Case Studies on Transport Policy 9 (2021) 1298–1312

Contents lists available at ScienceDirect

Case Studies on Transport Policy

journal homepage: www.elsevier.com/locate/cstp

Multi-aspect applications and development challenges of digital

twin-driven management in global smart ports
Kan Wang a, b, c, *, Qianqian Hu a, b, Mingjun Zhou a, Zhou Zun a, b, Xinming Qian d, *
a
College of Ocean Science and Engineering, Shanghai Maritime University, Shanghai, China
b
Department of Safety Science and Engineering, Shanghai Maritime University, China
c
Marine Public Safety Research Center, Shanghai Maritime University, Shanghai, China
d
School of Mechatronical Engineering, Beijing Institute of Technology, Beijing, China

A R T I C L E I N F O A B S T R A C T

Keywords: Digital twin (DT) creates a revolutionary opportunity for the smart ports’ authorities, with the capability of high-
Digital twin fidelity digital representation of real-world things. DT-driven management is able to predict potential risk, it is
Smart port considered to be a key integration involves transport and operation related with smart port supply chains. This
Visual-based management
paper presents DT information visualizations & outcome loop of global smart ports’ management, and core
Digital twin model
Systematic framework
technique towards smart ports’ processes based on DT-driven management. A systematic framework of DT-based
model for smart port’s management is presented, which can be categorized into physical layer, data layer, model
layer, service layer and application layer. Decision-makers engage in operational management through the DT-
based models and platform. To illustrate DT mechanism, multi-aspect functional applications of DT-driven
management are discussed, including cargo handling and transportation, containers operation and storage,
data communication and sharing, risk prediction, and environmental protection in global smart ports. Mean­
while, this paper proposes concerned challenges of DT-driven management integrated with optimized cross-
platform in smart ports’ processes, to indicate there are barriers ahead to get to an ideal visual-based manage­
ment system. Current challenges help decision-makers to determine where to find a proper strategy to win, in the
future, DT visual applications can be extended in several directions of global port industry. This work contributes
to address a comprehensive overview of DT managing capabilities together with details of necessary architecture
to support effective simulations in virtual port’s space. It provides DT-driven platform will enable better man­
agement across the smart port activities from design to transport to operation and sustainment of innovative
plan.

exact mathematical and virtual replica of a real scenario, including its

physical geometry, design characteristics, related parameters. This
1. Introduction digital model also interacts with the physical world to which the result
will ultimately be exposed (Tsega et al., 2020). Those interactions may
Smart port uses innovative technologies such as “Digital Twin” (DT) contain ambient temperature, humidity, air flow, water flow, chemical
in its management, it can help the port industry make better perfor­ reaction, electric current, etc. The model has a real-time connection with
mance of digitalization and move to a new era. DT-driven management the associated sensors that continuously transmit the conditions they are
in smart port provides a significant research direction for realizing living monitoring. Its algorithms allow computer servers to evaluate what is
spaces with visual monitoring, visual interaction and automation (Al happening in real smart port’s environment, and continuously run “what
Dakheel et al., 2020). To meet the increasing smart port requirements of if” scenarios of prediction or decision emerging next in port processes
management, the port authorities begin seeking for a range of digitali­ (Ritto and Rochinha, 2021). The suite of DT that enables model promises
zation and data-enabled virtual technologies. In aggregate, these tech­ to turn almost any smart port’s object into a source of information about
nologies form a DT-based model (Nacera et al., 2016). There are objects. This creates a way to manage port’s process and activities, and a
multiple different understandings of DT-based models (Tao et al., 2021; novel source of value that can be managed in its own right. With DT-
Cheng et al., 2020); in port industrial sector, a DT-based model is an

* Corresponding authors at: Address: No.1550, Haigang Ave., Shanghai, China.

E-mail addresses: [email protected] (K. Wang), [email protected] (X. Qian).

https://doi.org/10.1016/j.cstp.2021.06.014
Received 26 April 2021; Received in revised form 7 June 2021; Accepted 28 June 2021
Available online 4 July 2021
2213-624X/© 2021 World Conference on Transport Research Society. Published by Elsevier Ltd. All rights reserved.
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Nomenclature WDR Wide Dynamic Range

AI Artificial Intelligence
Acronym API Application Programming Interface
AES All Element Scene BIM Building Information Model
AIS Automatic Identification System CFD Computational Fluid Dynamics
AR Augmented Reality DTM Digital Terrain Model
BOM Bill of Materials ID Identity Document
DT Digital Twin IEC International Electrotechnical Commission
GPU Graphics Processing Unit ISO International Organization for Standardization
ICT Information and Communication Technology LGV Laser Guided Vehicle
IEEE Institute of Electrical and Electronics Engineers LTE Long Term Evolution
IoT Internet of Thing PDM Product Data Management
LiDAR light Detection and Ranging PSC Port Supply Chain
MoNArch Mobile Network Architecture PTO Port Transportation and Operation
PLM Product Lifecycle Management RFID Radio Frequency Identification
PTC Parametric Technology Corporation VR Virtual Reality
QR Quick Response WIP Work In Progress
UAV Unmanned Aerial Vehicle

driven management’ full potential being realized, a circle procedure that organization for standardization (ISO) encourage port industry to sup­
captures the sequence of activities is motivated, by which organizations port these formal managements to develop inputs for standardization
create outcome from information (Shao and Helu, 2020). As shown in (ISO/TC, n.d.). The significant meaning of DT standardization is to
Fig. 1, for digital information to complete the loop and create value, it define application programming interfaces (APIs) for DT-driven data
has to go through some essential stages. An actual object is monitored by access to securely and reliably collection, manage and retrieve records
related sensors that create data. The data passes through a network (Harper et al., 2019). The augmented intelligence is a generic term
system so that it can be communicated. DT-driven standards are meant to capture all manners of analytical support and collectively used
involving technical, legal, regulatory, or social elements, which allow to analyze data in smart port. Last, DT-based model will be completed by
the digital information to be aggregated across time and space in DT augmented behaviors, this technology that either enables automated
platform (Wang and Luo, 2021). The standards for a DT technologies are action or shapes decisions in a manner leading to improved management
under development (ISO/FDIS, n.d.). For example, the association of (Falah et al., 2020).
connecting electronics industries officially released International Stan­ Using DT-based models, the visualization performance and capacity
dard for Digital Twins (IPC-2551) (IPC, n.d.). IPC-2551 is the first in­ expansion of smart port’s scenario and objects can be realized through
ternational standard on DT, which targets the creation of smart supply lifecycle management (Zhang et al., 2019; Jones et al., 2020). Advances
chains with standardized data formats. Meanwhile, the international in Internet of Things (IoTs) (Jiang et al., 2021; Mathupriya et al., 2020)

Fig. 1. Visual information outcome loop of proposed management based on DT technologies.

