
Supply chain
logistics and supply chain industry, the DT and its potential benefits, and the motivations and scope of this
Framework
research. The literature review indicates research and practice gaps and needs that motivate proposing a new
Strategic planning
System
conceptual DT framework for LSCS. As each element of the new framework has different requirements and
goals, it initiates new research opportunities and creates practical implementation challenges. As such, the
future of DT computation involves advanced analytics and modeling techniques to address the new agenda’s
requirements. Finally, ideas on the next steps to deploy a transparent, trustworthy, and resilient DT for LSCS
are presented.
1. Introduction disrupting the global supply chain Lee and Wong (2021). Furthermore,
the Ukraine and Russia war started in February 2022 first created
The last decade has observed rapid technology development, in- worries about the wheat supply chain, but then quickly generated
cluding Industry 4.0, blockchains, 3D printing, new materials, and an energy crisis for European countries which heavily influenced the
others. Many sectors of society are moving to digitize because there supply chain of the area and worldwide (IEA, 2022; OECD, 2022).
could be enormous potential benefits, such as faster access to in- More recently, tension and the trade war between the US and China
formation, improved customer experience, increased productivity and
have created significant changes in manufacturing and business which
efficiency, improved decision-making, lower operating costs, and im-
have led to shifts in the global supply chain towards on-shoring, near-
proved safety (Khan, Khan, & Aftab, 2015; Rosin, Proksch, Stubner,
shoring, and friend-shoring (Mao & Görg, 2020; Steinbock, 2018; Wu,
& Pinkwart, 2020; Sabbagh et al., 2012). However, digitization also
faces challenges, for example, legal issues, a growing labor force gap Wood, Oh, & Jang, 2021; Yuan, Rodrigues, Wang, & Behrens, 2023).
in digital skills, security risks, and a shortage of technological re- Natural disasters, such as wildfires, hurricanes, and floods, that are
sources (Parviainen, Tihinen, Kääriäinen, & Teppola, 2017). Some lo- happening more frequently have also created additional challenges for
gistics and supply chain companies have digitized parts or whole chains supply chain performances (Gunessee & Subramanian, 2020; Ye & Abe,
to manage their business. 2012). To cope with the situation and support local industry, the US
Digital systems are surrounded by uncertainties that create chal- government established Freight Logistics Optimization Works (FLOW)
lenges for stakeholders to make prudent and timely decisions. Stake- in March 2022 with the aim of reducing delivery costs and generating
holders need to be informed of uncertain situations and often need to benefits for the supply chain system (The White House, 2022). Canada
act fast to respond to market challenges. An outstanding example of established the National Supply Chain Task Force 2022 to inform the
uncertainty is the pandemic that started in early 2020 when the Covid- future actions and measures (Government of Canada, 2022). The Euro-
19 disease spread worldwide. The supply chain was disrupted because pean Union has also drafted the Corporate Sustainability Due Diligence
of panic demand (e.g., hoaxed buying) and delay supply (e.g., lack of
Directive (CSDD) (i.e., the European Supply Chain Act) to facilitate
labor force in the manufacturing and supply chain system) (Chowd-
the fairness and sustainable global economy (European Commission,
hury, Paul, Kaisar, & Moktadir, 2021; Guan et al., 2020). In addition
Directorate-General for Justice and Consumers, 2022). These initiatives
to the supply chain crisis during the pandemic, the Ever Given ship
were created for handling a lack of supply chain transparency and
blocked the Suez Canal in Egypt for about a week in March 2021 further
∗ Correspondence to: 401 N. Grant St. Knoy Hall Room 145, West Lafayette, IN 47907, USA.
E-mail address: [email protected] (T.V. Le).
https://doi.org/10.1016/j.cie.2023.109768
Received 15 July 2023; Received in revised form 9 November 2023; Accepted 21 November 2023
Available online 29 November 2023
0360-8352/© 2023 Elsevier Ltd. All rights reserved.
T.V. Le and R. Fan Computers & Industrial Engineering 187 (2024) 109768
to address vulnerability in responding to uncertain human-made and Zhang, 2020), manufacturing (Rosen, Von Wichert, Lo, & Bettenhausen,
natural events. 2015; Schroeder, Steinmetz, Pereira, & Espindola, 2016), informat-
The logistics and supply chain system is becoming more complex ics (Canedo, 2016), and transportation (Schwarz & Wang, 2022; Wang,
because it is a highly connected and inter-operated system of systems. Gupta et al., 2022). DT also has strong potential in the logistics and
For instance, different parts of a Japanese car made in Thailand are supply chain domain since its highly dynamic and closed interaction
supplied by several local factories. Therefore, when a flood occurred in features correspond well to the strengths of DT.
Thailand in 2011, in total, Nissan, Honda, and Toyota lost more than
420,000 cars, and closed their factories for 245 days. Some of these 1.2. Potential benefits of digital twins
factories were not flooded but lacked parts from local suppliers that
were directly affected by the flood (Haraguchi & Lall, 2015). There Establishing a DT could lead to potential benefits to society, econ-
were also consequences and impacts on other factories, such as those omy, business, and environment (Centre for Digital Built Britain, 2022).
in Indonesia, the Philippines, Vietnam, and Malaysia. Another exam- DTs provide opportunities to improve LSCS modeling, simulating, op-
ple of a complex supply chain system is the semiconductor industry. erating, and managing tools by allowing greater interactions between
Semiconductors are the heart of modern electronic devices which are various data sources and models as well as different stakeholders. As
widely used in communications, computing, transportation, manufac- the interactions become intimate, communications among stakeholders
turing, healthcare, and military applications, among others. However, will grow more efficient. Consequently, the operational efficiency of the
most semiconductors are supplied by only a few foundries in Taiwan, LSCS will increase, allowing end-users to receive products quicker and
South Korea, and China. Therefore, the global semiconductors supply in a timely manner (especially for perishable goods and medicines).
chain is vulnerable to an extreme event in that region (e.g., epidemic, Meanwhile, the DT techniques connect virtual representations to the
wildfire, or flood). Complex systems are difficult to render transpar- physical infrastructure and the surrounding environment in near real-
ent. DTs provide an environment for tracking material/product flows; time. The virtual LSCS provides an opportunity to test and experiment
testing solutions; and exposing pre-existing vulnerabilities. DTs allow with various scenarios, enhancing the theoretical and practical basis
stakeholders to streamline operations, create transparency, and build for each decision and minimizing disruption to the real system. The
resilience for LSCS. DT techniques also enhance the understanding of analytical results
and facilitate tracking and tracing the origins of products/materials
1.1. Overview of digital twins (e.g., where do raw materials really come from?) by the end-users.
Therefore, the LSCS becomes more transparent, increases the systems’
Over the course of human history, many have imagined the creation credibility, and lays the foundation for further development. Addition-
of a virtual world that can reflect reality. For example, in 1993, a ally, the DT techniques support predictive maintenance, repair, and
book named Mirror World written by David Gelernter depicted such risk mitigation, and improve sustainability and robustness. As a result,
an indistinguishable world (Gelernter, 1993). Recently, with the rapid the system will become resilient to resist emergencies (e.g., epidemics,
development of technology, humans can make such imagination come natural disasters, and damages), which is vital to maintain the effective
true. For the Apollo 13 mission, the National Aeronautics and Space and efficient operations of LSCS in the context of global trending and
Administration (NASA) first used 15 simulators to train astronauts and economic fluctuation.
mission controllers. NASA initiated the Digital Twins (DTs) concept Furthermore, DT will boost the operational excellence and human
which the aerospace domain later adopted as a key element in the 2010 excellence in LSCS. Operational excellence in LSCS refers to the sys-
technology roadmap (Singh et al., 2021). In reality, DT techniques were tems’ capacity to maintain consistent and resilient operations in the
used to simulate the actual spacecraft conditions, explore solutions, and face of unforeseen risks, fluctuations, and challenges. It implies that the
predict results, which were important in planning, designing, operation, systems ensure competitive and productive procedures in dynamic and
and management (Allen, 2021). uncertain environments (Eisenhardt & Martin, 2000; Mishra, Singh, &
DT techniques have been defined in multiple ways. The common Subramanian, 2022; Sabahi & Parast, 2020; Sandberg, 2021). Human
definition of DT is a virtual representation of physical assets (Glaess- excellence in LSCS, on the other hand, refers to the stakeholders and
gen & Stargel, 2012; Lee, Lapira, Bagheri, & Kao, 2013). NASA first managers in LSCS sectors who are able to develop deliberate policies
introduced it as ‘‘An integrated multi-physics, multi-scale, probabilistic and utilize human resources to allocate responsibilities so that the
simulation of a vehicle or systems that uses the best available physical systems operate efficiently (Rahman, Hamid, Sahin, Wang, & Islam,
models, sensor updates, fleet history, etc., to mirror the life of its flying 2022). Operational excellence and human excellence have a close
twin.’’ (Machining4.0, 2023). Later, researchers proposed alternative reciprocal relationship as human excellence is the foundation of op-
definitions of DT: ‘‘The digital twin is a set of virtual information constructs eration excellence and operation excellence can boost human excellent
that fully describes a potential or actual physical manufactured product from development (Maré, 2004). DT implementations facilitate employees
the micro atomic level to the macro geometrical level’’ (Grieves & Vickers, in different parts of the system working with high efficiency that es-
2017). Our paper adopted the universal definition provided by IBM: sentially comes to high-quality decisions to achieve human excellence.
‘‘A digital twin is a virtual representation of an object or system that spans Consequently, the LSCS will be in effective and efficient operations and
its lifecycle, is updated from real-time data, and uses simulation, machine will achieve operational excellence.
learning, and reasoning to help decision-making ’’ (IBM, 2022). DT can also
be used for designing, modeling, visualizing, testing, and implementing 1.3. Motivations and research goals
new ideas without disrupting the current process.
DT has three basic layers (Far & Rad, 2022; Lv, Qiao, Li, Yuan, The planning, design, and implementation of a DT face challenges
& Wang, 2022; Zheng, Lu, & Kiritsis, 2022): (i) the physical layer and barriers. For instance, protection for the business secrets of data
contains all the real information, including data, decisions, actions, etc.; providers (Ha & Tran, 2023; Moshood, Nawanir, Sorooshian, & Ok-
(ii) the communication layer contains the data transmission tools, pro- falisa, 2021), computation time and storage (Abideen, Sundram, Pye-
cesses for converting the real information to machine-readable informa- man, Othman, & Sorooshian, 2021; Liu, Yeoh, Qu, & Gao, 2022),
tion and vice versa; and (iii) the digital layer includes computational establishment of a common platform for data and information ex-
and simulation techniques to process information, analysis, calcula- change (Cirullies & Schwede, 2021; Lyovin & Efimova, 2017), data
tion, and predictions. DT has applications in many different fields format standardization, requirement of a large amount of data (Car-
including aeronautics and space (Glaessgen & Stargel, 2012; Tuegel, lan & Vanelslander, 2021), complex integration, and cyber security
2012), robotics (Schluse & Rossmann, 2016; Wang, Jiao, Wang, & concerns (Del Giorgio Solfa, 2022; Liu et al., 2022), among others.
The paper "Digital Twins for Logistics and Supply Chain Systems: Literature Review, Conceptual Framework, Research Potential, and Practical Challenges" presents a comprehensive exploration
of the application of digital twin technology in the context of logistics and supply chain management.
2
The study begins with a thorough literature review, synthesizing existing research on digital twins and their relevance to logistics and supply chain systems. This review likely covers various
aspects, including the definition of digital twins, their evolution, and their potential benefits in optimizing supply chain operations.
Building on the literature review, the paper likely proposes a conceptual framework that outlines how digital twins can be applied within logistics and supply chain systems. This framework may
encompass the key components of digital twins, such as data acquisition, modeling and simulation, real-time monitoring, and decision support.
