Politecnico di Torino

Corso di Laurea Magistrale in Engineering and Management

A.a. 2022/2023
Sessione di Laurea Marzo/Aprile 2023

Applications of the Digital Twin in Industrial

Manufacturing

Relatore: Giulia Bruno Candidato: Iulian Burca

Summary

1. The Digital Twin Concept…………………………………………………….…...…1

1.1. Historical background………………………………………………………….....1
1.2. Concept definition and evolution…………………………………………….....1
1.3. Digital Twin basic structure……………………………………………………..3
1.4. Manufacturing application…………………………………………………….…4
2. Literature survey and data collection…………………………………………...….6
3. Cluster Analysis………………………………………………………………………....21
3.1. Production Line cluster………………………………………………….….21
3.1.1 Services………………………………………………………….....………….23
3.1.2 Communication Protocol and Architecture Network……………….……...25
3.1.3 Simulation Features…………………………………………………………..27
3.2. CNC Machine cluster…………………………………………………….....28
3.1.1 Services……………………………………………………………………….28
3.1.2 Communication Protocol and Architecture Network…………………...….31
3.1.3 Simulation Features……………………………………………………............32
3.3. Human-Robot Collaboration cluster…………………………………....33
3.1.1 Services………………………………………………………………………..33
3.1.2 Communication Protocol and Architecture Network…………………...….36
3.1.3 Simulation Features…………………………………………………………..36
3.4. Transportation System cluster………………………………………….....37
3.1.1 Services………………………………………………………………………..37
3.1.2 Communication Protocol and Architecture Network……………………...40
3.1.3 Simulation Features………………………………………………………......42

4. Overall findings and trends…………………………………………………….…....43

5. Summary………………………………………………………………………………….48
6. Acknowledgements………………………………………………....................................49
7. Appendix…………………………………………………………………………………50
8. Bibliography……………………………………………………….......................................51
1. The Digital Twin Concept

1.1. Historical background

The concept of Digital Twin (DT) is not new: the premises for its evolution date back to the space race
more than 50 years ago. NASA's need to ensure the safe return of the spacecraft to Earth under critical
conditions prompted engineers to develop simulators processed with real-time data from the physical
spacecraft in space, analyzing possible scenarios and calculating the optimal decision to instruct crew
members to maneuver the spacecraft. That application demonstrated the potential of virtual simulation
in linking physical and virtual spaces. A virtual simulation model reflects the constraints of physical
assets without errors and direct training on physical assets to extract appropriate solutions through
virtual simulation.

1.2. Concept definition and evolution

In 2002 M. Grieves defined DT for the first time in his course "Product Life Cycle Management:
Virtual representation of what has been produced."

Subsequently a milestone for the definition of Digital Twins was set by NASA itself in collaboration
with the United States Air Force in the field of aerospace equipment maintenance: ''A Digital Twin is a
multi-physics, multiscale, probabilistic integrated simulation of an as-built vehicle or system that uses
the best available physical models, sensor updates, fleet history, etc., to mirror the life of its
corresponding flying twin.''

In the following decade, the concept of DT began to grow exponentially in popularity. The most
relevant additions to the definition are as follows can be seen in Table1.1.

1
Table 1.1 Digital twin definition evolution
Author Definition

Chen 2017 [12] ‘‘A digital twin is a computerized model of a physical device or system that
represents all functional features and links with the working elements.’’

“The digital twin is actually a living model of the physical asset or system, which
Liu et al. 2018 [39]
continually adapts to operational changes based on the collected online data and
information and can forecast the future of the corresponding physical
counterpart.’’

ZHENG et al. 2018 ‘‘A Digital Twin is a set of virtual information that fully describes a potential or
[78] actual physical production from the micro atomic level to the macro geometrical
level.’’

‘‘A digital twin is a digital representation of a physical item or assembly using

integrated simulations and service data. The digital representation holds
VRABIČ et al. 2018 information from multiple sources across the product life cycle. This
[65] information is continuously updated and is visualized in a variety of ways to
predict current and future conditions, in both design and operational
environments, to enhance decision making.’’

2
Although the concept of DT has evolved over the years and achieved a high degree of complexity and
completeness, some aspects still make it difficult to distinguish it from other technological solutions: in
general, in many applications it is not easy to distinguish DT from general computational methods,
simulations and similar concepts such as digital model, digital shadow and digital twin.