1299
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

space are important for creating an ideal DT-based model. Sensors and the development of DT-driven management with its models is proposed,
IoTs embedded in physical objects of smart port and attached to staff can as shown in Fig. 2. In the process of realizing visual management, DT-
gather massive real-time data that allows DT-based model to mimic, and driven system can be categorized into five layers, namely, the physical
iteratively improves all interactions between actual objects and people. layer, model layer, data layer, service layer and application layer. Many
Combining building information modeling (BIM) to form a digital aspects have been involved, the physical layer contains smart port’s
terrain model (DTM), before matching it with related algorithms to transport and operation, which include the items that are expected to be
improve visualization performance of building is a direction worth covered in whole process of smart port (Rüßmann et al., 2015). Data
considering (Liu et al., 2020). However, there has been little research on layer is based on real-time data collected by sensors, combined with
applications of DT-based models into reinforcement managing of smart historical data to provide a database for the construction of DT-based
ports’ processes. Becoming a smart port means developing a solution to models in this layer. Service layer covers DT-based cloud server,
handle various future challenges including spatial constraints, pressure including monitoring of vessels and smart port status, classification of
on productivity, fiscal limitations, environmental sustainability, security accident types and risk levels, early warning of accident risks and sug­
risks and digital management. Today’s DT technology and its models are gestion of accident treatment. DT-based platform can deliver external
able to serve as a driving force behind smart port (Inkinen et al., 2021). output, which provides APIs or external calls with related authentica­
This paper aims to make an introductory research for DT-driven man­ tion. Meanwhile, DT-based platform affords the visualization of required
agement development in field of port supply chain (PSC). In global PSC, data, thereby assisting in virtual on-site support throughout smart port’s
the smart ports are veritable beneficiaries of digital process with DT- lifecycle management.
based model, for instance, DT can be used for deploying autonomous
pilots on laser guided vehicle (LGV) operating with fewer human in­ 2.2. System characteristics
terventions in container yard (Lin et al., 2021). Herein, DT-enabled
framework for management in smart ports is proposed, its systematic Digital models make use of a large amount of data monitored in real
characteristics and implementation steps are elaborated. This work fo­ world to dynamically present a certain physical individual, region or
cuses on a deep analysis of multi-aspect applications in smart ports with process in form of digitization. Effective feedbacks of DT system’s
regard to cargo handling, container storage, data communication, operating status allows the manager to take timely emergency actions on
optimized prediction, environmental sustainability. These DT-driven top of forecast information. The main body of DT technology adopted for
applications in management have been developed in some world- port transportation and operation (PTO) is the DT-based model, while its
leading smart ports. Understanding specific value of DT technology to core is effective twinning data. The models provide some data, while
smart port’s processes contribute to grasping potential high perfor­ those visual data is an integral part of DT-based model and functions as a
mance for better management in future. Further, the current challenges driving force (He et al., 2021). Therefore, whether DT-driven manage­
of performing a DT-driven management are presented, and these chal­ ment is realized in smart port, its system needs have following
lenges guidance future works concerning DT applications of manage­ characteristics:
ment in smart ports. It is expected DT-driven management would lead to
a statistical reduction in risks and incidents of smart ports. While the • Omni-directional virtual and real integration. Hyper realistic
techniques of creating the complex DT-based platform are coming down, models of physical entities are constructed, before a complete map­
the global port industry will gain a capable visualization ally in their ping of basic attributes of each physical entity (e.g. visual structure,
quest for a sophisticated management. geometric characteristics, material properties, performance param­
eters, behavior and rules). DT-based models provide real-time
2. Methodology interactive feedback to entity in real scene. The emergence of DT-
based system serves to break the barrier between physical world
2.1. Key framework of DT-driven management and virtual space, supplying digital information for decision makers
and managers in port industry (Coelho et al., 2021). This information
Physical entities, virtual models, data, integration methods, analysis helps to understand the state of real objects in smart port and
methods, and services are the core drivers of DT in its process for real­ broaden the boundary of object itself. On the other hand, smart port
izing a full-cycle process of smart port. Distributed sensors are deployed entities continuously provide valid data through sensors. It is more
throughout the smart port’s area to collect handling, operational and convenient for operators to carry out productions. Data-based pre­
environmental data related to the actual processes and activities. The dictions provide basis and direction for improvement of smart port
data types include historical data of smart port management and real- process.
time data from sensors, such as the inventory inside a materials • Data-driven in whole process. DT-based models can continuously
container. The integration method mainly uses communication interface track port activities links from loading, unloading, storage to light­
to realize data transmission between physical objects (e.g. smart ports) ening, meanwhile integrates real-time data (e.g. cargo attributes,
and digital models through integration technology. Technique in DT- transport, operation tool characteristics and environmental param­
driven management is to carry out data analysis through relevant al­ eters), thus facilitating DT-driven management through data fusion
gorithms and visualization programs, so as to realize the simulation and under models (Zhuang et al., 2021).
prediction of DT-based models. DT is partly designed to find smart port • Fusion of All Element Scene (AES). Data involving all elements,
operations deviation existing in the simulation process and return its scenarios, processes, services, and states of PTOs can be integrated
actual operations to the correct direction by optimizing those deviations. and fused. Data from operators, equipment, materials and environ­
If action is urgently needed during a real process, DT-based models ment, etc. can be transmitted into virtual space of DT-based system
generate signals or instructions through actuators within its platform to for integration and fusion, this system is characterized by the inte­
trigger physical behavior, which is subject to human intervention. gration of all elements, realizing real-time and historical data sharing
Obviously, the visualizations of DT-based models for actual smart port while ensuring the interaction among elements (Chakraborti et al.,
processes are extremely complicated. The DT-based models aim to fully 2020). The visualization data link formed during smart port process
demonstrate the matching, integration, and dynamic simulation of the of each element from the initial state to final state effectively sup­
physical and digital worlds. Decision-makers engage in actual manage­ ports cooperative PTOs. In addition, the performance index data
ment and emergency decision making through DT-based models and (port signal equipment, changes in environmental parameters in
their platforms. storage areas, etc.) (Eric and Mahadevan, 2021) help mitigate in­
According to above-mentioned concept, a systematic framework for formation isolation in smart port process through integration.

1300
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Fig. 2. Key framework of the DT-driven management in smart port processes.

• Overall optimization and upgrade. Twining data generated by DT- Akbarieh et al., 2020; Heaten et al., 2019), which stands for DT
based system covers all from PTOs lifecycle. Based on DT-based collaboration in built environments. BIM integrates the management
models, the virtual verification design scheme, transportation and and DT construction information throughout lifecycle of a smart
operation plan, visualization performance and function, manage­ port. When 3D models are elevated to DT, BIM can convey a diversity
ment, emergency plan are proposed to provide measures and of details that all construction phases need to get work done. These
schemes for each stage. Historical data are summarized to find out georeferenced 3D models can prescribe actions based on smart port’s
different defects existing in smart port process, before upgrading the simulations, physics, chemistry, statistics, machine learning, AI,
future management, while completing the iterative upgrade of DT- business logic or objectives. 3D models can be displayed via 3D
based models and platform. representations and augmented reality (AR) modelling for better
human understandings. DT models’ core are georeferenced 3D data
collected by different surveying technologies best suitable for DT’s
2.3. Implementation steps purpose. Data post-processing and combining of different datasets
require a thorough understanding of 3D data. Most of the utilization
Using graphical representation of DT helps to improve methods for possibilities need exact information of smart port’s location, the
management progress and enhance execution of on-site construction. valid locations of assets or observations in 3D model are foundation
The implementations require an extraordinary investment of time, when managing smart port process remotely. The accuracy should be
capital, and engineering might, and as with PTOs of this scale. There are guaranteed if decision-makers aim to use data for quay condition
a total of five steps to DT-driven modeling implementation success. inspection etc. This enables maintenance crew to react to possible
risks before they become a threat to safety.
• Step I: Data acquisition. DT technologies require both historical • Step III: Integrate real-time information to model. One funda­
data and real-time concerning an object in order for a virtual model mental ability of a DT-based model is to receive real-time informa­
to be created to represent the behaviors or states of real smart port’s tion and become a more accurate model for visual reality. This model
scenario item or procedure. These data are related with smart port’s can be used to forecast future development for predictive manage­
information, including equipment, materials in container, transport ment. In the interaction between physical layer and model layer,
vehicles, signals and quality control. Data can be collected by using entities in the former play a leading role through the latter. In actual
IoTs platform, which can apply for storage and visualization of real- process, each heterogeneous hardware in physical layer will upload
time data. Definitely, the data have a link to PTOs elements, such as the real-time data concerning PTOs state to the data layer. After that,
real-time feedback, historical analysis and maintenance records. DT-based model in virtual space can update its own model based on
Other data used in DT-based models involve business data or end-of- real-time information. DT-based model carries out virtual-real link­
life procedures (Kong et al., 2021). age and monitoring on PTOs state of entity units in physical layer.
• Step II: Create a georeferenced model. Once data are gathered, it For example, the real-time data can be gathered from IoTs’ sensors
can be used to create computational models to display the operating and integrated into DT-based models. IoTs empowers DT BIM models
effects and predict states. 3D models tailored for DT concept are to track, store, and display complex data, and effectively turns
expected to not only focus on smart port’s construction but also models into “LIVING” documents that update themselves automati­
interact with built environments. Thus DT-based models need cally. Integrations of external software or tools like automatic
project-wide information delivery tool like BIM (White et al., 2021;