Furthermore, the paper likely explores the research potential of digital twins in logistics and supply chain management. This could involve identifying areas where digital twin technology can offer
significant improvements, such as inventory management, predictive maintenance, or demand forecasting.
Fig. 1. Outline of research scope.
Such barriers and challenges not only impede the way for the de- ‘digital twin logistics’, or ‘digital twin supply chain’, or ‘digital twin
velopment of DT in LSCS but are also an obstacle to operational technique’, or ‘digital logistics’, or ‘digital supply chain’. On Google
excellent and human excellence development. The stakeholders could Scholar, for each keyword search, the first 1000 papers were screened,
not make effective decisions under insufficient communication, limited resulting in a total of 6000 papers. Fig. 2 shows the paper selection
information-sharing systems, or the unavailability of the relevant infor- process under different topics following the PRISMA approach (Page
mation at the right place and time in the right amount. Such delays in et al., 2021). From Google Scholar and Web of Science databases under
decisions could weaken the performance of the LSCS (Rahman et al., the six pre-mentioned topics, the total identified papers were 28,884,
2022). To overcome those challenges and barriers, there is a need to as of Apr. 1st, 2023. Before the screening, 302 papers were identified
review the state of the art and practice of efforts to build a DT for LSCS. as duplicated and were excluded from the databases. Upon a cursory
There is also a need to identify gaps in the DT development efforts review, 19,714 papers were excluded as they were not related to the
to develop a framework, identify research opportunities and current logistics and supply chain domain (e.g., they are about equipment
practical challenges, and identify technologies needed. building, machine operation monitoring, etc.), and a subset of papers
This research provides a comprehensive and systematic review of did not employ DT techniques. Even though these papers were omit-
DT for LSCS. It focuses primarily on developing a DT framework for ted, some of them present new research approaches relevant to DTs,
the deployment of a transparent, trustworthy, and resilient DT for therefore, they are cited in the research opportunities sections. Since
LSCS. This paper identifies research and practical opportunities for the this paper focuses on the system level, 7973 papers were excluded for
logistics and supply chain industry and for academic research. only concentrating only on specific parts of the system, for example,
manufacturing or processing. On checking the keywords and abstracts
1.4. Scope and paper organization for simultaneous mention of ‘digital twin’ and ’supply chain’, or ‘digital
twin’ and ‘logistics’, the pool of papers was narrowed down to 34
This research includes four topics which are depicted in Fig. 1. papers. In addition, the connectedpapers.com website was utilized to
The literature review indicates state-of-the-art and -practice gaps and obtain additional relevant papers connected to the 34 papers as well
needs that essentially motivate the development of a new DT framework as papers connected to the additional relevant papers themselves. As a
for LSCS. In addition, since each element of the new framework has result, additional 14 relevant papers were identified and incorporated
different requirements and goals, it initiates new research opportunities into the review. Therefore, a total of 48 papers were examined in
and creates practical implementation challenges. As such, the future of DT detailed analyses.
computation ties up with advanced analytics and modeling techniques
to address the new agenda.
2.2. Literature summary and analysis
The remainder of this paper is organized as follows. Section two
presents the literature review. Section three envisions a future DT and
Table 1 presents 48 papers about DTs for LSCS. Each paper is
develops a DT framework for LSCS. Section four discusses the research
assessed according to Scope, Domain, Objective, Method/technique, Focus
and practical foci of elements of the suggested DT framework. Section 5
problem, Dataset, Validation, and Online/offline. The abbreviations used
discusses the future of DT computation. Section six suggests the next
in the table are explained below it.
steps in building a transparent, trustworthy, and resilient DT for LSCS.
The Scope elucidates the primary purpose of each paper which is
The final section concludes the research.
divided into four categories: Framework, Implementation, a combina-
2. Literature review tion of both Framework and Implementation, and Review. The Domain
represents the specific area or field to which the paper is related. It
This section provides a review of the literature (journal papers, is noteworthy that within the principal application domains of supply
white papers, industry reports, etc.) to support the development of a chain and logistics, there are only 3 papers focusing on maritime
DT framework for LSCS. Only papers that address DTs for logistics or delivery, suggesting an area for future research.
supply chain issues at the system level are included in this review paper Except for the review papers, other papers’ Objective are classi-
as it focuses on DT strategic planning. A strategic planning ‘‘is used to fied into Design, Model, and Theory. For papers that are not re-
set priorities, focus energy and resources, strengthen operations, ensure views, most incorporate Computer Science or Operations Research
that employees and other stakeholders are working towards common methods/techniques, which is expected given that DT is inherently an
goals, establish agreement around intended outcomes/results, and as- AI and data-driven area.
sess and adjust the organization’s direction in response to a changing The Focus Problem highlights the main issue addressed in the paper.
environment’’ (Balanced Scorecard Institute, 2023). The system level Some papers aiming at developing frameworks or implementations
refers to the interrelationships and interactions of multiple elements employ data to validate the work, hence, the Dataset column shows the
of either DT logistics or DT supply chain systems. The effective and size of the dataset used in the paper. Validation refers to the methods
efficient DT operations could be improved by adopting techniques at used to validate the findings or results of the study. Online/offline
the system level, thereby boosting the development of the logistics and indicates whether the simulation was conducted in an online or offline
supply chain industry. environment.
In summary, as can be observed from Table 1, about half of the
2.1. Literature identifications efforts (25/48 papers) are either to develop DT frameworks or DT
frameworks and implementation, which apply to urban delivery, mar-
Six keywords were defined and selected to perform the search on itime delivery, or general logistics and supply chain fields. Except for
Google Scholar and Web of Science (WoS): ‘digital twin system’, or literature review papers, nearly all papers adopt operations research
3
Fig. 2. Identification of studies via Google Scholar, Web of Science, and Connectedpaper.com.
and data-driven techniques, like optimization, big data, simulation, for validation whereas the majority of papers utilize generative data
and artificial intelligence. With the progress of science and technology, sets. This revealed a potential shortfall in the practical implementation
the methodologies and techniques applied to DT in LSCS have un- of DTs for LSCS.
dergone significant innovations, moving from traditional approaches,
such as optimization, cyber–physical, and data mining, to AI-related 2.3. The trend
techniques, such as deep reinforcement learning, deep learning, sim-
ulation, and advanced optimization algorithms. Such advancements Fig. 3 presents a summary of literature papers by year. The blue
expand the capabilities and benefits of DT in LSCS and enhance the bars illustrate the publication frequency.
systems’ working efficiency. The 48 papers are chronologically and This graphical representation indicates a rising trend in DT research
alphabetically classified into different clusters of methodologies to concerning LSCS, which demonstrates the escalating importance and
investigate emerging and commonly used technologies and techniques interest in this area.
as well as the methodology gaps in the next section. Additionally, there
are a few papers on fresh foods, fruits, and perishable products but very 2.4. The current DT methods
limited research on DTs for LSCS of other essential and high-priority
areas, such as basic goods and services for physiology and safety For a deeper analysis of the methodology employed in the literature,
needs (based on Maslow’s Hierarchy of Needs) (e.g., medicine, shelter, VOSviewer is adopted. In VOSviewer, there are three ways to collect
energy) and semiconductors. Limited research is on conceptualized keywords: Author Keywords, All Keywords, and Keywords Plus. Author
DT for infrastructure (data, communication, computing, and storage), Keywords refer to the keywords that are defined by the authors when
vehicles, trucking drivers, bottleneck identifications, risk mitigation, publishing. All Keywords are the keyword occurrences in the given
cybersecurity, and network resilience. There is also a lack of studies articles and references in WoS database. Keywords Plus is the keyword
on implementation roadmap; literature on DT predictive maintenance, label provided by the WoS platforms.
repair, and management strategies; and methods for assessing DT op- Fig. 4 represents 61 Author Keywords in descending order of their
erations and benefits. There are few efforts to propose standards, Total Link Strength which is the ‘‘total strength of the co-authorship links
protocols, and strategies for integrating current systems, such as En- of a given researcher with other researchers’’ in WOS database (van Eck
terprise Resource Planning (ERP), Warehouse Management Systems & Waltman, 2017). All Author Keywords were derived from the Web
(WMS), and Transportation Management Systems (TMS), to create a of Science, and their total link strength values were calculated using
comprehensive DT. Finally, only a few papers employ actual data sets VOSviewer.
Finally, the paper likely addresses the practical challenges associated with implementing digital twins in logistics and supply chain systems. These challenges may include data integration issues,
privacy and security concerns, scalability, and the need for interdisciplinary collaboration.
4
Overall, the paper provides a comprehensive overview of digital twins in logistics and supply chain management, offering insights into their theoretical foundations, practical applications, research
opportunities, and challenges that need to be addressed for successful implementation.
Table 1
Summary of studies focus on DT in logistics and supply chain systems.
Paper Scope Domain Objective Method/Technique Focus Problem Dataset Validation On-
line/Offline
Cichosz et al. (2018) RE LO – – Summarized some changes in the logistics – – –
market when applying DT and other
digitization tools
Defraeye et al. (2019) IMP SC MOD Simulation Designed a DT fruits model on simulation to 24 days CS OFF
optimize the logistics process, reducing waste.
Gorodetsky, Kozhevnikov, FW LO DS Cyber-physical Designed a multi-agent system to enhance the – – OFF
Novichkov, and Skobelev multi-agent system operations in uncertain and dynamic
(2019) environments, which could be applied to smart
logistics or supply chain systems
Barykin, Bochkarev, RE SC – – Summarized the DT technique in supply chain – – –
Kalinina, and Yadykin management to reduce the operational risks
(2020) and risks caused by force majeure.
Jeong, Flores-Garcia, and FW LO DS – Presented design processes of DT in production – – –
Wiktorsson (2020) logistics, containing 5 dimensions of DT
Lind et al. (2020) RE MD MOD Optimization Used optimization techniques to improve the – – ONL
arrangement of the maritime ports to improve
the supply chain efficiency
Marmolejo-Saucedo (2020) FW LO MOD Simulation Developed decisions based on the DT to robust – CS OFF
the business process.
Yang et al. (2020) FW SC MOD Cyber-physical Introduced a CPS-based integration platform for – CS ONL
system global DT system in automotive parts suppliers
Wang, Wang, Yang, Zhu, BO SC DS Data mining Used big data-driven and DT techniques to – CS OFF
and Liu (2020) address the problem in the remanufacturing:
like the uncertainty reduction and lack of
innovative enabling technology applications.
Abideen et al. (2021) RE LO – – Summarized the papers related to the – – –
reinforcement learning application and
provided a research framework
Belfadel, Hörl, Tapia, and FW LO MOD Adaptive modeling Last Mile Urban Delivery – CS OFF
Puchinger (2021)
Bolender, Bürvenich, FW SC MOD Heuristic Designed an adaptive framework based on the 20 cases CS OFF
Dalibor, Rumpe, and optimization DT to reduce the manufacturing time and
Wortmann (2021) avoid waste
Busse et al. (2021) FW SC MOD Multimodel and Developed a model enabled improvements, – SIM OFF
intermodel simulations, and evaluations for the whole
supply chain
Chen and Huang (2021) RE SC – – Summarized the DT applied to remanufacture – – –
supply chain systems, including the challenges
and opportunities
Dai, Zhang, Maharjan, and BO SC MOD Deep reinforcement Applied DT to the IIoT to reduce the energy 2000 EXP OFF
Zhang (2021) learning consumption and improve data processing episodes
efficiency
Gutierrez-Franco, FW LO DS Reinforcement Provided a framework based on reinforcement 50 Nodes CS OFF
Mejia-Argueta, and Rabelo learning learning to maintain the sustainability of
(2021) last-mile logistics
Kegenbekov and Jackson FW SC DS Reinforcement Used some smart agents under reinforcement 30 periods EXP OFF
(2021) learning learning algorithms maintain the (modeling
demand-supply balance in the supply chain days)
system
Moshood et al. (2021) RE LO – – Summarized the papers about applying DT to – –
deal with the issues of the supply chain in
logistics
Nguyen, Tran, Thomassey, IMP SC MON Deep Learning Applied two data-driven methods for decisions 20 631 EXP OFF
and Hamad (2021) making in supply chain management
Pan, Zhou, Piramuthu, RE SC – – Summarized some new techniques in the – – –
Giannikas, and Chen sustainability of the smart city, containing the
(2021) supply chain and logistics
Park, Son, Ko, and Noh FW SC MOD Reinforcement Applied DT and reinforcement learning will 10 samples CS OFF
(2021) learning decide the ‘‘what-next’’ and ‘‘where-next’’ with
highly robust and efficiency
Rudskoy, Ilin, and FW SC MOD ArchiMate notation Designed a reference model for analyzing the – CS OFF
Prokhorov (2021) Intelligent Transport Systems
Shahat, Hyun, and Yeom RE LO – – Summarized the present and future benefits – –
(2021) and challenges of DT cities
Yao, Wang, Su, and Qi RE MD – – Summarized the techniques and research – – –
(2021) directions of DT applied to ports, showing the
way to improve it
Bányai (2022) RE LO – – Summarized the DT techniques and algorithms – – –
in supply chain and logistics in different
scenarios.
(continued on next page)
The rationale for the paper "Digital Twins for Logistics and Supply Chain Systems: Literature Review, Conceptual Framework, Research Potential, and Practical Challenges" stems from the
growing importance of digital transformation in the logistics and supply chain industry. As businesses increasingly rely on efficient and responsive supply chain operations to meet customer
demands and remain competitive, there is a need to explore innovative technologies that can optimize processes and enhance decision-making.
5
Digital twin technology has emerged as a promising approach to address these challenges by creating virtual replicas of physical assets, processes, and systems. However, while digital twins
have gained traction in various industries, their application within logistics and supply chain systems remains relatively underexplored. 1
Therefore, the paper seeks to fill this gap by providing a comprehensive examination of digital twins in the context of logistics and supply chain management. By conducting a literature review, the
paper aims to synthesize existing research, theories, and case studies related to digital twins, offering insights into their conceptual foundations and practical applications.
Furthermore, by proposing a conceptual framework, the paper aims to provide a structured approach for understanding and implementing digital twins within logistics and supply chain systems.
This framework serves as a guide for researchers, practitioners, and policymakers interested in leveraging digital twin technology to optimize supply chain operations. 2
Table 1 (continued).
Paper ScopeDomainObjectiveMethod/Technique Focus Problem Dataset ValidationOnline/Offline
Binsfeld and Gerlach (2022) FW SC DS – Provided a quantitative way to evaluate – EXP OFF
the advantages when applying DT
supply chain on several experiments
Cheng, Yang, Zou, and Zuo (2022) RE SC – – Summarized the developments of – – –
applying 5G and the DTs in
manufacturing in supply chain fields,
boosting communications and its
security in the future.
Dy, Olivares-Aguila, and Vital-Soto (2022) RE SC – – Summarized current research of the DT – – –
supply chain in different fields under the
disruption risks, focusing on the
improvements of resilience and agility
Garrow and Mohan (2022) BO LO DS Developed a Combined a heuristic method with DT to80 nodes SIM OFF
heuristic reduce time-consuming of drone delivery
optimization
Grieves (2022) RE SC – – Discussed that AIDT will facilitate in – – –
operations of complex systems.
Kamble et al. (2022) RE SC – – Summarized the DT techniques in – – –
sustainable manufacturing supply chain
systems, providing a framework to guide
future research
Leung, Lee, and Ouyang (2022) FW LO DS Machine Learning Presented a framework to simplify the 600 PI-SKUs EXP OFF
PI-hub in the city logistics system.
Lv, Li, Feng, and Lv (2022) FW LO THE Deep learning Applied improved deep learning 20 iterations SIM OFF
algorithms to solve the security
problems in DT cooperative intelligent
transportation systems.
Marmolejo-Saucedo (2022) FW SC MOD Optimization Designed a framework to integrate 366 days CS OFF
large-scale optimization problems in a
digital platform to make decisions in the
supply chain.
Shen et al. (2022) IMP SC MOD System dynamic Established a system dynamic model to 2942 CS OFF
model demonstrate the positive effect of
digitization on the tobacco supply chain
Tzachor, Sabri, Richards, Rajabifard, and Acuto (2022)IMP LO THE – Summarized the benefits and limitations – – –
when applied DT to sustainable
development in an urban and non-urban
environment
Vilas-Boas, Rodrigues, and Alberti (2022) RE LO – – Focused on the techniques used to – – –
deliver fresh food and their difference
Wang, Deng, Shen, Hu, and Qi (2022) FW LO MOD Optimization Designed a DT supply chain framework 20 days CS OFF
with optimization and data
collaboration, applying it to JD company
during the pandemic period.
Yao, Sun, and Wu (2022) FW LO DS Machine learning Designed a multi-mode intelligent – – OFF
storage and retrieval system in
automated three-dimensional
warehouses, facilitating efficient and
intelligent warehouse operations
Zhao, Zhang, Chen, Qu, and Huang (2022) FW LO DS Deep learning Developed a knowledge graph about 96 gateways CS OFF
production logistics resource allocation
on DT and deep learning
Belfadel et al. (2023) RE LO – – Summarized the literature to fill the – – –
gaps of the lack of a framework for
urban logistics arrangement
Grosse (2023) RE SC – – Summarized the papers about using DT – – –
techniques to facilitate warehouse
selection in supply chain systems
Kalaboukas, Kiritsis, and Arampatzis (2023) FW SC THE – Provided a holistic management – CS OFF
approach combining three different
perspectives
Kajba, Jereb, and Obrecht (2023) RE SC – – Provided a conceptual model to analyze – – –
the trends of DT techniques in supply
chain and logistics for investment
assessment.
Klar, Fredriksson, and Angelakis (2023) RE MD – – Used data analysis to make sure the – – ONL
three core requirements of ports DT and
the way using DT to improve the port
resources, facilities and arrangement
Marinagi, Reklitis, Trivellas, and Sakas (2023) FW SC – – Provided some factors to evaluate the – – OFF
resilience of DT supply chain
(continued on next page)
Moreover, the paper explores the research potential of digital twins in logistics and supply chain management, identifying areas where further investigation is needed to unlock their full
capabilities. By highlighting research gaps and opportunities, the paper aims to inspire future studies that contribute to advancing the field and driving innovation.
Finally, by addressing practical challenges associated with implementing digital twins, such as data integration, privacy, and scalability issues, the paper aims to provide actionable insights for
overcoming barriers to adoption. This practical focus ensures that the research is not only theoretically grounded but also relevant and applicable to real-world scenarios.
6
In summary, the rationale for the paper lies in its potential to advance understanding, drive innovation, and inform decision-making in the field of logistics and supply chain management by
exploring the application of digital twins and addressing the associated opportunities and challenges. 3
Table 1 (continued).
Paper Scope Domain Objective Method/Technique Focus Problem Dataset Validation Online/Offline
Shrivastava et al. (2023) FW SC MOD Simulation Simulated the conditions of fresh food 21 ventilated SIM OFF
delivery to make the supply chain packages
intelligence and high efficiency
Zdolsek Draksler, Cimperman, and Obrecht (2023) FW LO DS Machine learning Provided a simulation framework to – EXP OFF
analyze and improve the interruption of
logistic
Note: FW = Framework; IMP = Implementation; BO = Framework & Implementation; RE = Literature Review; SC = Supply Chain; MD = Maritime Delivery; LO = Logistic; IoT
= Internet of Things; UL = Urban Logistics; THE = theoretically; DS = Design; MON = Monitor; MOD = Model; CS = Case Study; SIM = Simulation; EXP = Experiments; OFF =
Offline; ONL = Online
Fig. 3. Publications on DT in logistics and supply chain systems by year.
Initially, there were 310 keywords about both methodology and optimization and neural network are combinations of Optimization
application fields. To select the keywords displayed in Fig. 4, two rules and Simulation. Further, there are emerging methods in other clusters
were adopted: (i) Select the methodology-related keywords; and (ii) including Cloud Computing, Blockchain, Control, Decision Making,
Combine nearly the same keywords (e.g., cyber physical system and Network Design, Digitization, AI, and ML.
cyber–physical system). More advanced technologies and methods are limited or absent
General terms, such as ‘‘simulation’’, ‘‘digitization’’ and ‘‘artificial in the current literature and applications, for instance: (i) critical
intelligence’’ have high link strength. Domain-specific keywords with enablers: (hyper-scale) edge computing, quantum computing, neuro-
high link strength include ‘‘self-driving’’, ‘‘internet of things’’, and morphic computing, and 6G; (ii) productivity revolution: edge AI,
"automated control". And top methodology-related keywords include generative AI, and self-supervised learning; (iii) transparency and pri-
‘‘neural network’’, "self-optimization", and "random forest’’. Fig. 4 pro- vacy: responsible AI, digital ethics, and cybersecurity; and (iv) smart
vides valuable information about the current methodological directions world: AR/VR, metaverse, digital human (AI avatars). These technolo-
of research in DTs for logistics supply chain systems. gies and methods, that enable transparent, trustworthy, and resilient
To cluster the current DT methodologies and techniques, VOSviewer DTs for LSCS, are further investigated and analyzed in Section 5 of this
was used for All Keywords analysis. Fig. 5 reveals clusters and linkages paper.
of all methodology-related keywords from the literature presented in In summary, the literature review reveals DT gaps in vital indus-
Table 1. In this clustering process, VOSviewer firstly calculates the try sectors, implementation roadmap and management, and advanced
frequencies of all the keywords from the papers and references and methods and techniques that future research and practice need to
presents the keywords as nodes in the network. The non-technical address. Motivated by the gaps and needs from research and practice,
keywords are manually removed. VOSviewer then analyzes the co- the next section presents a conceptual DT and provides a framework of
occurrence technical keywords and connects them with weighted links. DTs for LSCS.
Finally, the closely related keywords are presented near each other with
similar colors under the mapping technique and clustering algorithm in 3. Conceptual DT framework for logistics and supply chain
VOSviewer’s visualization. Each color code represents one cluster and
each cluster reveals one methodology/technical category (van Eck & This section emerges from several questions: What does the future
Waltman, 2017). There are eight main clusters that can be observed DT for LSCS look like? How does it work? What are the stakeholders
from the figure. The larger the size of the node, the greater the in the DT system? What is a framework that operationalizes a DT for
frequency of the keyword. The thicker the link between the nodes, the LSCS?
stronger the relationship between the two keywords. Essentially, a DT has three layers: the physical, the digital, and the
As can be seen from Fig. 5, Optimization, Simulation, and Big Data communication layers. Fig. 6 shows a DT of a product flow of a supply
are the three largest clusters that lead the methodological framework chain. The physical layer includes all physical entities in the supply
used in the literature for addressing problems and challenges of DTs for chain (e.g., vehicles, buildings, and companies), starting from raw
LSCS. These three largest methodology clusters serve as the foundation materials to products’ end-users. The transportation/logistics service
for more specialization and innovation. For example, Simulation gives plays a vital linkage between the entities. The digital layer contains
rise to more detailed methodology, such as Genetic Algorithms and the digital replica of the entities, their relationships and functions, and
Distributed Simulation. Meanwhile, further extensions of Big Data are the real-time working environment. The communication layer enables
in control, such as Model-Predictive/adaptive Control and Control Sys- the communication between the physical entity and its DT.
tems. From a different angle, some novel methods are developed based The three layers form a closed loop to enable the DT function that
on combinations of primary methodologies. For example, heuristic has seven processes: sampling, storage, modeling, simulation, learning,
7
Fig. 4. All Author Keywords related to methodology.