1301
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

identification system (AIS) data for vessels or weather stations will 3. Results and discussion
give valuable information to DT-based models. This move brings DT-
based models a lots beyond the conceptual design, due to the fact A fully-developed DT-driven management and the related DT-based
that DT-based models acts as a simulation of reality and can also be models have been prepared for use by smart port personnel to organize
utilized to manage various types of PTOs remotely, from smart port’s and “manage” their port processes with consideration of the environ­
allocation to management (Kuitunen, 2019). A plain visual repre­ ment in which those operations occur. The general port processes
sentation needs fleshes on framework to become fully utilizable in contain the cargo (e.g. dry bulk) handling & storage, cargo transport
PTOs. The attributes information attached to 3D data include tech­ (loading & unloading), non-bulk chemical storage & operation, cargo
nical information, data on users and ownership of an asset, guide­ handling equipment, rail & truck operations, vehicle & equipment
lines, rules, and more. A cloud-based management server enables an fueling, information networks, building & grounds maintenance, and
availability of information to decision-makers. This may be the most environment management of hazardous and non-hazardous waste
laborious phase of digitalization, and more often than not, infor­ generated by port activities.
mation is dispersed to various software, databases and archives. The
possibility to simultaneously upload a variety of information is 3.1. DT-driven management for cargo handling and transportation
essential in minimizing the workload of building a DT-based model
(Hu et al., 2018). Cargo handling and lightering is the most basic business operated on
• Step IV: Simulation, analysis & optimization. Simulation is an international smart ports. DT-based models enable efficient operation of
integral part of DT-based models, the purpose of which stretch cranes and other cargo handling equipment, but also ensure that rele­
beyond simulations (Glatt et al., 2021). In the initial stage, the vant equipment is properly maintained, helping terminal operators in­
system-level models use Simplorer design software running on crease throughput and productivity. DT-driven system can assist with
computer. 3D models are connected with sensor-based data and networked cargo handling equipment to do the job in real-time. As can
mimic the hydraulic system. Meanwhile, the models are further be seen in Port of Valencia (Lind et al., 2018; Barbieri and Gutierrez,
connected to a human–machine interface developed by using 2021), whose container terminal is testing this network. In this smart
computational software with the same gauges and dials as physical port, over 200 cranes, trucks and forklifts have been fitted with high-
objects. With the corresponding configuration, 3D models can be precision sensors that gather information about the equipment’s loca­
disconnected from the physical objects and operated offline to tion, PTOs status and energy consumption. The improving the efficiency
explore operating scenarios in smart port process. Different from of cargo handling is not the only contribution of DT-based models to port
traditional simulations, DT-based models are used for simulating the freight. The DT-based models can properly guide freight trucks and
entire design, execute, change, and decommission of lifecycle in real- trains in ports and optimize the congestion path. Another example as
time. DT-based model knows everything regarding with smart port’s Port of Singapore, this smart port is testing a traffic monitoring system
moment, the vital performance statistics collected on a regular basis, based on this model (Ivars-Baidal et al., 2021). Its system is able to track
parts replaced in the past, previous service records, potential issues the real-time movement of a truck in Singapore port, notifies it when the
as observed by sensors. In PTOs scenario, DT-based models help vehicle is approaching its key facilities, and instructs it on how to pro­
deliver time, cost and effort savings. All operators stand to benefit ceed. The Port of Singapore, on the other hand, has installed embedded
from the enhanced and relevant information available at the traffic monitoring sensors along its main road, which are linked with its
opportune moment. DT-based models at best not only can boost platform “Orbit 3DM Content Manager” to assist in accelerating truck
crucial understanding of what may happen in real smart port process, turnover (D’Amico et al., 2021). In addition to handling cargo more
but also of how to analyze in real world (Arrichiello and Gualeni, efficiently, smart ports need to improve the handling of cargo infor­
2019). The models offer a real-time view on a human-friendly 3D mation, including trade licenses, import and export licenses, customs
visual format and platform, generating analytical results that help clearance, etc. “DT-based model + Blockchain” can reduce PTOs costs
prevent wrong decisions, conduct preventive maintenance and related to shipping documents and customs fee payment. It can
reduce untoward accidents. Those models provide insights into in­ distribute data in multiple locations, which not only speeds up infor­
teractions of physical objects, acting as a source for understanding mation processing, but also reduces risks of data tampering. “DT-based
their clashes and conflicts (He and Bai, 2021). When assets in smart model + Blockchain” can save shippers, shipping lines, ports, terminals
port intersect with other products, people and processes, various and customs authorities hundreds of dollars in labor and operating costs
unexpected behaviors may arise. Gartner posits the evolution of DT- per container. The smart port has attempted to use “DT-based model +
based models beyond assets to include whole organizations covering Blockchain” to guarantee container transport and track the status of
people, transport and operations. They help operators to optimize an goods (Alamoush et al., 2020).
interactions and incorporate into the simulations, which are further Nowadays, with the rapid development of digital information age,
expected to deliver hitherto unknown insights on PTOs before the port industry is embarking on an innovative journey of connected
deciding on a plan. shipping (Peng et al., 2019; Olba et al., 2019). Just as the driverless
• Step V: Sharing and delivering insights. It is significant to un­ technology in land transport, connected vessels in a smart port can
derstand potential benefits from DT-based models, to share analyt­ operate autonomously and communicate with each other to avoid col­
ical results that can be communicated through maps and dashboards lisions. Port of Rotterdam, covers an area of over 106.2 km2 and serves
in platform. IoTs allows models to connect themselves with oper­ nearly 30,000 sea-going vessels each year, which is undergoing a
ating processes using platform, as a result, PTOs could be better magnificent transformation. A goal of Port of Rotterdam is to enhance its
understood with optimized performance. Port authorities use DT vessels’ capacity to enter and leave the smart port autonomously in
platform to detect and isolate faults, perform diagnostics and trou­ 2030. To turn such a great vision into reality, the Port of Rotterdam
bleshooting, recommend corrective action, determine ideal mainte­ authority’s project aims to transforming it from a physical port to a
nance schedule based on specifics of individual asset, optimize asset digital one with BIM-based DT technology. Is is a common language
and generate insights that further improve smart port’s next gener­ used for exchanging digital information for models about infrastructure
ation. DT-driven platform helps decision-makers to make accurate (Witte et al., 2018; Pauwels et al., 2017). DT-driven visualizations in
and timely decisions digitally and apply them to physical network smart port area relies on 3D computer-aided design models extracted
with greater efficiency. It is potential to be a key to unlocking this from BIM or custom 3D models. However, the information provided by
additional value for both risk assessment and management (Mailly ordinary BIM is often limited to facility or building level, while lacking
et al., 2010). in real-time dynamic data information of personnel, materials and

1302
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

equipment on-site. To address this problem, IoTs technology is used to computer based on real geographic information and various elements of
enable the installation of sensors along wharf walls, mooring posts and physical world (Liu et al., 2013; D’Addona et al., 2018). Once applied to
roads, both on land and at sea, within 42 km of smart port shoreline. construction of DT-based models, AES is able to facilitate seamless
These sensors collect hydrological and meteorological data such as tide, scaling, so as to achieve fully digital restoration. AES provides full
tidal current, temperature, wind speed and direction, water level, berth support for both centimeter-level reduction and 4 K resolution, with
availability and visibility. The information of goods, equipment and high accuracy. It can realize the dynamic simulation of container, ve­
personnel on-site and transmit it to background of information system hicles, operating personnel, etc. It also synchronize the real world
through radio frequency identification (RFID) tag (Costin et al., 2018), meteorological data, the track of container vehicles and the mechanism
sensing devices, quick response (QR) code, perception technology such of cranes (Kumar et al., 2020; Szpytko and Duarte, 2019). 3D scene in
as video surveillance, thereby realizing the real-time tracking of basic DT-based models of smart port is loaded in graphics processing unit
information, location information and transportation status of moni­ (GPU) of cloud server, which realizes lifecycle management simulation
tored objects. BIM-aided DT can mobilize a variety of sensor networks coverage from cargos to storage, from loading and unloading, reloading,
within smart port to create a real-time view of models. Fig. 3 shows DT- to warehousing and departure.
based model of the Port of Rotterdam, digitization and automation allow Fig. 4 shows DT-based models and real-time monitoring platform for
collaboration in an increasingly complex environment in this smart port. Mawan port in China. This smart port’s area covering nearly 1 million
BIM and working with DT-based models (Abbassi et al., 2015; Akbarieh square meters is cloned through AES scene DT-driven platform, with an
et al., 2020) help Rotterdam port authority with information-aided actual landform and environmental features of this port accurately
management of various assets (e.g. quay walls, bollards, and etc.). The stored. Meanwhile, unmanned aerial vehicles (UAVs) and sensors are
collected data is analyzed by IBM’s IoTs cloud server, before being used to dynamically generate and synchronize the DT-based models of
converted into decision support data to reduce vessels waiting times, the containers, ultimately creating a 1:1 DT for more than 100,000
optimize berthing and cargo handling. All data related to real-time wind containers in Mawan port’s terminal. DT-driven platform provides a
speed, relative humidity, sea level, flow rate and tide can be used to complete super-APIs interface, which comprises a set of programming
predict visibility in port area on a given day, meanwhile calculate interfaces developed in JavaScript to mediate between Web pages and
headroom height of its vessels. Under favorable conditions of sea water cloud rendering server. It is used to call the scene action of cloud
and weather, the fuel consumption rate of vessels can be generally rendering server in Web application, complete DT-based 3D scene
reduced. While the payload of each vessel can be improved, thus interaction under B/S architecture and conduct two-way interaction.
ensuring the safe arrival of goods. Real-time and synchronous moni­ Super-APIs interface accesses various kinds of box information (attribute
toring data can be fused through artificial intelligence (AI) algorithm data like box owner, box type, box size, box number and other, as well as
and data mining, before predicting the hydrological and wind conditions box location, etc.). Mawan port authority enters the attributes and
to determine whether vessels smoothly enter. Further, Rotterdam au­ location data according to due requirements, after which they can
thority presented the 1st version DT application program, which com­ generate the container matching those attribute at specified location in
bines public data with data retrieved directly from participating DT-driven platform. The platform can also render the information con­
companies and forecasts from AI applications (Abaglo et al., 2017) to cerning container logo, color and container owners and print container
generate accurate management information. DT-driven management number, so as to achieve a higher legibility of information. By directly
applications allow users to work an optimal plan, execute it and risk clicking from DT-based scene, Mawan port managing personnel are able
monitor activities throughout entire port’s call. to understand all commonly used information of containers, while get­
ting rid of the tedious work of consulting various forms and searching
from the system. DT-based models give each container in Mawan port its
3.2. DT-driven management for container terminal operation own “identity card”. Combined with IoTs and spatiotemporal data, DT-
based models generally follow physical attributes and operational rules,
DT-based system has the ability to collect massive data of whole endows real-time spatiotemporal posture and motion state for the port-
process of the automated container terminal before building a big data related elements (trailers, quay Bridges, tower Bridges, ships, etc.), and
knowledge base for intelligent decision making. DT-based models can realizes a high-precision restoration. DT-driven platform provides
complete data collection after unloading the containers, also facilitate standardized API interfaces for movement of mechanical equipment of
the summary and display of container-related indicators. DT-driven quay bridge, field bridge and trailer, including function of setting
platform creates profiles for each container entering and leaving smart different motion parameters, in an effort to realize two-way communi­
port, this platform is implemented under AES scenario. AES is a kind of cation synchronization between real equipment and models. As Fig. 4
interactive 3D simulation scene generated by real-time rendering

Fig. 3. 3D BIM-based DT model of the visual route at Port of Rotterdam.