Note: IM Systems: Intelligence Manufacturing Systems; AI-decision making: Artificial Intelligence-based Decision-making Algorithm; N-ANP: Neutrosophic Analytic Network Process.
prediction, and actuation. The beginning of this closed loop is sam- equipment). Then the sensory data is sent to the digital layer through
pling, where the physical entity is in charge of sampling data from its the communication layer, which is stored in databases, and also used
equipped sensors (e.g., on trucks, warehouses, and handling material for modeling the DT. Simulation of the DT adopts algorithms, such as
8
Fig. 5. High-Level Literature Clustering (All Methodology-related Keywords).
Fig. 6. Conceptual DT for logistics and supply chain systems.
machine learning/AI, and the prediction results are generated and sent In this study, the components of the DT for LSCS are structured
back to the physical layer. Such results are then utilized by the physical into three dimensions following the Reference Architectural Model for
entity to actuate and support the decision-making process of the LSCS. Industry 4.0 (RAMI 4.0) approach (Hankel & Rexroth, 2015). This
The impacts of such decisions can be sampled by sensors in the next approach is well accepted for showing the deployment of Industry 4.0
round (thus completing the loop). in a systematic way. Even though RAMI is developed and mainly used
for industries that produce physical products, it still makes sense to
Stakeholders in this DT system could include the government, the
apply RAMI to the both manufacturing and service industries. There-
DT operator, data providers, and data users. The government may reg-
fore, RAMI 4.0 ideas are adapted and modified as mechanisms to
ulate the DT system to ensure ethical and transparent operations. The
operationalize the DT framework for the LSCS which are presented in
DT operator manages the DT platform where advanced techniques and Fig. 7.
computation methods are embedded. The DT operator takes informa- The three dimensions of this RAMI framework are Layers, Hierarchy
tion inputs from data providers and provides processed information to Levels, and Life Cycle and Value Stream. The Layers dimension presents
data users. The data providers supply data to the DT operator whereas the structures of DT comprising six inter-related components (assets,
the data users request useful information from the DT operator. In integration, communication, information, functional, and business),
some cases, the data owner (any entity in the physical layer, such as which are categorized into three layers: physical (Assets and Integra-
companies, manufacturing, and enterprises) and data user are the same. tion), communication (Integration, Communication, and Information),
9
Fig. 7. Framework of the DT for logistics and supply chain systems with its components structured following RAMI approach.
Source: Adapted and modified from Hankel and Rexroth (2015).
and digital (Functional and Bussiness). The Assets component refers to related and closely integrated. Layers help categorize and structure the
physical equipment, goods, vehicles, etc., in the physical layer. Data various functional aspects of an LSCS. Hierarchy Levels provide a way
collected by sensors and other equipment is transformed into digital to organize and manage the complexity within each layer, allowing
format within the Integration component. Whereas the Communication for modularization and scalability. Whereas, Life Cycle & Value Stream
component handles the information exchange between the physical and cut across Layers and Hierarchy Levels, emphasizing the importance
digital layers. Data is processed and integrated to provide accurate and of considering the entire lifecycle of a system to maximize its value
useful information. In the Functional component, data is utilized for and efficiency. The suggested framework combines these dimensions
predictions, system management, route optimization, etc. The Business to provide a comprehensive framework for understanding, designing,
component digitally represents the business processes and services and implementing DT solutions for LSCS, ensuring that technology,
with data analytical results in LSCS. The Layers dimension depicts the organization, and processes work together harmoniously to achieve
vertical relationship and information exchange flows among the three the goals of increased automation, transparency, efficiency, and value
layers in the LSCS. creation.
The Life Cycle & Value Stream dimension represents different states The next section discusses in more detail the elements of the frame-
from ideas to service development until the end life of the service work.
(e.g., the end of the contract). Specifically, this dimension focuses on
processes in LCSC from initial concepts to final service delivery. The 4. Research and practical foci
aim of the Life Cycle & Value Stream dimension is to maximize flows of
business value through the service life cycle. Whereas, the Hierarchy Motivated by the new proposed framework, this section discusses
Levels dimension provides a perspective from operations to decision- the research potentials and practical challenges for the implementation
making, illustrating the different management levels in LSCS. Within of transparent, trustworthy, and resilient DTs for LSCS. Three subsec-
the Hierarchy Levels dimension, the Services refer to transportation, tions feature RAMI framework dimensions: Layers, Hierarchy Levels,
logistics, and supply chain services where the start of the chain is raw and Life Cycle and Value Stream. Because the boundaries between
materials, then suppliers, manufacturing, distribution, retail, and end some framework’s elements are porous, some literature studies are
users. Flexible systems provide the service and integrate it into the presented in the related work of an element, but not the others to
connected world. prevent duplication.
An operational activity can be pinpointed in the Hierarchy Levels
and the Layers dimensions. In LSCS, an operational activity is referred 4.1. The layers dimension
to as one part of the Hierarchy Levels. Meanwhile, the technique
associated with this operation can be located in the Layers dimension. The layer dimension comprises three main parts: the physical layer,
For example, ‘‘tracking shipment’’ can be at the ‘‘Work center’’ in the the digital layer, and the communication layer. The physical layer
Hierarchy Levels dimension, while the tracking techniques that ensure includes physical assets and their digital integration into the network.
accurate and timely information exchange belong to the ‘‘Communica- The digital layer incorporates two sub-layers which are functions of
tion’’ of the Layers dimension. The Layers dimension and the Hierarchy services and businesses (the things that customers are willing to pay
Levels dimension specify what activities and the corresponding re- for). The communication layer involves information/data, ways to
quired techniques are in a certain operational and management process access information, and partial integration of assets into the digital
in LSCS. Meanwhile, the activity is located in its corresponding Life network.
Cycle & Value Stream dimension. Furthermore, the Hierarchy Levels
and the Life Cycle & Value Stream depict which states are associated • The physical layer (in Fig. 7: Assets and Integration)
with a specific operation of the given process. Goals/motivations: Vehicles and infrastructure in LSCS aim to max-
In conclusion, the proposed framework represents a holistic ap- imize usage by making them more efficient and effective while
proach where the three dimensions of a DT for LSCS are strongly minimizing empty trips, waiting time, transaction time, etc.
10
Related work: The current literature presents two directions for for supply chain management to make predictions using time
handling assets and integration problems in the logistics and series data: Long-Short-Term-Memory (LSTM) and One-Class Sup-
supply chain physical layer. One direction is increasing the syn- port Vector Machine (OCSVM). The LSTM model was used for
chronization of each element. Leung et al. (2022) designed a forecasting, while the OCSVM was used for anomaly detection.
Digital twin-based Total Inbound Synchronization (DTIS) model, Taking numerical generated and real data into experiments, these
in which components are divided into three targets: order syn- two algorithms improved the performance of supply chain man-
chronization, process synchronization, and information synchro- agement.
nization. The authors lay a foundation for the future research on Improving algorithm efficiency is another important topic within
highly efficient data pre-processing, prediction, and optimization the digital layer. The system needs to process a considerable
in the digital layer. amount of data in a short period. Nikolopoulou and Ierapetri-
A few studies handle uncertainty and asymmetric problems in tou (2012) used a hybrid simulation approach to overcome the
the physical layer. For instance, Wang, Wang et al. (2020) used complexity of solving mixed integer linear problems, which al-
a data-driven control mechanism and cyber–physical system to lowed highly efficient solutions to supply chain management
reduce uncertainty in transportation. problems. Lv, Li et al. (2022) invented a new model, Coop-
Research opportunities: During DT operations, the physical layer erative Intelligence Transportation System (CITS) and DTs, to
generates dynamic information which is analyzed in the DT layer. solve security problems in transportation. Compared with previ-
However, this information is complex and hard to deal with ous models, CITS-DT has higher prediction accuracy and lower
directly. Therefore, automatic information pre-processing, includ- data transmission delay.
ing data cleaning and data selection, plays a crucial role in the Research opportunities: Ideas for advanced algorithms in the digital
physical layer. Additionally, developing algorithms with weak layer can be drawn from similar systems. Dai et al. (2021) applied
assumptions will be more practical (Leung et al., 2022). Further- the Lyapunov optimization technique to transfer the stochastic
more, it is essential to improve supply chain systems’ automation problems to a deterministic per-time slot problem in IoT systems,
to increase productivity. The deep learning algorithms developed then used a deep actor-critic algorithm to solve it. Qiao and
by Nguyen et al. (2021) were universal and could be used in Riddick (2004) designed and tested a simulation model aiming
different supply chain fields which needed abnormal forecasting at the manufacturing supply chain system.
and detection. These algorithms could connect with GPS or social Challenges for practical implementation: First, many stochastic fac-
networks to acquire and analyze data automatically, enhancing tors exist in LSCS because of humans and nature. Therefore, the
the model learning and prediction ability. approach to dealing with stochasticity is the first challenge. The
Challenges for practical implementation: First, enhancing data pro- LSTM algorithm of Nguyen et al. (2021) minimized the computa-
cessing, including city socio-economic, and building integration tional complexity and showed meaningful results in real-time, but
between the physical and digital layers are challenging (Shahat it ignored some stochastic elements, which became a limitation.
et al., 2021). Second, when designing the inbound warehouse In addition, the LSTM algorithm has high requirements for the
synchronization strategy, Leung et al. (2022) made assumptions input data that should be sequential. Consequently, data pre-
including the number of Store Keeping Units (SKU), the size of the possessing is an essential part of LSTM which is time-consuming.
forward and reserve areas, order fulfillment cycle time, and re- • The communication layer (in Fig. 7: Integration, Communication,
plenishment policy. These assumptions could be more relaxed for and Information)
more flexible practical implementation. Meanwhile, evaluating Goals/motivations: The communication layer contains three as-
stochastic factors in logistics is a complicated process. Therefore, pects: internal communication within DT, data transmission from
adding a proper statistical model to the synchronization strategy the physical layer to the digital layer, and data transmission
is helpful (Leung et al., 2022). from the digital layer back to the physical layer. The integration,
• The digital layer (in Fig. 7: Functional and Business) communication, data, and information are expected to be fast,
Goals/motivations: The digital layer processes information using stable, and secure.
algorithms (e.g., AI/ML), simulation, and reasoning to inform Related work: (1). Internal communications within DT: A DT
the decision-making of stakeholders given the complex, uncer- system contains elements for optimization, storage, and predic-
tain, and stochastic characteristics of the physical layer. The tion. These elements should communicate and transfer informa-
data need to be processed quickly to provide timely support for tion efficiently. Layaq, Goudz, Noche, and Atif (2019) used the
decision-making. blockchain method to solve the information transmission in sup-
Related work: AI techniques and algorithms facilitate the DT for ply chain systems, then Dietrich, Turgut, Palm, and Louw (2020)
LSCS by utilizing historical data to enrich real-time data for better compared blockchain-based solutions in supply chain manage-
analysis and prediction. As a result, the DT capacity is expanded, ment. The convolution method had fast implementation but could
especially under abnormal situations. For example, reinforcement not control complex systems. The extended method could con-
learning can enhance automatic functions in LSCS (Abideen et al., trol complex systems with low efficiency. Both methods could
2021), and facilitate stakeholders to decide the next steps under avoid some risks under information opaque. Moreover, cloud
certain curriculums (Park et al., 2021). Kegenbekov and Jackson computing techniques can accelerate the communication speed
(2021) provided an adaptive supply chain system using smart in dealing with dynamic information in logistics supply chain
agents to maintain the demand-supply balance which was trained systems (Campos et al., 2020).
under the reinforcement learning algorithms. Chen and Huang (2). Data flow from the physical layer to the digital layer: Sensors
(2021) found that the DT is a valuable tool for dealing with collect data (of trucks, robots, material handling machines, con-
physical information sharing in the supply chain of the manu- tainer temperature, etc.) in LSCS and transfer it to the digital layer