1303
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Fig. 4. DT-based models based on AES and the real-time data for Mawan port, China.

shown, every quay bridge in DT-driven platform can be selected by 3.3. DT-driven data communication and sharing
clicking. The models of equipment includes parameter settings such as
cart position, trolley position, sprig position and big arm angle. Due to DT technology makes data communicating and sharing more selec­
continuous incoming real-time data, the quay axle equipment in DT- tively. DT-based model’s integration with its platform that ensure
based model can respond to operational process in real-time and personnel in smart port’s transport chain is correctly informed. Infor­
realize the restoration of real motion dynamics. DT-driven platform also mation sharing and interconnection affect a goal of making the man­
supports the input of predictive data to perform visual simulations of agement of port smarter. Port authority provides a great example of
operational process in container yard to evaluate the feasibility of efficient and sustainable implementation of new information sharing
Mawan port management plan. For example, the multi-view simulation solutions (Darko et al., 2020). It’s a solution center for various questions,
of the trailer operation in terminal serves as key link of boosting the water and coastal navigation, security, railway, port facilities, property
container’s flow efficiency. In trailer DT-based models (Szpytko and management and economic conditions in local area. In order to improve
Duarte, 2020), the planned instruction track and actual operation track the rapidity and effectiveness of the correlation between DT-based
that the trailer is executing can be displayed from different perspectives, models and real-time data, the Hamburg port authority has success­
so as to assist the management personnel to check whether there is a fully integrated 5th Generation Mobile Communication Technology
deviation and give timely warning to trailers running in violation of (5G) technology into their DT-based models. 5G-aided Mobile Network
related rules. In deduction and prediction of container depot plan, DT- Architecture (MoNArch) is tested in Hamburg port’s DT-based models,
driven platform reasonably divides the area according to current and which have deployed three parts to address different requirements
future containers information, thus percolates an appropriate location within port area, as shown in Fig. 5.
for each container to improve the utilization rate of depot. DT-based
models offer Mawan port an extremely high practical value for profes­ • PART I: An ultra-reliable low latency communication (Kapkaeva
sional managers and operators working. Mawan port dispatchers can et al., 2021) designed to connect a traffic light to the Port of Ham­
control its overall situation, operators can observe the execution result burg’s vessels traffic control center, in an attempt to improve the
of the equipment in real-time and give timely feedback, while container traffic flow within port area.
handlers can rapidly check the stock with one-click. Moreover, man­ • PART II: An enhanced mobile broadband (Gupta et al., 2021) serving
agement personnel monitor the abnormal state of the container yard in interactive AR for engineering teams to improve certain the Port of
real-time via DT-driven platform to ensure containers’ safe storage. In Hamburg’s PTOs (e.g. asset maintenance, management).
future, a large number of APIs interfaces will be opened for the conve­ • PART III: A massive machine type communication (Mancuso et al.,
nience of Mawan port managers to introduce more types of business data 2019) developed to connect environmental sensors installed on
and improve scalability of DT-based models and platform through big mobile barges to the application cloud of Hamburg port authority for
data and AI. measuring environmental quality in port area.

The Port of Hamburg integrates several independent DT-based

1304
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Fig. 5. Testbed schematic setup and use cases based on DT-driven system for Port of Hamburg.

models with 5G MoNArch system to form an interconnected platform concerning the Barcelona port and its DT-based models gathered, port
available across whole area. DT-driven platform brings together data managers can reproduce the status of the PTOs at any given moment on
from different sources, including sensors, mobile devices, and databases a two dimension (2D) map. Although the data and digital models have
of various partners. Such digital platform helps Hamburg port authority already ripe for simple analysis, the prediction function of DT that fails
and its partners improve the internal PTOs, meanwhile it collects real- to collect real-time data has not been achieved yet. To promote the
time traffic information to minimize traffic bottlenecks at the depot management and decision-making in Barcelona port, the authority
and terminal gates. Thanks to 5G MoNArch, DT-driven platform has further optimizes its numerical models based on DT technology, with a
geolocation capabilities to pinpoint incoming trucks, facilitating the major challenge and purpose of improving its “port activities” repre­
port to plan container truck traffic more efficiently. With 5G MoNArch sentation, obtaining a real-time data for models and managing Barce­
integrated into DT-based models, the Port of Hamburg is maintaining lona port more smartly. As shown in Fig. 6, the Port of Barcelona
smarter control over each moving part. It means upgrading the services continues to take advantage of DT’s rapid development to complete the
in a number of ways, including improving traffic light control for better optimization process of smart port area. In the initial phase, Barcelona
traffic flow, and integrating mobile sensors on barges that are connected port authority uses geographic graphics and high-definition video
through 5G MoNArch (Onel et al., 2019). With bigger bandwidth, AR acquisition technology (Van der Berg et al., 2012) to present a 3D rep­
and virtual reality (VR) 360 pictures of Port of Hamburg are able to be resentation of Barcelona port area in digital form. It adds sensors to the
implemented, which all can be monitored through their DT system. port’s operating area and connects all of its sensors with advanced 5G
As Spain’s largest smart port, the Port of Barcelona plays an impor­ communications and Wi-Fi. In this way, communications in real-time
tant role in worldwide cargo shipping and logistics, managing 63 million can be greatly improved. The additional sensors mean more data on
tons of cargo each year. Thanks to collected information data (Nokia PTOs, yet too many data limit port authority’s ability to have a timely
drives key European, n.d.), the construction of DT-based models for port translation of this information into actionable intelligence. Most data-
helps a significant effect. In the first stage, with all information driven AI are implemented through innovations in sensor applications

Fig. 6. DT-driven data communication and sharing for Barcelona port’s management in city zone.

1305
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

that can generate large volumes of business data by going beyond simple machinery and more. This leads to a lot of inefficiencies and evaluation
PTOs detection and binary function-like capabilities. For this reason, risks, when handling freights on dock and in loading/unloading phases
Barcelona port authority adopts Tsenorflow AI system, which enables on vessels. DT-driven engineers in automation and terminal operations
the sensors to make interpretation with valid data and make decisions focus on three parts of smart port’s DT-based models, including the first
automatically (Goel et al., 2020). Next, based on video image analysis, it part is forecasting with a static model based on existing data, the second
is like a powerful detector that can realize count, detect changes, send part is creating a dynamic model using DT-based program to manually
and notice, measure available space, etc. Finally, through APIs program support PTOs, and the last part is bringing about automated algorithms
developed to facilitate the interconnection between DT-based models on DT-based models. By providing wireless control over existing tech­
and data, the port operators are able to improve dataflow in real-time. In nologies, communication latency is significantly increased compared to
fact, the primary goal of making technical innovation from Barcelona standard Ethernet (Shirowzhan et al., 2020). The Port of Livorno serves
port’s DT-based model is to represent timely data on within framework. as an experimental area covered by a live 5G network (shown in Fig. 7).
The transparency of PTOs is improved for end clients by providing in­ Through 5G connectivity, Livorno port authority enables the capture of
formation relevant to business in real-time, thus implementing the full- mass data from smart sensors, 3D light detection and ranging (LiDAR)
process prediction of vessels loading and unloading, as well as the and wide dynamic range (WDR) cameras (Jazdi et al., 2021) in opera­
container delivery and pick-up conducted in Barcelona port. DT-based tional area. According to measurements, with an ideal Wi-Fi, there is a
models and visual interface provide detailed data on all incoming and decrease of 0.05% observed with respect to the reference Ethernet case
outgoing traffic. Combined with data from customs, they allow for ac­ (Martino et al., 2020). This difference is higher for non-optimized Wi-Fi
curate 24 to 48 h forecast on Barcelona port’s traffic, enhancing and private 4G long term evolution (LTE) cases (0.31% and 0.38%,
container terminals management for its volume increase. It helps Bar­ respectively). In 5G case, the overall data appear to be comparable to
celona port’s managers to make decisions in real-time and guarantee Ethernet reference case (0.01%), which are much better than one pre­
safety of businesses and workers. Meanwhile, the optimized DT-driven dicted for other wireless technologies. Those data captured through 5G
management is conducive to improving the productivity and efficiency feeds a DT-driven engine, which elaborates a virtual replica of Livorno
of smart PTOs. port area in real-time.
DT-based modeling is conducted for Livorno port’s outdoor area
where the general cargo freights are stored. As shown in Fig. 7, there is a
3.4. DT-driven visualization and risk prediction satellite view of Livorno port’s real area inside terminal, which is built
by DT-based models. Meanwhile, another basic element for DT-based
The predictive optimization is one of the characteristic applications models is modeling of freight such as wooden boxes. It can be seen
of DT-driven management in smart port. Using a DT-based model, it that Livorno port where the stocked goods are monitored by a set of
predicts failures in advance and reduce the downtime of PTOs proced­ WDR cameras. These devices handle bright and dark conditions at the
ure. It is helpful for managing area equipment, monitoring forklift same time, resulting in normal running status in Livorno port’s outdoor
trucks, doing predictive simulations and further optimizing manage­ area where the light intensity and direction change continuously due to
ment in ports. DT-based models are used to determine equipment’s the weather conditions and time of a day. WDR cameras play a signifi­
remaining useful life and most opportune time to service or replace it. cant role in positioning and tracking tasks, which are modeled in DT
The variables such as weather conditions, vessels state, containers size, virtual scenario as well. The accurate goods localization is fundamental
or performance factors are employed to trigger hundreds of predictive to optimize the storage space and allow for a fast retrieval of proper
simulations, in order to evaluate readiness and necessary adjustments to objects at the right time for an efficient loading of vessels. WDR cameras
current port management set points. This application leads port staff to stream videos are enabled through 5G connectivity to have a software
optimize their operations, which aims to mitigate risk, gain system ef­ application in charge of identifying objects and positions. The available
ficiencies, reduce cost and ensure safety. It envisions that DT allows port data in real-time are captured by Livorno port’s WDR cameras and
workers to operate at the intersection of cyber and physical world sensors from vehicles operating in operational area. The management
(Telukdarie et al., 2018). Unlike the cargoes transported in standard system based on AI and deployed by Ericsson for DT project determines
containers, the cargo type in smart port is characterized by irregular and the sequence of logistics tasks and activities correlated with a real-time
non-standard geometries like pipes, cranes, components for industrial

Fig. 7. Visualization and its main components of DT-driven management in Port of Livorno.

1306
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

information flow. The tasks consist of movement, tracking and posi­ IoTs equipment and sensors, these multi-point source data can be rep­
tioning of forklifts and freights, and inventories of goods, orders of resented on DT-based models, thus improving efficiency of port envi­
loading and unloading of vessels. The tasks also mean identifying spe­ ronmental protection. The advanced DT-based models and predictions
cific activities, such as using cranes and moving larger loads in Livorno of tool-modules assist Oulu port authority to formulate better environ­
port, as with a “high risk” flagging. The detailed DT-driven view of mental plans and sustainable measurements for PTOs, and optimize the
everything happening allows AI engine to provide feedback on processes use of Oulu port’s infrastructure and assets. Aided by Nordic visual
and share live updates with supervisors. Livorno port authority develops design firm Sitowise (Alzamora et al., 2019), Oulu port has implemented
a management platform, which provides on-demand identification of the DT-based models of its own. The models covers all basic elements
any vehicles in operational area to accomplish a specific task. Based on a including the objects structure and environment of port area, though
multi-stereo vision technology, a positioning system is developed to which the internal relations and laws in port process of environmental
guarantee a positioning estimation with high precision when freights are pollution can be clearly described. Meanwhile, DT-based models can be
far and angled with WDR cameras. Moreover, VR application is used by mobilized to simulate the regional, river basin and spatial environment
operators to virtually navigate inside DT-based models (Ashraf et al., within the scope of Oulu port, before predicting the transmission as well
2016). All the information related to digital objects are identified in as dilution and diffusion characteristics of pollutants in environment
database, e.g. identity document (ID), content, dimensions, weight, etc. media. During this process, environmental variables, disturbance and
These data appear in a virtual clipboard that can be manipulated in VR decision variables are accurately collected. The starting point is to
environment. It also deploys the specific algorithm which makes it collect spatial data from Oulu port’s distributed sources, and implement
possible to achieve an automated and optimal sorting of freights and data in a real-time DT-driven platform called “AURA” (Audouin and
enables a quality check by inspecting terminals virtually from a remote Finger, 2018). Oulu port’s extensive infrastructure contain industrial
site. The algorithm is fully integrated with VR applications so that the lights, monitors and systems, all provide data support for DT-driven
result of optimal sorting is visible in DT-driven platform before being platform. Fig. 8 presents the monitoring processes of Oulu port imple­
executed in field. It is greatly realistic way of visiting the operational mented by AURA. It enables data collection, analysis and simulation in
area while remaining in back-office with a significant increase of man­ order to fully use the potential of resources and make more scientific
agement for operators. In addition, Livorno port pilot contains specific management in Oulu port. There are totally four types of data from
AR applications for forklift drivers (Wang et al., 2019; Mogensen et al., distributed sources for DT-based models simulation input data. The 1st
2021). When driving these vehicles, they are equipped with a rugged type is existing appointment system, which contains information about
smart headset built for protective needs, through which they can see registered vessels for current day in Oulu port. 2nd type is historical
graphical signs and instructions of freight they should move next. These data, which consists of processing times in Oulu port for different vessels
directions appear laid on top of operator’s vision of real port area. DT- and traffic information for the route from inbound lane to berth area.
driven visualization are revolutionizing way to Livorno ports, for 3rd type of data has a potential to use external sources while the 4th is
boosting operational efficiency, managing high massive data, avoiding responsible for supplying the simulation with real-time data via sensors
accidental events, loss prevention and risk prediction. (e.g. from cameras to proximity sensors, registering the arrival vessels at
different process stages) (Hasan et al., 2018). Above-mentioned data are
3.5. DT-driven management for environment protection and sustainability communicated to Oulu port’s AURA IoTs-platform. AURA platform
provides DT-driven simulations, which uses work in progress (WIP) in­
Besides the operational efficiency and safety, the core weights to formation (Hofmann and Branding, 2019) from database to initialize the
measure development quality of a world-class smart port include its current system status. Subsequently, different dispatching algorithms
environmental sustainability status. It is a critical factor to determine are evaluated and the results are displayed in a Web-based application.
whether smart port achieves a good balance between economic interests Thanks to DT technology and IoTs-platform, the personnel responsible
and sustainable development in future. The key of DT-driven manage­ for environmental supervision, management and emergence planning of
ment for port environmental protection are expected to involve main Oulu port are able to efficiently and intuitively analyze performance,
sources (e.g. dangerous chemical containers), environmental quality, test various planning scenarios, assess the consequences, and track
pollutant discharge management and environmental emergency distributed data in an integrated, visual 3D environment. As a result,
response, etc. DT-based models’ sharing capability determine that data- Oulu port authority is able to execute better environmental management
driven and physical port resource optimization can be realized by data and make more scientific decisions regarding further sustainable
flow. Therefore, various environmental data of port, including pollution development. Moreover, Oulu port’s DT-driven management allows all
source data, meteorological data, environmental data and others, are participants of PTOs a simultaneous, direct access to an up-to-date,
collected by sensing equipment. Then, DT-driven system is used to carry comprehensive information on these presented in an intuitive, spatial
out real-time monitoring, synchronous description and dynamic pre­ visual context. These participants contain vessels, trains and trucks,
diction of the entity elements of port environmental management. construction projects, area maintenance and waste management, cus­
Meanwhile, DT-driven management loaded with network technology toms and seamen’s mission, emergency and environmental protections.
helps smart port reduce energy consumption and waste. In Port of Above-mentioned analyzed applications of DT-driven management
Valencia, DT-driven management dynamically tracks the port lighting in global smart ports can be summarized as following Table 1.
system, providing a lighting-up service only when vehicles approach
docks. Valencia port authority saves nearly 80% of energy consumption 4. Current challenge and future work
after adopting the dynamic sensitive lighting system (Qiu et al., 2019).
Using DT-driven management to carry the drones, monitors at the port 4.1. Challenges to DT application in smart port
are able to check for fuel leakage, monitor oil spills on sea surface and
supervise cleanup, effectively reducing the cost of environmental Challenge to create a mirrored DT model. Although many benefits
management. are associated with implementation of DT-based model, there are still
In Finland, Oulu port authority hopes to enable comprehensive limitations in its creation. Construction of DT-based model for port
monitoring of environmental protection and sustainability, with a view process should include geometric information and process information
to improving management in PTOs processes. Oulu port builds DT-based models. Geometric information model covers port environment, build­
models that enable real-time sharing of dynamic environmental moni­ ing, goods, equipment and auxiliary systems. The appearances, sizes and
toring data through interactive DT-driven platform. By integrating dig­ positions of its modeling objects should be exactly the same as those in
ital representation, environmental data and real-time data monitored by actual port scene. At present, there are many DT geometric modeling

1307
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Fig. 8. DT-driven platform “AURA” for environmental monitoring in Port of Oulu, Finland.

Table 1
Summary of effects based on DT-driven applications in the global smart ports.
Port name Location Software Model type Data acquisition Platform DT service

Valencia port Spain, Europe BIM Hydraulic Model IoTs GoAigua Improved efficiency of cargo
handling
Singapore Singapore, Asia Bentley 3D Map + Model GPS Lands Orbit 3DM Content Manager Transport traffic monitoring system
port
Rotterdam the Netherlands, BIM 3D Model IoTs Cloud Server 1st version DT application Cargo handling and transportation
port Europe program
Mawan port China, Asia 51 World 3D Model IoTs AES scene Operation and storage in containers
terminal
Hamburg port Germany, Europe BIM 3D Model 5G Technology 5G MoNArch Data communication & sharing
Barcelona Spain, Europe BIM 3D Model 5G Communications and Tsenorflow AI system Data communication & sharing
port Wi-Fi
Livorno port Italy, Europe Corealis Geometrical LiDAR and WDR Cameras Ericsson’s AI AR/VR visualization & risk prediction
Model
Oulu port Finland, Europe Sitowise 3D Model IoTs AURA Environment protection &
sustainability

tools (Rudin et al., 2020; Qi et al., 2021; Tran et al., 2021), however the problem is DT data transmission in large smart ports is not stable
original intentions of these are not to serve DT technology. So in enough. In data collection or instruction processes, DT data can be lost
modeling process, it is necessary to consider 3D modeling and para­ when transmitted in real-time. While a traditional port is transforming
metric modeling, meanwhile to ensure that each modeling object should itself into a smart port, it can no longer merely depend on the stability
be assigned a unique code. For the core equipment required by PTOs and reliability of wireless communication network for data trans­
process, its behavior needs to be defined and coded (Grégorio et al., mission, which is not enough to meet higher construction requirements
2021; Bellalouna, 2021; López-Estrada et al., 2019). The process infor­ of real-time DT-based model interaction (Koulamas and Kalogeras,
mation model belongs to pure data model. This kind of model needs to 2018). Third limitation is once DT-based models generate a huge
take the geometric information model as its data carrier, developing amount of data, its storage and processing are actually presenting a great
plug-in on modeling software. Currently, the design of basic process challenge. It is especially true after the introduction of some complex
information model is still not advanced in DT-based modeling. There is equipment in smart ports, whose fault diagnosis and preventive main­
still a lack of a proprietary paradigm that can integrate the engineering tenance pose difficulties of storing and analyzing massive data coming
data of physical system and basic models needed to support the con­ from different data sources, as well as improving data storage and
struction of manual system. computing capacity (Tao et al., 2019). Fourth challenge is from different
Challenge to connectivity of DT data. The first limitation is to types of data transmission and interactions between different systems
ensure the consistency of multi-dimensional and multi-scale data and devices are required during the integration process of DT-based
collection in PTOs (Stark et al., 2019). The consistency of scale or unit of models derived from different dimensions. However, the existing
measurement of data collection involved in port process includes communication interface protocols and data standards of different sys­
physical data, geometric data, time data, etc. If the coordinates of tems and devices are not unified, which also leads to an obstacle in DT-
models are not consistent with measurement units, it leads to a fusion based model construction. Fifthly, data sharing and opening are a lim­
failure between different DT-based models (Huang et al., 2021). Second itation when a DT-based model is applied to IoTs platform or cloud