facturing industry, especially for handling dynamic asymmetric for analysis and prediction. Lu, Olofsson, Jensen, and Simonsson
information. Chen and Huang (2021) combined the IoT and DT (2011) combined the Building Information Model (BIM) with the
in the supply chain to enhance the efficiency of information Discrete Event Simulator (DES), providing symmetric information
communication. (for example, delivery time and cost) in the logistics supply chain
Furthermore, combined with deep learning algorithms, like GNN systems. Lee and Lee (2021) clarified that the BIM acquired
or CNN, a DT can analyze historical data to support current information in the supply chain system, including the locations
decision-making. Nguyen et al. (2021) proposed two approaches and connections of elements, and converted the information to
11
computer-readable signals under the Internet of Things (IoT) Regarding methodology, Belfadel et al. (2023) designed a data-
sensors, sharing with all DT participants. driven conceptual framework facing urban logistics, and divided
Moreover, augmented reality (AR) is another development di- it into three parts: the physical world, data and model manage-
rection that brings potential benefits for DTs for logistic supply ment, and storage, to minimize the cost and delivery time. The
chain systems. AR enhances the connection between the real and authors also developed models to simulate empirical decisions
the virtual, improves processes, and saves time spent to optimize made by stakeholders.
every element in the systems (Kajba et al., 2023). Research opportunities: Firstly, empirical decisions could be con-
(3). Control flow from the digital layer back to the physical layer: sidered for integration into the DT framework. During logistic
After completing analysis and prediction in the DT layer, the and transportation processes, empirical functions and decisions
results are transferred to reality. Lee and Lee (2021) proposed a are generated. Incorporating empirical decision-making into DT
framework that comprises IoT, BIM, and GIS to convert results could improve their reliability (Belfadel et al., 2023). de Bok and
into reality decisions (for example, an optimal delivery route or Tavasszy (2018) built an agent-based urban logistic framework
strategy). called MASS-GT which includes simulation of empirical decisions,
Research opportunities: There are several research opportunities, and performed well after testing (Davidsson, Henesey, Ramstedt,
including the combination of data flow and control flow, how Törnquist, & Wernstedt, 2005). Secondly, improving DT robust-
to improve communication efficiency, and adopting algorithms
ness and stability are important research directions. Only after
from similar fields. First, the control and data flow can be in-
intensive training, empirical decisions can be effectively included
tegrated under back-end architectures (Lee & Lee, 2021). Under
in a DT framework.
this integration, simulator predictions are not only much more
Challenges for practical implementation: Some frameworks are de-
accurate but they reduce risks and provide more reliable results.
veloped for unique fields that make transferring results to other
This integration approach can be applied to other industries be-
fields difficult. Conversely, studies that developed models under
cause of its flexibility and easy construction. Moreover, enhancing
strong assumptions make them less practical.
communication efficiency is also needed, as it is one of the critical
• Flexible Systems
factors that promotes the wide use of DT for logistics supply chain
systems. Goals/motivations: A flexible system has the ability to respond to
Lastly, some useful algorithms or systems in the communica- supply chain uncertainties and disruptions. The system is resilient
tion layer of DT from other fields can also be applied to logis- to extreme events. It could also handle high-demand seasons
tics supply chain systems. Lv, Chen, Feng, Wei, and Lv (2022) while adjusting to lean operations during low-demand seasons.
designed a marine communication network system to collect The system’s functions are distributed through the network that
and transmit underwater monitoring data. Redelinghuys, Bas- crosses company boundaries where stakeholders interact across
son, and Kruger (2020) developed a DT six-layer architecture of hierarchical levels.
the Cyber-physical Product System (CPPS) with the Internet of Related work: The synergy between cloud computing and DT can
Things (IoT) to improve communication with minimum process bolster interactions among the various components within LSCS,
disruption. O’Dwyer, Pan, Charlesworth, Butler, and Shah (2020) thereby fostering flexibility. DT could serve as a platform for
designed a Sustainable Energy Management System (SEMS) to homogenizing datasets, that facilitates more streamlined and ef-
avoid system constraint violations during coordination in differ- fective subsequent analysis. This combination of cloud computing
ent sectors through timely data communications. and DT holds great promise for improving the robustness and
Challenges for practical implementation: The communication layer adaptability of LSCS (Mylonas et al., 2021; Qian, Luo, Du, & Guo,
between the digital layer and the physical layer needs to convert 2009; Tzachor et al., 2022).
the machine-readable language to human-readable language (and Research opportunities: First, data standardization is a key topic to
vice versa). Therefore, the development of a precise interpretation achieve flexible DT logistic and supply chain systems. Different
and reliable communication technique is one of the Challenges for companies or agencies provide related data to the DT platform.
practical implementation. To create a flexible data exchange process that supports com-
Another challenge is the security in the communication layer. putational techniques, a win-win collaboration model should be
Security is always a topic of concern for data/information trans- developed. Second, promoting full-scale digitalization is a poten-
mission. Careful data security measures are needed. tial research direction. The flexible system is not only limited to
one country but could expand globally. If the technique of DT is
4.2. The hierarchy levels dimension owned privately or handled by only a few agencies or companies,
it challenges the construction and operation of the global DT
The hierarchy dimension presents integrations between stakehold- flexible system.
ers in the chain. Challenges for practical implementation: It is challenging to en-
• Service (transportation, logistics, and supply chain services) sure all stakeholders in the system share a common perspective,
Goals/motivations: Transportation, logistics, and supply chain ser- understanding, and responsibility. The communication, standard-
vices should be accessible to society, regardless of urban deliv- ization, and integration in the system facilitate a big volume
eries (e.g., same day, next day, first-last mile, etc.) or long-haul of data and information to be quickly and efficiently processed
transport (air, maritime, road, rail, pipeline). and interpreted. However, international co-operations towards a
Related work: In the LSCS for agriculture, building a DT for food global standard seem to be a challenging mission to be achieved
logistics, including monitoring processes and decision-making, soon.
improves on-time delivery and food freshness (Shrivastava et al., • Connected World (interconnected global supply chain)
2023; Vilas-Boas et al., 2022). In the maritime supply chain Goals/motivations: To establish an interconnected and resilient
systems, DT facilitates shipping companies to optimize transporta- global supply chain where information is processed at the local
tion fleets, port arrangements, terminal scheduling, etc. (Lind level before integrating at the global level.
et al., 2020) that help to reduce the total freight delivery time Related work: Busse et al. (2021) claimed that synchronization
in the supply chain. Moreover, DT can be applied to the port of transport and transshipment processes is the key factor in
organization to facilitate collaborations and governance (Klar global supply chain systems. The authors designed a Digital Sup-
et al., 2023) and promote cost reduction (Yao et al., 2021). ply Chain Twin (DSCT) to handle a multi-model supply chain.
12
Under the technical combination (5 Generation [5G] technolo- simulation models, productions, and applications still faces un-
gies, Cloud Computing, Artificial Intelligence, etc.), the DSCT predictable challenges like material, environmental, and ethical
had good performance in predicting delivery time and optimizing issues (Kajba et al., 2023).
delivery networks, especially in maritime transportation–the main • Service provision and sales
component of the global supply chain systems. Goals/motivations: New supply chain and logistics services are
Research opportunities: In the realm of global supply chain man- needed to fulfill different niche-market segments (e.g., different
agement, the application of graph theory offers valuable insights sectors or industries) whose requirements change from time to
into network characteristics, including but not limited to connec- time. The new supply chain and logistics services need to keep
tivity, centrality, network diameter and spread, detours, and the up and be well-integrated with the technology and development
identification of industry-specific sub-networks. These insights of the sector/industry/customer that it serves. At this stage, the
can significantly augment the resilience of the DT system, par- services are offered to the market.
ticularly in instances of disruptions such as broken links or nodes Related work: (Jeong et al., 2020) designed a logistics process
within the network. for the production industry based on DTs. The design work con-
Challenges for practical implementation: There are several chal- tained all the elements in the logistics tasks, including warehouse
lenges, such as the bullwhip effect, misleading information, and locations, storage and material handling, etc. Through DT imple-
data policies. Even with a small discrepancy in demand fore- mentations, the logistics process could avoid wasting resources
casting at the retail level, it can cause significant impacts on and opportunities, then improve the operation efficiency. Thus,
operations of the lower chain level, including wholesale, distri- this design was helpful for the technical and beneficial problems
bution, manufacturing, suppliers, and the raw material industry. in logistics. In addition, Marmolejo-Saucedo (2022) applied DTs
Additionally, if the information submitted to the system by the jointed with complex large-scale optimization algorithms (3D-BPP
data providers is wrongly created either on purpose or by in- and GVRP) focusing on the parking supply chain, accelerating
cident/innocent, it may cause severe impacts on a global scale. information processing in real-time.
Moreover, data-sharing policies vary and could be more strict Wang, Deng et al. (2022) developed a DT supply chain system
in one place than in others making the DT even more frag- (DTSC) that contains suppliers, manufacturing factories, etc. This
ile. Keeping a DT up-to-date, universal data policies is a real DT includes basic features (e.g., organized levels) and business
challenge. processes (e.g., transportation available) of fundamental supply
chain systems. The DTSC platform was applied to the JD com-
4.3. The life cycle and value stream dimension pany, and reported excellent performance in high-efficiency data
processing and disruptions prediction, especially during COVID-
The life cycle of transportation, logistics, and supply chain services 19.
starts from the service development and ends after the service is Research opportunities: Building a DT to support the development
delivered to customers. The business value created during this life cycle of resilient LSCS is a potential research opportunity. A small dis-
is provided across the temporal (i.e., time) and spatial (i.e., space). This
ruption in a system element may break the balance of the whole
dimension includes service development, quality assurance, service
Logistics and supply chain system, not to mention a large shock,
creation, and maintenance and usage.
such as the COVID-19 pandemic. Therefore, it is worthwhile to
• Development investigate how to increase the systems’ resilience using DTs and
Goals/motivations: In this first stage of the life cycle, new ser- minimize the negative influence when a disruption happens.
vices are developed, prototyped, and constructed. Continuous Challenges for practical implementation: It is challenging to collect
development adds value to the chains by improving efficiency, useful big data for customer service improvement. It is also a
reducing costs, and gaining more benefits. New technologies, such challenge to apply new technology, such as DNA tracking, arti-
as connected and automated vehicles, electric vehicles, delivery ficial intelligence, and blockchains to trace raw materials from
robots, and drones could improve transportation, logistics, and the original source to the store.
supply chain services. • Maintenance and usage (Quality assurance and customer service)
Related work: Literature presents several works that focus on Goals/motivations: In the Type phase in Fig. 7, maintenance and
developing new DT algorithms or systems. Rudskoy et al. (2021) usage refer to the quality assurance steps that model the mainte-
designed a new intelligent transport system (ITS) with the help of nance by preparing software, instruction manuals, product
DTs. Compared to the original control system, ITS could improve changes, etc. In the Instance phase, maintenance and usage refer
traffic flow and control the service in an emergency. Meanwhile, to customer service (e.g., optimization, updates, troubleshooting)
ITS could make predictions based on the DT data with artificial which plays a significant role in retaining customers.
intelligence algorithms. Marmolejo-Saucedo (2020) developed a Related work: The nature of the usage and maintenance of DT
DT for a pharmaceutical company based on mathematical models LSCS is to promote sustainable development. Pan et al. (2021)
and data analysis, using optimization and dynamic programming highlighted the essence of sustainability in LSCS and provided
to facilitate decision-making. With the help of this DT technology, the emergence of DT techniques that facilitate the maintenance
the information shared is more accurate and predictable, which and usage of these systems. Another direction to improve the
brings financial benefits to companies. Recently, Kalaboukas et al. maintenance and usage of the systems is to optimize the last-
(2023) developed a cognitive DT focuses on holistic supply chain mile delivery that is the most time-consuming and expensive