1308
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

server. DT data sharing mechanism and service system have not yet been emulators of physical objects for a higher precision to strengthen
perfected, with their sharing among different subjects bringing great future DT-driven implements for smart port.
security risks and interest conflicts (Caputo et al., 2019). Finally, multi- • Great importance should be attached to autonomous renewal capa­
source heterogeneous data involve port’s spatial model, internet infor­ bility of DT-driven management in future technological develop­
mation, real-time perception of IoTs, professional knowledge, etc., and ment. As DT-driven autonomous updating is characterized by
even a large volume of texts, audios and videos. How to integrate, fuse dynamic changes during PTOs, a key to facilitating DT modeling is to
and uniformly manage these multi-source, heterogeneous and multi- ensure virtual-real consistency. The parameter adjustment and
module data (Cattaneo and Macchi, 2019) is one of critical issues. replacement of traditional models depend more on manual inter­
Challenge to data security of DT system. After DT-based models vention, resulting in a deviation triggered by time delay. In future,
are adopted, previously closed system of transport, warehousing and the virtual DT-based models are expected to have capacity to sense
logistics in enterprises is gradually opening. As a result, it is reflected in and analyze real-time data, self-adjust and update, and realize
problems including data loss and network attacks arising from transport autonomous evolution in continuous data supply. When should
data transmission. In fact, the security risks occur in virtual and real different sub-models be updated, how to update (in structure or
interaction process of DT-based models. During data storage, a large parameter), whether updated models have an impact on PTOs are the
amount of production operational data, transport data and factory key works to be considered (Breda and Kiss, 2020).
external data are generated and stored. Once one of these storage forms • DT-driven management is a complex interoperating system based on
has security problems, it may potentially cause data leakage in smart interaction and understanding of virtual and real space, which is
ports (Feldt et al., 2020). Self-organization of resources and self-decision expected to be used to achieve high precision reconstruction target
of process are largely required in managing PTOs with DT-based models. before increasing the efficiency of port management. The features of
However, the virtual system may have a variety of unknown safety high-precision PTOs [95] raise high requirements for data accuracy
loopholes, making it vulnerable to external attacks. It leads to a disorder in DT-based models. DT-based models are largely relied on big data
of DT-driven management system, resulting in wrong instructions to acquisition in realization process, the future implement focus needs
port operating area while triggering severe security events (Rebmann to be placed on historical data selection, mass volume processing and
et al., 2020; Akhlaghi et al., 2020). big data timeliness. Biggest advantage of DT-driven management
Challenge of promoting ongoing DT application. DT has already applied in PTOs is intelligent fusion of multi-source data, and it is
been able to be applied in industry scenarios, but potential application enabled to merge location information from transport ships, weather
modes formed by reconstruction and integration of port industry ele­ conditions, data inside containers, operating equipment, and logis­
ments are full of challenges. The accuracy of a DT-driven design needs to tics data. Obviously, there is a long way to go for text mining of single
be further improved, while its performance in real transport environ­ data set or multi-source data in future implements.
ment needs to be verified. DT can be used to speed up logistics intro­ • The core element of DT-driven management with its integration
duction time, improve transport efficiency, avoid operational risks, and platform is basic software used to perform twinning during PTOs
reduce management cost. It is combined with a large number of sensors before correlating it with real-time data. DT-driven platform is sup­
to collect data concerning the environment and state of PTOs, thus posed to be a vendor agnostic integration and management which
improving remote monitoring and predictive maintenance (Shvedenko connects all existing and future systems, subsystems, sensors and
and Volkov, 2019). The other limitation comes from large-scale forma­ data bases in one common and easy-to-use user interface [96]. It
tion of DT-driven industry chain. DT is a comprehensive integrated allows management, coordination, correlation, escalation and
technology for multi-intelligent transport and multi-deployment envi­ sharing of all data through a single platform. The command and
ronment, which can achieve the integration of “man-machine-material- control centers, field staff and connected external resources benefit
law-circle”, as well as efficient collaboration and optimal resource from the centralized platform as re-occurring tasks, responses on
allocation across regions, industries, enterprises and departments (Liu alarms and incidents are handled more efficiently and without
et al., 2021). The current parties of DT-driven industrial chain are still in operator randomness. The integration platform provides a develop­
fragmental stage, and its formation and development process based on ment foundation for DT developers, enabling start up quickly and
port industry will inevitably encounter more challenges in the future. scale up efficiently.
• Standardizations of DT-based models and platform should be
4.2. Future works for DT-driven management in smart port constantly given close attention. The International Organization for
Standardization (ISO) currently promotes ISO/TC 184/SC 4, it al­
There are a lot of potential applications for DT-driven management lows a high-level framework for a digital representation to ensure the
including monitoring, simulation, and remote control of smart ports’ consistency and interoperability of standard. IPC-2551 is developed
assets with virtual objects (Saddik, 2018). Challenges focus on the by generic requirements for digital twin task group, by using this
model, data, process, and system twinning. The implementation of DT- standard, port can initiate application interoperability to assess their
driven management in global ports will enable completed synchroni­ current DT readiness level. A number of international standards
zation information of PTOs, which will extend a potential application to related to DT-driven management have been proposed by Interna­
the following aspects. tional Electro Technical Commission (IEC), Information and
Communication Technology (ICT), Institute of Electrical and Elec­
• DT-based models can be used to build a “good enough” digital for­ tronics Engineers (IEEE), National Intelligent Manufacturing Stan­
mation by constantly iterating with reality of physical world through dardization General Group and other standardization organizations.
its construction, thereby providing assistance for lifecycle informa­ Such as Concepts and Terms of Digital Twin (ISO/IEC JTC1
tion flow in PTOs. Boundary of existing implements and methods AWI5618), Application Cases of Digital Twin (ISO/IEC JTC1 AWI
needs to be further extended. The abstract methods employed during 5719), Reference Framework of Intelligent Manufacturing Virtual
DT construction process actually show that DT-based model should Factory (20182046-T-339) [97]. It should be encouraged that the
be gradually optimized in continuous iteration process, to enable it port authorities to support these formal activities and to develop
to map the multi-physical field, multi-scale features and lifecycle inputs for standardization of DT technologies. Standards can define
information of physical entities (Wang et al., 2021). Calibration of a APIs for DT-driven data access to securely and reliably store, manage
DT-based model can be conducted through analyzing the behaviors and retrieve records. DT implementations may restrict updates and
of port process. The variations are analyzed and recognized by deletes to support data consistency goals.
monitoring the real PTOs scenario. These changes will be updated in