governance. part of the logistics system (Gutierrez-Franco et al., 2021). More-
Research opportunities: Some obstacles impede the implementa- over, Cho, Lee, Ahn, and Hwang (2012) provided a framework
tion of DT development and will become future research op- based on the fuzzy analytic hierarchy process (fuzzy-AHP), which
portunities: sensor installation (Rudskoy et al., 2021), mecha- focuses on the service measurement in a supply chain. Under this
nism of liability sharing (Kalaboukas et al., 2023), and possible framework, the service processes can be visualized and accessible.
modifications of the model (Kajba et al., 2023). Research opportunities: First, to enhance the service quality in
Challenges for practical implementation: Even though computation logistic and supply chain systems, it will be helpful to improve
methods and DT technologies have developed for several years, the quality assurance approaches. Effective and sufficient ser-
producing rich theoretical research related to algorithms and vice measurement methods will improve the efficiency of the
13
Fig. 8. Data and information flows between physical and digital layers.
LSCS. Second, as the 5G cellular network is more popular, in- to train the computer version to make better predictions. Such
vestigating ways to utilize customer information (e.g., searching improvements are beneficial for the construction of sustainable
activities, location sharing, feedback) to improve the LSCS is an- and flexible LSCS, especially by incorporating 3D images and
other interesting research direction. Third, more research efforts VR techniques (Shekari & Rajabzadeh Ghatari, 2013; Xu, Qi,
are suggested to take advantage of blockchain technologies to et al., 2022). Moreover, videos or text messages were used to
improve the efficiency of the LSCS. train natural language processing (NLP) algorithms for improving
Challenges for practical implementation: Technologies change over production logistics resource allocation (Zhou, Awasthi, & Stal-Le
time and create challenges for maintenance, as software, devices, Cardinal, 2021). On the other hand, DT possesses the capability to
etc., need to be updated regularly and the related labor force rapidly analyze and visualize data, thereby offering an effective
needs to be retrained. platform for the validation of complex questions or algorithms
through experiments without disrupting the real system. For in-
5. The future of DT computation stance, DT can be utilized to find a solution with a minimal
cost objective function under soft assumptions derived from real
The proposed DT framework creates new research opportunities data in the first and last-mile delivery (Garrow & Mohan, 2022).
and practical challenges that initiate a new agenda for future DT Or DTs could also facilitate experimentation aimed at tackling
computation. The DT guiding computational techniques discuss how the complex logistical challenges, such as the traveling salesman
future DT requirements will facilitate robust, reliable, and explainable problem (Garrow & Mohan, 2022).
computation. The computational techniques guiding DT considers how Research opportunities: In literature, scientists designed a DT sys-
advanced techniques enable transparent, efficient, and trustworthy DT tem named Automatically Configuring DT, which breaks the re-
systems. We also discuss the design and evaluation of a future DT strictions of the reinforcement learning model. The model op-
platform at the end of this section. timizes the final decisions based on the human experience or
Fig. 8 illustrates a conceptualization of the DT computation. The relevant contextual information and has been verified as effec-
data flow transfers real-time information from the physical layer to tive in manufacturing (Bolender et al., 2021) and human-in-the-
the digital layer. Real-time information can be captured by sensors, loop adaptive systems (Yigitbas, Karakaya, Jovanovikj, & Engels,
cameras, etc., equipped on assets and infrastructure systems or from 2021). Those techniques could be adapted to improve DT for
information management systems. After being processed by computa- LSCS. Additionally, real-time simulation techniques that support
tional techniques, the data is converted to a machine-readable format decision-making should also be developed.
for further computation and prediction in the digital layer. The control Challenges for practical implementation: The physical facilities or
flow, on the other hand, is directed from the digital layer to the physical hardware of DT should be durable and easy to update. For ex-
layer. Within the digital layer, AI computational models utilize both ample, as the techniques update, the data transmission facilities
historical and real-time data to forecast and make predictions. The ac- should be easily adapted to maintain high transmission efficiency.
tuators’ interpretations translate predictive results to assist stakeholders As computation methods and techniques are also changed over
in the physical layer in making reliable decisions (thus completing the time, it is challenging to maintain harmonious operations be-
whole loop). tween new hardware and software without causing delay or
disruption to the DT system.
• DT guiding computational techniques • Computational techniques guiding DT
Goals/motivations: The need to design, build, implement, and Goals/motivations: On the other hand, computational techniques
operate DT for LSCS requires computational techniques to be will guide DT in the sense that the descriptive computation tech-
robust, reliable, and explainable. More data flows in from dif- nique is informed by information from sensors embedded in the
ferent sources and maybe even in different formats requiring real system as presented in Fig. 8. The computational techniques
computational techniques to handle timely and effectively. The will facilitate the DT to be designed, performed, and managed.
predictive technique that was trained and tested from DT will be Related work: As the LSCS are applied to many new fields, the
converted by an actuator to guide the real system as can be seen complexity of the systems is increasing dramatically (Grieves,
from Fig. 8. 2022). Computational techniques are used to design algorithms
Related work: The developments of DT for LSCS would stimu- and integrated network flow to handle complex systems (Kittur
late computation technique inventions. On one hand, sensors et al., 2013). Grieves (2022) provided a conceptual model called
collect abundant data from various sources to train different the Intelligence Digital Twins system (IDT), combining AI, model-
algorithms in the systems. For instance, images can be gathered ing, and simulation techniques, which can dynamically calculate
14
and monitor the DT status and avoid emergency disruption sit- 6. Next steps
uations of the real system. Almasan et al. (2022) summarized
algorithms for solving DT network problems; for example com-
This section is motivated by several questions: How can we move
bining the traditional algorithms (e.g., Constraint Programming
towards building a DT for LSCS that is integrated, secure, and reliable?
(CP) and Integrated Linear Programming (ILP), etc.) with some
More importantly, what technologies and techniques are needed to
new algorithms (e.g., Graphical Neural Networks (GNN), etc.).
design a DT system that is fair, transparent, and resilient; protects
These algorithms would help to integrate and optimize the DT
business secrets, and benefits all stakeholders? What is the roadmap
system and make it more efficient. Besides that, some AI can
be treated as a reflective tool. Compared with purely utilizing for implementation? and What are the maintenance and management
reinforcement learning or machine learning in AI, Deng et al. strategies for the DT life cycle? This section presents initial ideas for
(2021) provided a self-optimization model with better perfor- the next steps that facilitate the deployment of a DT for LSCS.
mance, combing the experts’ knowledge, reinforcement learning, Fig. 9 presents initial ideas on the next steps to build a reliable
and DT, to solve high-dimensional network problems. Research op- and transparent DT for LSCS that comprises data-driven value (orange
portunities: Explainability is one of the key parts of computational triangle in the middle) and data-informed value (green triangle on
techniques guiding DT. Even though deep learning models have the right) methods. In a complex system, technology may be allowed
great performance in prediction and optimization, especially for to make some decisions (i.e., data-driven decisions) whereas other
some neural network models, the explainability of their results decisions should be made by humans (i.e., data-informed decisions).
is still a big challenge. Future research should find a balance Because tacit knowledge is still hard to be adequately modeled by
between the number of model layers and the model’s explain- current methods and techniques. Humans are informed by the data and
ability. Furthermore, computationally tractable models should be use their experience to evaluate the situation and make decisions ac-
developed to address uncertainty and transparent issues. cordingly. Therefore, data-driven and data-informed value approaches
Challenges for practical implementation: Operations of a large-scale should both be utilized for decision-making to take the strengths while
DT can be a challenge for computation methods and techniques. overcoming the limitations of the two approaches.
Moreover, the complexity of the real system together with an
The ideas of the data-driven value are motivated by Maslow’s
uncertain environment create obstacles for computational tech-
Hierarchy of Needs as the ultimate goal is to build a responsible,
niques.
trustworthy, and resilient DT. Data and information are the foundation
• DT platform
of the process: good data, good decisions. There is a need to collect
Goals/motivations: In the near term, building a fully compre-
and prepare quality data for analytics which essentially provides a
hensive DT platform, including all aspects of supply chain and
better understanding of the physical systems. Communication needs
logistics, including R&D, is a big and ambitious mission that
to ensure smooth transactions and linkages of the physical and the
involves many stakeholders. A DT platform is expected to be
digital layer while safely and securely preserving the information. As
secure, reliable, and resilient to provide benefits for all partners.
Related work: DT platform can be integrated into an ecosystem an engine of a DT, modeling and data-driven methods and techniques
to optimize routes, predict performance, and make decisions by (e.g., explainable AI, metaverse) enable prediction requirements and
analyzing the current and historical data (Marmolejo-Saucedo, explainability.
2020). A reliable DT platform needs a high organizational level On the other hand, the data-informed value approach dictates a
to manipulate every part of LSCS simultaneously. Currently, there sequence of using data to inform decision making and create value
are only a few research papers focusing on virtual simulation of which is essentially the ultimate goal of for-profit businesses. Data
DT platforms. Marmolejo-Saucedo (2020) modeled a DT platform is processed and converted into information that informs employees.
of a pharmaceutical company, from purchasing raw materials to Based on the provided information and the employee’s experience and
market sales. This research was primarily based on computer insights, further searches and investigations are conducted to inform
simulations and lacked practical applications. On the other hand, decisions. Actions would be made accordingly that result in outcomes
some major scientific companies have invented general DT plat- and value creations.
forms, such as the Omniverse Enterprise by NVIDIA (Nvidia, Data-driven value and data-informed value play essential roles in
2023) and Azure Digital Twins by Microsoft (Baanders, 2023), enhancing operational excellence and human excellence in LSCS. The
for various purposes. These platforms have been used to optimize data-driven value and data-informed value in DT can boost operational
distribution center throughput and train autonomous warehouse excellence and human excellence. The two values facilitate improve-
robots, among other applications.
ment and (i) maintain favorable customer relationships and make
Research opportunities: In the literature, some DT platforms have
improvements in the systems’ revenue growth; (ii) integrate the systems
been applied to solve industrial problems, but there is still a lack
and form firm cooperation between companies among LSCS; (iii) high-
of work that builds a DT platform for LSCS, as it has special
light the essential of human resources, cultivation, and management for
features compared to other industrial fields, such as long-distance
making constructive decisions; (iv) provide a measurement for opera-
collaboration and making joint decisions timely (Marmolejo-
tions, such as the cycle time for delivery, the timeliness of after-sales
Saucedo, 2020). To fulfill such special features, there are much
service, and the productivity (Rai, Patnayakuni, & Seth, 2006); and (v)
more strict requirements for DT platforms in logistics and sup-
ensure transparent, trustworthy, sustainable, and resilient operations
ply chain than in other industrial fields regarding inter-sector
collaborations, computing efficiency, and predictive accuracy. regards to the uncertain and dynamic environment under which the
Challenges for practical implementation: Challenges and obstacles LSCS is operated. While (i) and (iii) refer to human excellence, the other
still remain as some key questions that need to be addressed, such items concern operational excellence.
as who should build and manage this DT platform? What contri- Above all, there are maintenance and management needs that sus-
butions of different stakeholders are expected to make the system tain the operations of a DT system in its life cycle. More detailed
work effectively and efficiently? How to assess the performance discussions on methods and techniques that enable the implementation
and robustness of the DT models? How to protect the DT system of a DT for LSCS are left for future research, together with maintenance
from cyber-attacks? and management strategies.
15
Fig. 9. Initial ideas on the next steps for building a transparent, trustworthy, and resilient DT for logistics and supply chain systems that comprise data-driven value (orange
triangle in the middle) and data-informed value (green triangle on the right-texts adapted from Wells (2023)) methods.
7. Conclusions and the components in each dimension guide the DT construction

step by step. Furthermore, the research identifies crucial research and
DT has garnered considerable attention for its systematic integra- implementation gaps in the field that can help inform future research
tion, collaborations, and promising benefits. Applications of DT tech- efforts and investments, enabling decision-makers to focus on areas
niques have spanned diverse sectors, from autonomous systems to where innovation and improvement are most needed. By highlighting
manufacturing, robotics, transportation, etc. Logistic and supply chain these gaps, the research encourages researchers to develop innovative
systems, with their stringent requirements for timely corporation and methodologies and explore new research directions. Decision-makers
information sharing, are in line with the features of DT. Moreover, such can leverage these emerging technologies and strategies to enhance
innovative DT could stimulate the advancement of global logistics and their logistics and supply chain operations. Moreover, the research em-
supply chain development, potentially boosting the global economy.
phasizes the need for stronger interaction and integration of advanced
Consequently, this paper concentrates on the DT applied to LSCS aiming
technologies and methods to build a transparent, responsible, trustwor-
to collect and synthesize useful information and techniques to guide
thy, and resilient DT system. This insight can guide decision-makers in
future developments in this domain and to support industry planning
adopting and integrating cutting-edge technologies to optimize their
and government decision-making.
supply chain operations. In summary, the research paper provides
This paper first systematically reviewed and synthesized literature
studies concerning methodologies, frameworks, and implementations valuable insights and a structured approach to decision-makers in the
of DT techniques for LSCS in several ways, such as comparison form, field of logistics and supply chain operations. It not only reviews
keywords link strength, and keywords cluster. Then, this paper pro- existing knowledge but also encourages innovation and the integration
posed a conceptual framework of DT for LSCS, divided into three of advanced technologies to enhance decision-making processes and
dimensions: the hierarchy levels dimension, the layers dimension, and overall performance in this critical industry.
the life cycle and value stream dimension. Detailed analysis of elements Future research should provide more details about emerging tech-
in the three dimensions enriched by structuring discussions to include nologies and techniques that enable transparent, trustworthy, and re-
goals/motivations, related research, research opportunities, and Challenges silient DTs for LSCS. Another direction for future research is to propose
for practical implementation. Several research and implementation gaps a roadmap for implementing a DT. Finally, maintenance and manage-
were identified, for instance, only a few studies on the DT targeting ment strategies need to be suggested to ensure effective and efficient
real-time/online simulations or lack of practical experiment capacity in operations as well as sustain the DT life cycle.
the current DT techniques and methods. The syntheses and analyses of
the DT for LSCS facilitate researchers to develop innovative method-
ologies and new research directions addressing the identified gaps. CRediT authorship contribution statement
Future DT computation requires stronger interaction and integration
of advanced technologies/methods to build a transparent, responsible, Tho V. Le: Conceptualization, Investigation, Methodology, Project
trustworthy, and resilient DT system. administration, Supervision, Validation, Visualization, Writing – orig-
The paper’s systematic review and synthesis of literature studies inal draft, Writing – review & editing. Ruoling Fan: Investigation;
provide a comprehensive understanding of existing methodologies, Formal analysis, Visualization, Writing – original draft.
frameworks, and implementations of DT techniques for LSCS. This
offers a valuable resource for decision-makers in LSCS by providing
insights into proven strategies and best practices. The proposed con- Declaration of competing interest
ceptual framework for DT in LSCS serves as a practical guide for
decision-makers. This framework helps in structuring and aligning The authors declare that they have no known competing finan-
decision-making processes within logistics and supply chain opera- cial interests or personal relationships that could have appeared to
tions. The framework provides the technical depth (Layer dimension), influence the work reported in this paper.
the organization width (Hierarchy Levels), and the temporal continu-
ity (Life Cycle & Value Stream). To build DT for LSCS, stakeholders
should adopt essential techniques, adjust the LSCS organization and Data availability
construction, and optimize operational states in LSCS. These three
steps correspond to the three dimensions of the proposed framework, No data was used for the research described in the article.
16
Acknowledgment Davidsson, P., Henesey, L., Ramstedt, L., Törnquist, J., & Wernstedt, F. (2005). An
analysis of agent-based approaches to transport logistics. Transportation Research
part C: emerging technologies, 13(4), 255–271.
The authors want to thank Professor. Matthew Roorda of the Uni-
de Bok, M., & Tavasszy, L. (2018). An empirical agent-based simulation system for
versity of Toronto; Professor. Lóri Tavasszy, Associate Professor. Aaron urban goods transport (MASS-GT). Procedia Computer Science, 130, 126–133.
Ding, and Assistant Professor. Mahnam Saeednia of the Delft University Defraeye, T., Tagliavini, G., Wu, W., Prawiranto, K., Schudel, S., Kerisima, M. A., et al.
of Technology for reading and commenting on an earlier draft of this (2019). Digital twins probe into food cooling and biochemical quality changes for
paper. There is no funding associated with this research. reducing losses in refrigerated supply chains. Resources, Conservation and Recycling,
149, 778–794.
Del Giorgio Solfa, F. (2022). Impacts of cyber security and supply chain risk on digital
References operations: Evidence from the pharmaceutical industry. International Journal of
Technology, Innovation and Management (IJTIM), 2.
Abideen, A. Z., Sundram, V. P. K., Pyeman, J., Othman, A. K., & Sorooshian, S. (2021). Deng, J., Zheng, Q., Liu, G., Bai, J., Tian, K., Sun, C., et al. (2021). A digital
Digital twin integrated reinforced learning in supply chain and logistics. Logistics, twin approach for self-optimization of mobile networks. In 2021 IEEE wireless
5, 84. http://dx.doi.org/10.3390/logistics5040084. communications and networking conference workshops (pp. 1–6). IEEE.
Allen, B. D. (2021). Digital twins and living models at NASA. In Digital twin summit. Dietrich, F., Turgut, A., Palm, D., & Louw, L. (2020). Smart contract-based blockchain
Almasan, P., Ferriol-Galmés, M., Paillisse, J., Suárez-Varela, J., Perino, D., López, D., solution to reduce supply chain risks. In Advances in production management systems.
et al. (2022). Digital twin network: Opportunities and challenges. arXiv preprint Towards smart and digital manufacturing: IFIP WG 5.7 international conference, APMS
arXiv:2201.01144. 2020, Novi Sad, Serbia, August 30–September 3, 2020, Proceedings, Part II (pp.
Baanders (2023). What is azure digital twins? - azure digital twins. Azure Digital Twins 165–173). Springer.
| Microsoft Learn, Accessed 26-April-2023, URL https://learn.microsoft.com/en- Dy, K. J., Olivares-Aguila, J., & Vital-Soto, A. (2022). A survey of digital supply chain
us/azure/digital-twins/overview. twins’ implementations. In IFIP international conference on advances in production
Balanced Scorecard Institute (2023). Strategic planning basics. Accessed 28- management systems (pp. 502–509). Springer.
August-2023, https://balancedscorecard.org/strategic-planning-basics/#:~: Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: what are they? Strategic
text=Strategic%20planning%20is%20an%20organizational,assess%20and% Management Journal, 21(10–11), 1105–1121.
20adjust%20the%20organization’s. European Commission, Directorate-General for Justice and Consumers (2022). Corpo-
Bányai, T. (2022). Transformation of conventional manufacturing and service systems rate sustainability due diligence directive (CSDD). Accessed 17-November-2022,
into a cyber-physical environment: Review of potential solutions. In Vehicle and URL https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX:52022PC0071.
automotive engineering 4: Select proceedings of the 4th VAE2022, Miskolc, Hungary Far, S. B., & Rad, A. I. (2022). Applying digital twins in metaverse: User interface,
(pp. 456–471). Springer. security and privacy challenges. Journal of Metaverse, 2(1), 8–15.
Barykin, S. Y., Bochkarev, A. A., Kalinina, O. V., & Yadykin, V. K. (2020). Concept
Garrow, S. G., & Mohan, S. (2022). Extension of traveling salesperson problem with
for a supply chain digital twin. International Journal of Mathematical, Engineering
drone for manufacturing digital twin data collection. In 2022 American control
and Management Sciences, [ISSN: 24557749] 5, 1498–1515. http://dx.doi.org/10.
conference (pp. 4056–4061). IEEE.
33889/IJMEMS.2020.5.6.111.
Gelernter, D. (1993). Mirror worlds: Or the day software puts the universe in a Shoebox...
Belfadel, A., Hörl, S., Tapia, R. J., Politaki, D., Kureshi, I., Tavasszy, L., et al. (2023).
how it will happen and what it will mean. Oxford University Press.
A Conceptual Digital Twin Framework for City Logistics.
Glaessgen, E., & Stargel, D. (2012). The digital twin paradigm for future NASA and
Belfadel, A., Hörl, S., Tapia, R. J., & Puchinger, J. (2021). Towards a digital twin
US Air Force vehicles. In 53rd AIAA/ASME/ASCE/AHS/ASC structures, structural dy-
framework for adaptive last mile city logistics. In 2021 6th International conference
namics and materials conference 20th AIAA/ASME/AHS adaptive structures conference
on smart and sustainable technologies (pp. 1–6). IEEE.
14th AIAA (p. 1818).
Binsfeld, T., & Gerlach, B. (2022). Quantifying the benefits of digital supply chain
Gorodetsky, V., Kozhevnikov, S., Novichkov, D., & Skobelev, P. O. (2019). The frame-
twins—A simulation study in organic food supply chains. Logistics, 6(3), 46.
work for designing autonomous cyber-physical multi-agent systems for adaptive
Bolender, T., Bürvenich, G., Dalibor, M., Rumpe, B., & Wortmann, A. (2021). Self-
resource management. In Industrial applications of holonic and multi-agent systems:
adaptive manufacturing with digital twins. In 2021 International symposium on
9th International conference, HoloMAS 2019, Linz, Austria, August 26–29, 2019,
software engineering for adaptive and self-managing systems (pp. 156–166). IEEE.
Proceedings 9 (pp. 52–64). Springer.
Busse, A., Gerlach, B., Lengeling, J. C., Poschmann, P., Werner, J., & Zarnitz, S.
Government of Canada (2022). Final report of the national supply chain
(2021). Towards digital twins of multimodal supply chains. Logistics, 5, 25. http:
task force 2022. Accessed 10-November-2022, URL chrome-extension:
//dx.doi.org/10.3390/logistics5020025.
//efaidnbmnnnibpcajpcglclefindmkaj/https://tc.canada.ca/sites/default/files/2022-
Campos, J., Sharma, P., Cañizo, M., Jantunen, E., Baglee, D., Charramendieta, S., et al.
10/supply-chain-task-force-report_2022.pdf.
(2020). The use of relational and nosql databases in industrial asset management.
Grieves, M. (2022). Intelligent digital twins and the development and management
In Engineering assets and public infrastructures in the age of digitalization: Proceedings
of complex systems [version 1; peer review: 4 approved]. Digital Twin, 2(8),
of the 13th world congress on engineering asset management (pp. 302–308). Springer.
http://dx.doi.org/10.12688/digitaltwin.17574.1.
Canedo, A. (2016). Industrial IoT lifecycle via digital twins. In Proceedings of the eleventh
IEEE/ACM/IFIP international conference on hardware/software codesign and system Grieves, M., & Vickers, J. (2017). Digital twin: Mitigating unpredictable, undesirable
synthesis (p. 29). ACM. emergent behavior in complex systems. Transdisciplinary Perspectives on Complex
Systems: New Findings and Approaches, 85–113.
Carlan, V., & Vanelslander, T. (2021). Economic aspects of introducing artificial
intelligence solutions in logistics and port sectors: The data entry case. Frontiers Grosse, E. H. (2023). Application of supportive and substitutive technologies in manual
in Future Transportation, 2, Article 710330. warehouse order picking: A content analysis. International Journal of Production
Centre for Digital Built Britain (2022). Digital twins. Accessed 28-June-2022, URL Research, 1–20.
https://www.cdbb.cam.ac.uk/research/digital-twins. Guan, D., Wang, D., Hallegatte, S., Davis, S. J., Huo, J., Li, S., et al. (2020). Global
Chen, Z., & Huang, L. (2021). Digital twins for information-sharing in remanufacturing supply-chain effects of COVID-19 control measures. Nature Human Behaviour, 4(6),
supply chain: A review. Energy, 220, Article 119712. 577–587.
Cheng, J., Yang, Y., Zou, X., & Zuo, Y. (2022). 5G in manufacturing: A literature review Gunessee, S., & Subramanian, N. (2020). Ambiguity and its coping mechanisms in sup-
and future research. International Journal of Advanced Manufacturing Technology, ply chains lessons from the Covid-19 pandemic and natural disasters. International
1–23. Journal of Operations & Production Management.
Cho, D. W., Lee, Y. H., Ahn, S. H., & Hwang, M. K. (2012). A framework for Gutierrez-Franco, E., Mejia-Argueta, C., & Rabelo, L. (2021). Data-driven methodology
measuring the performance of service supply chain management. Computers & to support long-lasting logistics and decision making for urban last-mile operations.
Industrial Engineering, 62(3), 801–818. Sustainability, 13(11), 6230.
Chowdhury, P., Paul, S. K., Kaisar, S., & Moktadir, M. A. (2021). COVID-19 pandemic Ha, H. M., & Tran, T. P. (2023). Physical internet towards sustainable supply chain:
related supply chain studies: A systematic review. Transportation Research Part E: The significance of digital interoperability. Business & IT, XIII(1), 129–140.
Logistics and Transportation Review, 148, Article 102271. http://dx.doi.org/10.14311/bit.2023.01.15.
Cichosz, M., et al. (2018). Digitalization and competitiveness in the logistics service Hankel, M., & Rexroth, B. (2015). The reference architectural model industrie 4.0 (rami
industry. e-mentor, 77(5), 73–82. 4.0). ZVEI, 2(2), 4–9.
Cirullies, J., & Schwede, C. (2021). On-demand shared digital twins – an information Haraguchi, M., & Lall, U. (2015). Flood risks and impacts: A case study of thai-
architectural model to create transparency in collaborative supply networks. In land’s floods in 2011 and research questions for supply chain decision making.
Proceedings of the annual Hawaii international conference on system sciences. http: International Journal of Disaster Risk Reduction, 14, 256–272.
//dx.doi.org/10.24251/hicss.2021.202. IBM (2022). What is digital twin?. Accessed 18-November-2022, URL https:
Dai, Y., Zhang, K., Maharjan, S., & Zhang, Y. (2021). Deep reinforcement learning //www.ibm.com/topics/what-is-a-digital-twin#:~:text=A%20digital%20twin%
for stochastic computation offloading in digital twin networks. IEEE Transactions 20is%20a,reasoning%20to%20help%20decision%2Dmaking.
on Industrial Informatics, 17(7), 4968–4977. http://dx.doi.org/10.1109/TII.2020. IEA (2022). World energy outlook 2022. Paris https://www.iea.org/reports/world-
3016320. energy-outlook-2022, License: CC BY 4.0 (report); CC BY NC SA 4.0 (Annex A).
17
Jeong, Y., Flores-Garcia, E., & Wiktorsson, M. (2020). A design of digital twins for Nguyen, H. D., Tran, K. P., Thomassey, S., & Hamad, M. (2021). Forecasting and
supporting decision-making in production logistics. 2020-December, (pp. 2683– anomaly detection approaches using LSTM and lstm autoencoder techniques with
2694). Institute of Electrical and Electronics Engineers Inc., [ISSN: 08917736] ISBN: the applications in supply chain management. International Journal of Information
9781728194998, http://dx.doi.org/10.1109/WSC48552.2020.9383863, Management, 57, Article 102282.