1309
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

5. Conclusions Science (20692193100) and National Natural Science Foundation of

China (52001196). Kan WANG acknowledges State Key Laboratory of
DT technology is a burgeoning evolution of digital models from Explosion Science and Technology in BIT, China Institute of Commu­
multiple tools including 3D simulation, real-time data acquisition, AI, nications Education (JTYB20-60) for providing supporting. More
cloud computing and pre-execution. It goes beyond simulations to help importantly, the authors would like to gratefully acknowledge the edi­
port authorities understand what is happening in real-world. As DT- tors and reviewers for the valuable suggestions and comments.
driven management, the smart ports can be expected to develop
further optimizations and more efficient port’s processes. With in-depth References
analysis, conclusions are further summarized as follows:
Al Dakheel, J., Del Pero, C., Aste, N., Leonforte, F., 2020. Smart buildings features and
key performance indicators: A review. Sustainable cities and society 61, 102328.
• This study brings contributions to research on exploring the core Nacera, B., Bouabdellah, K., Hafid, H., 2016. Cotransportation and operation between
framework of DT-driven management for transport and operation in Intelligent Autonomous Vehicles to enhance container terminal transportation and
smart port. It found the key driver of DT-driven management in operations. Journal of Innovation in Digital Ecosystems 3 (1), 22–29.
Tao, F., Anwer, N., Liu, A., Wang, L., Nee, A.Y.C., Li, L., Zhang, M., 2021. Digital twin
lifecycle of smart port including physical entities, virtual models, towards smart manufacturing and industry 4.0. Journal of Manufacturing Systems
data, integration methods, analysis methods and services. DT-based 58, 1–2.
system for port management can be categorized into physical layer, Cheng, J., Zhang, H., Tao, F., Juang, C., 2020. DT-II: Digital twin enhanced Industrial
Internet reference framework towards smart manufacturing. Robotics and
data layer, model layer, service layer and application layer. Computer-Integrated Manufacturing 62, 101881.
• This study adopts multi-aspect applications of DT-driven manage­ Tsega, Y.M., Valentina, D.P., Stefano, R., 2020. Digital twin models in industrial
ment to realize visibility and traceability of on-site situations in real transportation and operations: a systematic literature review. Procedia
Manufacturing 42, 267–272.
smart port. Relying on DT-based models and visualization platform,
Ritto, T.G., Rochinha, F.A., 2021. Digital twin, physics-based model, and machine
DT-driven management for smart ports are realized in the aspect of learning applied to damage detection in structures. Mechanical Systems and Signal
cargo transportation, containers storage, real-time data sharing, Processing 155, 107614.
predictive optimization and environmental sustainability. DT-driven Shao, G., Helu, M., 2020. Framework for a digital twin in manufacturing: Scope and
requirements. Manufacturing Letters 24, 105–107.
management gives opportunity to create 5th generation smart ports, Wang, P., Luo, M., 2021. A digital twin-based big data virtual and real fusion learning
while simultaneously presenting the advantages of intelligent, vir­ reference framework supported by industrial internet towards smart manufacturing.
tual and green. Journal of Manufacturing Systems 58, 16–32.
ISO/FDIS 23247-1. Automation systems and integration – Digital twin framework for
• The creating and maintaining processes of DT-based models are still manufacturing – Part 1: Overview and general principles.
difficult and risky, this study proposes the concerned challenges of IPC-2551. International Standard for Digital Twins.
DT integrated with digital models in PTOs. The main limitation is to ISO/TC 184/SC 4. Industrial data.
Harper, K.E., Malakuti, S., Ganz, C., 2019. Digital Twin Architecture and Standards. IIC
build a consistent model for the holistic use of DT in entire port Journal of Innovation 11, 1–12.
process, which covers the topics of common understanding interfaces Falah. M. F., Sukaridhoto. S., Al. Rasyid., Wicaksono. H., 2020. Design of virtual
and standardizations, as well as efficient data flow. DT technology engineering and digital twin platform as implementation of cyber-physical systems.
Procedia Manufacturing, 52: 331-336.
extends well beyond the life spans of the formats for proprietary Zhang, H., Ma, L., Sun, J., Lin, H., Thürer, M., 2019. Digital twin in services and
design software that are likely used to create them and means of industrial product service systems: review and analysis. Procedia CIRP 83, 57–60.
collected data. DT-driven management creates in proprietary design Jones, D., Snider, C., Nassehi, A., Yon, J., Hicks, B., 2020. Characterizing the Digital
Twin: A systematic literature review. Journal of Manufacturing Science and
software formats have a connectivity and security problem of being
Technology 29, 36–52.
unreadable at some point in the service practices of smart port. Jiang, Z., Guo, Y., Wang, Z., 2021. Digital twin to improve the virtual-real integration of
• DT-driven management can be extended in several development industrial IoT. Journal of Industrial Information Integration 22, 100196.
directions of port industry. The boundary of existing methods needs Mathupriya. S., Banu. S. S., Sridhar. S., Arthi. B., 2020. Digital twin technology on IoT,
industries & other smart environments: A survey. Materials Today: Proceedings, in
to be renewal so that the better DT-based models can be established. press.
DT-based models are expected to have the capacity to sense and Liu, Z., Bai, W., Du, X., Zhang, A., Xing, Z., Jiang, A., 2020. Digital Twin-based Safety
analyze real-time data, self-adjust and update, and realize autono­ Evaluation of Restressed Steel Structure. Advances in Civil Engineering 7, 1–10.
Inkinen, T., Helminen, R., Saarikoski, J., 2021. Technological trajectories and scenarios
mous evolution in continuous data supply of smart ports. DT-driven in port digitalization. Research in Transportation Business & Management 1,
management can form a centralized platform with multi- 100633.
technologies and AI fusion of multi-source data. With the contin­ Lin, T., Shi, G., Yang, C., Zhang, Y., Wang, J., Jia, Z., Guo, L., Xiao, Y., Wei, Z., Lan, S.,
2021. Efficient container virtualization-based digital twin simulation of smart
uous standardization constructions of DT technology, it is expected industrial systems. Journal of Cleaner Production 281, 124443.
to promote an integrated DT-driven supply chain in management of Rüßmann. M., Lorenz. M., Gerbert. P., Waldner. M., Justus. J., Engel. P., Harnisch. M.,
global port industry. 2015. Industry 4.0: The future of productivity and growth in manufacturing
industries, Consulting Group. Technical Report – Boston.
He, B., Cao, X., Hua, Y., 2021. Data fusion-based sustainable digital twin system of
CRediT authorship contribution statement intelligent detection robotics. Journal of Cleaner Production 280, 124181.
Coelho, F., Relvas, S., Barbosa-Póvoa, A.P., 2021. Simulation-based decision support tool
for in-house logistics: the basis for a digital twin. Computers & Industrial
Kan Wang: Conceptualization, Methodology, Writing - original
Engineering 153, 107094.
draft, Visualization. Qianqian Hu: Data curation, Software, Formal Zhuang, C., Gong, J., Liu, J., 2021. Digital twin-based assembly data management and
analysis. Mingjun Zhou: Writing - review & editing. Zhou Zun: Vali­ process traceability for complex products. Journal of Manufacturing Systems 58,
dation, Writing - review & editing. Xinming Qian: Supervision, Inves­ 118–131.
Chakraborti, A., Heininen, A., Koskinen, K.T., Lämsä, V., 2020. Digital twin: multi-
tigation, Resources. dimensional model reduction method for performance optimization of the virtual
entity. Procedia CIRP 93, 240–245.
Declaration of Competing Interest Eric, V.D.H., Mahadevan, S., 2021. Digital Twin: Generalization, characterization and
implementation. Decision Support Systems 1, 113524.
Kong, T., Hu, T., Zhou, T., Ye, Y., 2021. Data construction method for the applications of
The authors declare that they have no known competing financial workshop digital twin system. Journal of Manufacturing Systems 58, 323–328.
interests or personal relationships that could have appeared to influence White, G., Zink, A., Codecá, L., Clarke, S., 2021. A digital twin smart city for citizen
feedback. Cities. 110, 103064.
the work reported in this paper. Akbarieh, A., Jayasinghe, L.B., Waldmann, D., Teferle, F.N., 2020. BIM-Based End-of-
Lifecycle Decision Making and Digital Deconstruction_ Literature Review.
Acknowledgements Sustainability 12, 2670.

This work is financially sponsored by Shanghai Key Projects of Soft

1310
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Heaton, J., Parlikad, A.K., Schooling, J., 2019. Design and development of BIM models to Van der Berg, R., Langen, P.W.D., Costa, C.R., 2012. The role of port authorities in new
support transportation and operations and maintenance. Computers in Industry 111, intermodal service development; the case of Barcelona Port Authority. Research in
172–186. Transportation Business & Management 5, 78–84.
Kuitunen, V., 2019. Added value of Gisgro information system for the Port of Goel, P., Jain, P., Pasman, H.J., Pistikopoulos, E.N., Datta, A., 2020. Integration of data
HaminaKotka[D]. South-Eastern Finland University of Applied Sciences. analytics with cloud services for safer process systems, application examples and
Hu, L., Nguyen, N., Tao, W., Leu, M.C., Liu, X.F., Shahriar, M.R., Nahian, S.M., 2018. implementation challenges. Journal of Loss Prevention in the Process Industries 68,
Modeling of cloud-based digital twins for smart manufacturing with MT connect. 104316.
Procedia Manufacturing 26, 1193–1203. Telukdarie, A., Buhulaiga, E., Bag, S., Gupta, S., Luo, Z., 2018. Industry 4.0
Glatt, M., Sinnwell, C., Yi, C., Donohoe, S., Ravani, B., Aurich, J.C., 2021. Modeling and implementation for multinationals. Process Safety and Environmental Protection
implementation of a digital twin of material flows based on physics simulation. 118, 316–329.
Journal of Manufacturing Systems 58, 231–245. Shirowzhan, S., Tan, W., Sepasgozar, S., 2020. Digital twin and CyberGIS for improving
Arrichiello, V., Gualeni, P., 2019. Systems engineering and digital twin-a vision for the connectivity and measuring the impact of infrastructure construction planning in
future of cruise ships design, production and transportation and operations. smart cities. International Journal of Geo-Information 9 (4), 240.
International Journal on Interactive Design and Manufacturing 14, 115–122. Jazdi, N., Talkhestani, B.A., Maschler, B., Weyrich, M., 2021. Realization of AI-enhanced
He, B., Bai, K.J., 2021. Digital twin-based sustainable intelligent manufacturing-a review. industrial automation systems using intelligent digital twins. Procedia CIRP 97,
Advances in Manufacturing 9, 1–21. 396–400.
Mailly, P., Haber, S.N., Groenewegen, H.J., Deniau, J., 2010. A 3D multi-modal and Martino, M.D., Magnotti, F., Morvillo, A., 2020. Port governance and value creation in
multi-dimensional digital brain model as a framework for data sharing. Journal of the supply chain: The case of Italian ports. Case Studies on Transport Policy 8 (2),
Neurosci Methods 194 (1), 56–63. 373–382.
Lind, M., Bjørn-Andersen, N., Ward, R., Watson, R.T., Bergmann, M., Andersen, T., 2018. Ashraf, S.A., Holfeld, B., Wieruch, D., Wirth, T., Thiele, L., 2016. Wireless
Synchronization for Port Effectiveness. Port Technology Journal 79, 82–84. Communication for Factory Automation: an opportunity for LTE and 5G systems.
Barbieri, G., Gutierrez, D.A., 2021. A Gemma-Grafcet methodology to enable digital twin IEEE Communications Magazine 54 (6), 36–43.
based on real-time coupling. Procedia Computer Science 180, 13–23. Wang, H., Hu, J., Wan, W., Gui, H., Qin, F., Yu, F., Liu, J., Lü, L., 2019. A wide dynamic
Ivars-Baidal, J.A., Celdrán-Bernabeu, M.A., Femenia-Serra, F., Perles-Ribes, J.F., Giner- range and high resolution all-fiber-optic turbidity measurement system based on
Sánchez, D., 2021. Measuring the progress of smart destinations: The use of single photon detection technique. Measurement 134, 820–824.
indicators as a management tool. Journal of Destination Marketing & Management Mogensen, R.S., Rodriguez, I., Schou, C., Mortensen, S., Sørensen, M.S., 2021. Evaluation
19, 100531. of the impact of wireless communication in production via factory digital twins.
D’Amico, G., Szopik-Depczyńska, K., Dembińska, I., Ioppolo, G., 2021. Smart and Manufacturing Letters 28, 1–5.
sustainable logistics of port cities: a framework for comprehending enabling factors, Qiu, C., Zhou, S., Liu, Z., Gao, Q., Tan, J., 2019. Digital assembly technology based on
domains and goals. Sustainable Cities and Society 69, 102801. augmented reality and digital twins: a review. Virtual Reality & Intelligent Hardware
Alamoush, A.S., Ballini, F., Ölçer, A.I., 2020. Ports’ technical and transportation and 1 (6), 597–610.
operational measures to reduce greenhouse gas emission and improve energy Alzamora. F. M., Carot. M. H., Campos. J. C., Conejos. P., 2019. Development and Use of
efficiency: A review. Marine Pollution Bulletin 160, 111508. a Digital Twin for the Water Supply and Distribution Network of Valencia (Spain).
Peng, P., Poon, J.P.H., Yang, Y., Lu, F., Cheng, S., 2019. Global oil traffic network and 17th International Computing & Control for the Water Industry Conference, CCWI
diffusion of influence among ports using real time data. Energy 172, 333–342. 2019At: Exeter (UK).
Olba, X.B., Daamen, W., Vellinga, T., Hoogendoorn, S.P., 2019. Multi-criteria evaluation Audouin, M., Finger, M., 2018. The development of mobility-as-a-service in the Helsinki
of vessel traffic for port assessment: A case study of the Port of Rotterdam. Case metropolitan area: a multi-level governance analysis. Research in Transportation
Studies on Transport Policy 7 (4), 871–881. Business & Management 27, 24–35.
Witte, P., Slack, B., Keesman, M., Jugie, J.H., Wiegmans, B., 2018. Facilitating start-ups Hasan, R., Hossain, M., Khan, R., 2018. Aura: An incentive-driven ad-hoc IoT cloud
in port-city innovation ecosystems: A case study of Montreal and Rotterdam. Journal framework for proximal mobile computation offloading. Future Generation
of Transport Geography 71, 224–234. Computer Systems 86, 821–835.
Pauwels, P., Zhang, S., Lee, Y.C., 2017. Semantic web technologies in AEC industry: A Hofmann, W., Branding, F., 2019. Implementation of an IoT- and Cloud-based Digital
literature overview. Automation in Construction 73, 145–165. Twin for Real-Time Decision Support in Port Transportation and operations. IFAC-
Costin, A., Adibfar, A., Hu, H., Chen, S.S., 2018. Building Information Modeling (BIM) for Papers OnLine 52 (13), 2104–2109.
transportation infrastructure – Literature review, applications, challenges, and Rudin, R.S., Fischer, S.H., Damberg, C.L., Shi, Y., Shekelle, P.G., Xenakis, L.,
recommendations. Automation in Construction 94, 257–281. Khodyakov, D., Ridgely, M.S., 2020. Optimizing health IT to improve health system
Abbassi, R., Khan, F., Garaniya, V., Chai, S., Chin, C., Hossain, K.A., 2015. An integrated performance: A work in progress. Healthcare 8 (4), 100483.
method for human error probability assessment during the maintenance of offshore Qi, Q., Tao, F., Hu, T., Anwer, N., Liu, A., 2021. Enabling technologies and tools for
facilities. Process Safety and Environmental Protection 94, 172–179. digital twin. Journal of Manufacturing Systems 58, 3–21.
Akbarieh, A., Jayasinghe, L.B., Waldmann, D., Teferle, F.N., 2020. BIM-Based End-of- Tran, H., Nguyen, T.N., Christopher, P., Bui, D., Khoshelham, K., Ngo, T.D., 2021.
Lifecycle Decision Making and Digital Deconstruction-Literature Review. A digital twin approach for geometric quality assessment of as-built prefabricated
Sustainability 12 (7), 2670. façades. Journal of Building Engineering, in press 102377.
Abaglo, A.J., Bonalda, C., Pertusa, E., 2017. Environmental Digital Model: Integration of Grégorio, J., Lartigue, C., Thiébaut, F., Lebrun, R., 2021. A digital twin-based approach
BIM into environmental building simulations. Energy Procedia 122, 1063–1068. for the management of geometrical deviations during assembly processes. Journal of
Liu, L., Wang, K.Y., Yip, T.Y., 2013. Development of a container port system in Pearl Manufacturing Systems 58, 108–117.
River Delta: path to multi-gateway ports. Journal of Transport Geography 28, 30–38. Bellalouna, F., 2021. Case Study for an Transportation and operation-based Topology
D’Addona, D.M., Conte, S., Lopes, W.N., De Aguiar, P.R., Bianchi, E.C., Teti, R., 2018. Optimization Using the Digital Twin Approach. Procedia CIRP 98, 342–347.
Tool condition monitoring of single-point dressing transportation and operation by López-Estrada, L., Fajardo-Pruna, M., Gualoto-Condor, S., Ríos, J., Vizán, A., 2019.
digital signal processing of AE and AI. Procedia CIRP 67, 307–312. Creation of a micro cutting machine tool digital-twin using a cloud-based model-
Kumar. K., Ramkumar. K. R., Kaur. A., 2020. A lightweight AES algorithm based PLM Platform: first results. Procedia Manufacturing 41, 137–144.
implementation for encrypting voice messages using field programmable gate arrays. Stark, R., Fresemann, C., Lindow, K., 2019. Development and transportation and
Journal of King Saud University – Computer and Information Sciences. operation of Digital Twins for technical systems and services. CIRP Annals 68 (1),
Szpytko, J., Duarte, Y.S., 2019. Digital Twins Model for Cranes Operating in Container 129–132.
Terminal. IFAC-Papers OnLine 52 (10), 25–30. Huang, H., Yang, L., Wang, Y., Xu, X., Lu, Y., 2021. Digital Twin-driven online anomaly
Szpytko, J., Duarte, Y.S., 2020. A digital twins concept model for integrated detection for an automation system based on edge intelligence. Journal of
maintenance-a case study for crane transportation and operation. Journal of Manufacturing Systems 59, 138–150.
Intelligent Manufacturing 7, 927. Koulamas, C., Kalogeras, A.P., 2018. Cyber-Physical Systems and Digital Twins in the
Darko, A., Chan, A.P.C., Yang, Y., Tetteh, M.O., 2020. Building information modeling Industrial Internet of Things. Computer 51 (11), 95–98.
(BIM)-based modular integrated construction risk management – Critical survey and Tao. F., Zhang. M., Nee. A. Y. C., 2019. Chapter 9 – Digital twin and big data. Editor(s):
future needs. Computers in Industry 123, 103327. Digital Twin Driven Smart Manufacturing, Academic Press, 183-202.
Kapkaeva, N., Gurzhiy, A., Maydanova, S., Levina, A., 2021. Digital platform for Caputo, F., Greco, A., Fera, M., Macchiaroli, R., 2019. Digital twins to enhance the
maritime port ecosystem: Port of Hamburg case. Transportation Research Procedia integration of ergonomics in the workplace design. International Journal of
54, 909–917. Industrial Ergonomics 71, 20–31.
Gupta, R., Reebadiya, D., Tanwar, S., 2021. 6G-enabled Edge Intelligence for Ultra Cattaneo, L., Macchi, M., 2019. A digital twin proof of concept to support machine
-Reliable Low Latency Applications: Vision and Mission. Computer Standards & prognostics with low availability of run-to-failure data. IFAC-Papers OnLine 52 (10),
Interfaces 77, 103521. 37–42.
Mancuso, V., Quirós, M.P., Midoglu, C., Moulay, M., Comite, V., Lutu, A., Alay, O., Feldt, J., Kourouklis, T., Kontny, H., Wagenitz, A., 2020. Digital twin: revealing
Alfredsson, S., Rajiullah, M., Brunström, A., Mellia, M., Khatouni, A.S., Hirsch, T., potentials of real-time autonomous decisions at a manufacturing company. Procedia
2019. Results from running an experiment as a service platform for mobile CIRP 88, 185–190.
broadband networks in Europe. Computer Communications 133, 89–101. Rebmann, A., Knoch, S., Emrich, A., Fettke, P., Loos, P., 2020. A multi-sensor approach
Onel, L., López, A., Alves, H., Nardelli, P.H.J., Latva-aho, M., 2019. Hybrid resource for digital twins of manual assembly and commissioning. Procedia Manufacturing
scheduling for aggregation in massive machine-type communication networks. Ad 51, 549–556.
Hoc Networks 94, 101932. Akhlaghi, Y.G., Badiei, A., Zhao, X., Aslansefat, K., Xiao, X., Shittu, S., Ma, X., 2020.
Nokia drives key European research project 5G-MoNArch to bring 5G mobile network A constraint multi-objective evolutionary optimization of a state-of-the-art dew
architecture from concept to real world. https://www.nokia.com/. point cooler using digital twins. Energy Conversion and Management 211, 112772.

1311
K. Wang et al. Case Studies on Transport Policy 9 (2021) 1298–1312

Shvedenko, V.N., Volkov, A.A., 2019. A method for digital twin generation based on the Saddik, A.E., 2018. Digital twins: the convergence of multimedia technologies. IEEE
aggregation of information objects. Automatic Documentation and Mathematical Multi Media 25 (2), 87–92.
Linguistics 53 (3), 122–126. Wang, X., Wang, Y., Tao, F., Liu, A., 2021. New paradigm of data-driven smart
Liu, S., Lu, Y., Li, J., Song, D., Sun, X., Bao, J., 2021. Multi-scale evolution mechanism customization through digital twin. Journal of Manufacturing Systems 58, 270–280.
and knowledge construction of a digital twin mimic model. Robotics and Computer- Breda, G., Kiss, M., 2020. Overview of information security standards in the field of
Integrated Manufacturing 71, 102123. special protected Industry 4.0 areas & industrial security. Procedia Manufacturing
46, 580–590.

1312