Kajba, M., Jereb, B., & Obrecht, M. (2023). Considering IT trends for modelling Nikolopoulou, A., & Ierapetritou, M. G. (2012). Hybrid simulation based optimization
investments in supply chains by prioritising digital twins. Processes, 11(1), 262. approach for supply chain management. Computers & Chemical Engineering, 47,
Kalaboukas, K., Kiritsis, D., & Arampatzis, G. (2023). Governance framework for 183–193.
autonomous and cognitive digital twins in agile supply chains. Computers in Nvidia (2023). Nvidia omniverse for digital twins. Accessed 17-June-2023, URL https:
Industry, 146, Article 103857. //www.nvidia.com/en-us/omniverse/solutions/digital-twins/,
Kamble, S. S., Gunasekaran, A., Parekh, H., Mani, V., Belhadi, A., & Sharma, R. (2022). O’Dwyer, E., Pan, I., Charlesworth, R., Butler, S., & Shah, N. (2020). Integration
Digital twin for sustainable manufacturing supply chains: Current trends, future of an energy management tool and digital twin for coordination and control of
perspectives, and an implementation framework. Technological Forecasting and Social multi-vector smart energy systems. Sustainable Cities and Society, 62, Article 102412.
Change, 176, Article 121448. OECD (2022). OECD Economic Outlook, Interim Report March 2022 (p. 16). OECD,
Kegenbekov, Z., & Jackson, I. (2021). Adaptive supply chain: Demand–supply URL https://www.oecd-ilibrary.org/economics/oecd-economic-outlook/volume-
synchronization using deep reinforcement learning. Algorithms, 14(8), 240. 2022/issue-2_4181d61b-en.
Khan, S., Khan, S., & Aftab, M. (2015). Digitization and its impact on economy. Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C.
International Journal of Digital Library Services, 5(2), 138–149. D., et al. (2021). The PRISMA 2020 statement: an updated guideline for reporting
Kittur, A., Nickerson, J. V., Bernstein, M., Gerber, E., Shaw, A., Zimmerman, J., et al. systematic reviews. International Journal of Surgery, 88, Article 105906.
(2013). The future of crowd work. In Proceedings of the 2013 conference on computer Pan, S., Zhou, W., Piramuthu, S., Giannikas, V., & Chen, C.
supported cooperative work (pp. 1301–1318). ACM. Park, K. T., Son, Y. H., Ko, S. W., & Noh, S. D. (2021). Digital twin and reinforcement
Klar, R., Fredriksson, A., & Angelakis, V. (2023). Digital twins for ports: Derived from learning-based resilient production control for micro smart factory. Applied Sciences,
smart city and supply chain twinning experience. arXiv preprint arXiv:2301.10224. 11(7), 2977.
Layaq, M. W., Goudz, A., Noche, B., & Atif, M. (2019). Blockchain technology as a risk Parviainen, P., Tihinen, M., Kääriäinen, J., & Teppola, S. (2017). Tackling the
mitigation tool in supply chain. International Journal of Transportation Engineering digitalization challenge: how to benefit from digitalization in practice. International
and Technology, 5(3), 50–59. Journal of Information Systems and Project Management, 5(1), 63–77.
Lee, J., Lapira, E., Bagheri, B., & Kao, H.-A. (2013). Recent advances and trends in Qian, L., Luo, Z., Du, Y., & Guo, L. (2009). Cloud computing: An overview. In Cloud
predictive manufacturing systems in big data environment. Manufacturing Letters, computing: First international conference, CloudCom 2009, Beijing, China, December
1(1), 38–41. 1-4, 2009. Proceedings 1 (pp. 626–631). Springer.
Lee, D., & Lee, S. (2021). Digital twin for supply chain coordination in modular
Qiao, G., & Riddick, F. (2004). Modeling information for manufacturing-oriented
construction. Applied Sciences, 11(13), 5909.
supply-chain simulations. In Proceedings of the 2004 winter simulation conference,
Lee, J. M.-y., & Wong, E. Y.-c. (2021). Suez canal blockage: an analysis of legal impact,
2004, vol. 2 (pp. 1184–1188 vol.2). http://dx.doi.org/10.1109/WSC.2004.1371446.
risks and liabilities to the global supply chain. In MATEC web of conferences, vol.
Rahman, N. S. F. A., Hamid, A. A., Sahin, B., Wang, M., & Islam, S. (2022). A new
339. EDP Sciences.
human capital development framework in logistics and supply chain incorporating
Leung, E. K., Lee, C. K. H., & Ouyang, Z. (2022). From traditional warehouses to
industry 4.0. International Journal of Applied Logistics (IJAL), 12(1), 1–17.
physical internet hubs: A digital twin-based inbound synchronization framework
Rai, A., Patnayakuni, R., & Seth, N. (2006). Firm performance impacts of digitally
for PI-order management. International Journal of Production Economics, 244, Article
enabled supply chain integration capabilities. MIS Quarterly, 225–246.
108353.
Redelinghuys, A., Basson, A. H., & Kruger, K. (2020). A six-layer architecture for
Lind, M., Becha, H., Watson, R., Kouwenhoven, N., Zuesongdham, P., & Baldauf, U.
the digital twin: A manufacturing case study implementation. Journal of Intelligent
(2020). Digital twins for the maritime sector. Smart Maritime Network, 10.
Manufacturing, 31, 1383–1402.
Liu, J., Yeoh, W., Qu, Y., & Gao, L. (2022). Blockchain-based digital twin for supply
Rosen, R., Von Wichert, G., Lo, G., & Bettenhausen, K. D. (2015). About the importance
chain management: State-of-the-art review and future research directions. arXiv
of autonomy and digital twins for the future of manufacturing. IFAC-PapersOnLine,
preprint arXiv:2202.03966.
48(3), 567–572.
Lu, W., Olofsson, T., Jensen, P., & Simonsson, P. (2011). BIM-based lean-agile supply
Rosin, A. F., Proksch, D., Stubner, S., & Pinkwart, A. (2020). Digital new ventures:
chain for industrialized housing. In International conference on construction applica-
Assessing the benefits of digitalization in entrepreneurship. Journal of Small Business
tions of virtual reality: 03/11/2011-04/11/2011 (pp. 262–270). Bauhaus-Universität
Strategy, 30(2), 59–71.
Weimar.
Rudskoy, A., Ilin, I., & Prokhorov, A. (2021). Digital twins in the intelligent transport
Lv, Z., Chen, D., Feng, H., Wei, W., & Lv, H. (2022). Artificial intelligence in underwater
systems. Transportation Research Procedia, 54, 927–935.
digital twins sensor networks. ACM Transactions on Sensor Networks, 18(3), 1–27.
Lv, Z., Li, Y., Feng, H., & Lv, H. (2022). Deep learning for security in digital twins Sabahi, S., & Parast, M. M. (2020). Firm innovation and supply chain resilience:
of cooperative intelligent transportation systems. IEEE Transactions on Intelligent A dynamic capability perspective. International Journal of Logistics Research and
Transportation Systems, 23(9), 16666–16675. http://dx.doi.org/10.1109/TITS.2021. Applications, 23(3), 254–269.
3113779. Sabbagh, K., Friedrich, R., El-Darwiche, B., Singh, M., Ganediwalla, S., & Katz, R.
Lv, Z., Qiao, L., Li, Y., Yuan, Y., & Wang, F.-Y. (2022). Blocknet: Beyond reliable spatial (2012). Maximizing the impact of digitization. The Global Information Technology
digital twins to parallel metaverse. Patterns, 3(5). Report, 2012, 121–133.
Lyovin, B. A., & Efimova, O. V. (2017). Digital logistics and electronic data exchange Sandberg, E. (2021). Dynamic capabilities for the creation of logistics flexibility–A
in freight transportation. World of Transportation, 15(2), 142–149. conceptual framework. The International Journal of Logistics Management, 32(2),
Machining4. 0 (2023). Digital twin applications in manufacturing | machining4.eu. Ac- 696–714.
cessed 28-May-2023, URL https://www.machining4.eu/technology-sheets/digital_ Schluse, M., & Rossmann, J. (2016). From simulation to experimentable digital twins:
twin_ts. Simulation-based development and operation of complex technical systems. In 2016
Mao, H., & Görg, H. (2020). Friends like this: The impact of the US–China trade war IEEE international symposium on systems engineering (pp. 1–6). IEEE.
on global value chains. The World Economy, 43(7), 1776–1791. Schroeder, G. N., Steinmetz, C., Pereira, C. E., & Espindola, D. B. (2016). Digital twin
Maré, D. C. (2004). What do endogenous growth models contribute? data modeling with automationml and a communication methodology for data
Marinagi, C., Reklitis, P., Trivellas, P., & Sakas, D. (2023). The impact of industry exchange. IFAC-PapersOnLine, 49(30), 12–17.
4.0 technologies on key performance indicators for a resilient supply chain 4.0. Schwarz, C., & Wang, Z. (2022). The role of digital twins in connected and automated
Sustainability, 15(6), 5185. vehicles. IEEE Intelligent Transportation Systems Magazine, 14(6), 2–11. http://dx.doi.
Marmolejo-Saucedo, J. A. (2020). Design and development of digital twins: A case org/10.1109/MITS.2021.3129524.
study in supply chains. Mobile Networks and Applications, 25(6), 2141–2160. Shahat, E., Hyun, C. T., & Yeom, C. (2021). City digital twin potentials: A review
Marmolejo-Saucedo, J. A. (2022). Digital twin framework for large-scale optimization and research agenda. Sustainability, [ISSN: 2071-1050] 13(6), http://dx.doi.org/10.
problems in supply chains: A case of packing problem. Mobile Networks and 3390/su13063386, URL https://www.mdpi.com/2071-1050/13/6/3386.
Applications, 27(5), 2198–2214. Shekari, H., & Rajabzadeh Ghatari, A. (2013). Promoting corporate image: A reflection
Mishra, R., Singh, R. K., & Subramanian, N. (2022). Impact of disruptions in agri-food on green supply chain management approach. International Journal of Management
supply chain due to COVID-19 pandemic: contextualised resilience framework to and Business Research, 3(4), 311–324.
achieve operational excellence. The International Journal of Logistics Management, Shen, X., Zhang, Y., Tang, Y., Qin, Y., Liu, N., & Yi, Z. (2022). A study on the impact of
33(3), 926–954. digital tobacco logistics on tobacco supply chain performance: Taking the tobacco
Moshood, T. D., Nawanir, G., Sorooshian, S., & Okfalisa, O. (2021). Digital twins driven industry in guangxi as an example. Industrial Management & Data Systems, 122(6),
supply chain visibility within logistics: A new paradigm for future logistics. Applied 1416–1452.
System Innovation, [ISSN: 25715577] 4, http://dx.doi.org/10.3390/asi4020029. Shrivastava, C., Schudel, S., Shoji, K., Onwude, D., da Silva, F. P., Turan, D., et al.
Mylonas, G., Kalogeras, A., Kalogeras, G., Anagnostopoulos, C., Alexakos, C., & (2023). Digital twins for selecting the optimal ventilated strawberry packaging
Muñoz, L. (2021). Digital twins from smart manufacturing to smart cities: A survey. based on the unique hygrothermal conditions of a shipment from farm to retailer.
Ieee Access, 9, 143222–143249. Postharvest Biology and Technology, 199, Article 112283.
18
Singh, M., Fuenmayor, E., Hinchy, E. P., Qiao, Y., Murray, N., & Devine, D. (2021). Wu, J., Wood, J., Oh, K., & Jang, H. (2021). Evaluating the cumulative impact of
Digital twin: Origin to future. Applied System Innovation, 4(2), 36. the US–China trade war along global value chains. The World Economy, 44(12),
Steinbock, D. (2018). US-China trade war and its global impacts. China Quarterly of 3516–3533.
International Strategic Studies, 4(04), 515–542. Xu, D., Qi, J., et al. (2022). Comprehensive evaluation method of supply chain logistics
The White House (2022). Fact sheet: Biden-harris administration announces new system quality based on 3D image processing technology. Advances in Mathematical
initiative to improve supply chain data flow. URL https://www.whitehouse. Physics, 2022.
gov/briefing-room/statements-releases/2022/03/15/fact-sheet-biden-harris- Yang, J., Lee, S., Kang, Y.-S., Do Noh, S., Choi, S. S., Jung, B. R., et al. (2020).
administration-announces-new-initiative-to-improve-supply-chain-data-flow/. Integrated platform and digital twin application for global automotive part sup-
(Accessed 28 June 2022). pliers. In Advances in production management systems. Towards smart and digital
Tuegel, E. manufacturing: IFIP WG 5.7 international conference, APMS 2020, Novi Sad, Serbia,
Tzachor, A., Sabri, S., Richards, C. E., Rajabifard, A., & Acuto, M. (2022). Potential and August 30–September 3, 2020, Proceedings, Part II (pp. 230–237). Springer.
limitations of digital twins to achieve the Sustainable Development Goals. Nature Yao, W., Sun, J., & Wu, S. (2022). Construction and simulation research of logistics
Sustainability, 5(10), 822–829. automation stereoscopic storage system based on digital twin. In 2nd International
van Eck, N. J., & Waltman, L. (2017). Vosviewer manual. Accessed 25-May-2023, URL conference on applied mathematics, modelling, and intelligent computing (CAMMIC
https://www.vosviewer.com/documentation/Manual_VOSviewer_1.6.6.pdf, 2022), vol. 12259 (pp. 848–852). SPIE.
Vilas-Boas, J. L., Rodrigues, J. J., & Alberti, A. M. (2022). Convergence of distributed Yao, H., Wang, D., Su, M., & Qi, Y. (2021). Application of digital twins in port system.
ledger technologies with digital twins, IoT, and AI for fresh food logistics: Journal of Physics: Conference Series, 1846(1), Article 012008.
Challenges and opportunities. Journal of Industrial Information Integration, Article Ye, L., & Abe, M. (2012). The impacts of natural disasters on global supply chains.
100393. Tech. rep., ARTNeT working paper series.
Wang, L., Deng, T., Shen, Z.-J. M., Hu, H., & Qi, Y. (2022). Digital twin-driven smart Yigitbas, E., Karakaya, K., Jovanovikj, I., & Engels, G. (2021). Enhancing human-in-the-
supply chain. Frontiers of Engineering Management, [ISSN: 2095-7513] 9, 56–70. loop adaptive systems through digital twins and VR interfaces. In 2021 International
http://dx.doi.org/10.1007/s42524-021-0186-9. symposium on software engineering for adaptive and self-managing systems (SEAMS),
Wang, Z., Gupta, R., Han, K., Wang, H., Ganlath, A., Ammar, N., et al. (2022). (pp. 30–40). IEEE.
Mobility digital twin: Concept, architecture, case study, and future challenges. Yuan, R., Rodrigues, J. F., Wang, J., & Behrens, P. (2023). The short-term impact of
IEEE Internet of Things Journal, 9(18), 17452–17467. http://dx.doi.org/10.1109/ US-China trade war on global GHG emissions from the perspective of supply chain
JIOT.2022.3156028. reallocation. Environmental Impact Assessment Review, 98, Article 106980.
Wang, Q., Jiao, W., Wang, P., & Zhang, Y. (2020). Digital twin for human-robot Zdolsek Draksler, T., Cimperman, M., & Obrecht, M. (2023). Data-driven supply chain
interactive welding and welder behavior analysis. IEEE/CAA Journal of Automatica operations—The pilot case of postal logistics and the cross-border optimization
Sinica, 8(2), 334–343. potential. Sensors, 23(3), 1624.
Wang, Y., Wang, S., Yang, B., Zhu, L., & Liu, F. (2020). Big data driven hierarchical Zhao, Z., Zhang, M., Chen, J., Qu, T., & Huang, G. Q. (2022). Digital twin-enabled
digital twin predictive remanufacturing paradigm: Architecture, control mechanism, dynamic spatial-temporal knowledge graph for production logistics resource
application scenario and benefits. Journal of Cleaner Production, [ISSN: 0959-6526] allocation. Computers & Industrial Engineering, 171, Article 108454.
248, Article 119299. http://dx.doi.org/10.1016/j.jclepro.2019.119299, URL https: Zheng, X., Lu, J., & Kiritsis, D. (2022). The emergence of cognitive digital twin: vision,
//www.sciencedirect.com/science/article/pii/S0959652619341691. challenges and opportunities. International Journal of Production Research, 60(24),
Wells, D. (2023). From analysis to action: Data-driven vs. Data-informed. (Accessed 2- 7610–7632.
June-2023), URL https://www.eckerson.com/articles/from-analysis-to-action-data- Zhou, R., Awasthi, A., & Stal-Le Cardinal, J. (2021). The main trends for multi-tier
driven-vs-data-informed, supply chain in industry 4.0 based on natural language processing. Computers in
Industry, 125, Article 103369.
19

Paper 1 - Digital Twins

Uploaded by

Copyright:

Available Formats

Paper 1 - Digital Twins

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Paper 1 - Digital Twins

Uploaded by

Copyright:

Available Formats

Computers & Industrial Engineering 187 (2024) 109768

Contents lists available at ScienceDirect

Computers & Industrial Engineering

ARTICLE INFO ABSTRACT

Fig. 1. Outline of research scope.

Fig. 3. Publications on DT in logistics and supply chain systems by year.

Fig. 4. All Author Keywords related to methodology.

Fig. 5. High-Level Literature Clustering (All Methodology-related Keywords).

Fig. 6. Conceptual DT for logistics and supply chain systems.

7. Conclusions and the components in each dimension guide the DT construction

You might also like