As can be seen in Figure 1.1. The discriminant with respect to the technological solutions mentioned
above is the flow of data: a digital model is described as a digital version of a pre-existing or designed
physical object; to properly define a digital model, there must not be an automatic exchange of data
between the physical model and the digital model. The digital shadow, on the other hand, is a digital
representation of an object that has a unidirectional flow between the physical object and the digital
object. A change in the state of the physical object leads to a change in the digital object, not vice versa.
In the digital twin if data flows between an existing physical object and a digital object, and is fully
integrated in both directions, this constitutes the ''Digital Twin'' reference. A change made to the
physical object automatically leads to a change in the digital object and vice versa.

Figure 1.2. Differences among Digital Model, Digital Shadow, and Digital Twin.

3
1.3. Digital Twin basic structure

Originally proposed as a three-dimensional model by Grieves, the five-dimensional extension

model proposed by Tao et al. (2018) [62] laid the foundation most of the frameworks proposed in
the subsequent literature. The dimensions (three initial and two extended) are:

1. Physical;

2. Virtual;

3. Service;

4. Data;

5. Connections.

It is important to note that the DT is not just a virtual representation of an object, but can encapsulate
an entire process, i.e., a complete diagnostic procedure, with the necessary equipment.

data acquisition, flow and management, connections, and algorithms. The correlation between these
layers is illustrated in Figure 1.2.

DT data can be in a multitude of forms, such as physical sensor signals, virtual signals, manuals, tables,
databases. Localization can occur simultaneously in the system itself, in adjacent (auxiliary) systems that
may or may not be part of the DT itself (even if their data are), and in the cloud. In addition, this data
may be raw (e.g., voltage, current, flow, counts, size) or processed (e.g., health indices, state values,
grouped or labeled). Therefore, proper management is of critical importance.

4
Figure 1.2. DT layers correlation.

1.4. Manufacturing application

Manufacturers are always seeking ways to track and monitor products to save time and money, a core
factor and motivation for any manufacturer. That is why Digital Twins seem to have the most
significant impact in this context. Connectivity is one of the main drivers for the manufacturing sector
to use Digital Twins.

The Digital Twin can provide real-time machine performance status and feedback from the production
line. This enables the manufacturer to predict problems in advance. The use of the Digital Twin
increases connectivity and feedback between devices, in turn improving reliability and performance.
Artificial intelligence algorithms coupled with the Digital Twins have the potential for greater accuracy
as the machine can hold large amounts of data, which is needed for performance analysis and
forecasting.

The Digital Twin is creating an environment for testing products and a system that operates on real-
time data, which in a production environment has the potential to be an extremely valuable asset.

This review explored the services that can be grouped into these categories:
5
 1) Real-time monitoring of the performance and health status of the physical asset;
2) Energy efficiency analysis;
3) Detection and diagnosis of product failures;
4) Predictive maintenance: optimized maintenance strategy obtained by analyzing historical and
real-time data of system status by using simulation models and optimization algorithms;
5) Performance prediction: historical and real-time data are used to perform a reliable prediction of
future production system state through simulation;
6) Human Operator Analysis, used to obtain the operations done from the users and/or giving
them user guidance to visualize the system updates with a user-friendly HMI (Human-Machine
Interface);
7) Scheduling optimization in the production planning process;
8) Plant layout optimization in a reconfigurable manufacturing system;
9) Exchange data between other systems such as interaction between Digital Twin and other ERP
systems;
10) Controlling, automated feedback from the digital twin to the physical system.

2. Literature survey and data collection

This review aims to investigate the practical implementations of DTs in manufacturing systems in
industrial or laboratory environments by understanding the main application purpose of the created DTs,
what services are offered among those described in Chapter 1, and how the architecture for data
acquisition and simulation (including dataset acquisition and protocol, model, and software used) was
constructed.

The literature search was conducted using the Scopus database: search strings were limited to article
titles, abstracts, and keywords.

As can be seen from Fig. 2.1a. the concept of DT increased in popularity especially in 2018, so the scope
of research was narrowed by limiting the time frame between 2018 and 2023.

6
A large percentage of the research results came from the United States, Germany and China (Fig 2.1.b.,
Fig 2.1.c), which are leading the race toward Industry 4.0. A small number of researchers and research
organizations contributed nearly 40 percent of the total number of articles on this topic.

Figure 2.1a. Statistics from Scopus database (TITLE-ABS-KEY ("digital twin" or ("virtual twin") and
("manufacturing" or ("production system"). Documents by year.

Figure 2.1b. Documents by country or territory.

7
Figure 2.1c. Documents by affiliation.

Table 2.1. and Fig 2.2. illustrate the procedure for selecting articles in this review and the number of
articles identified accordingly are as follows. A total of 2698 papers were identified after searching
Scopus using the search strings presented in Table 3 for paper selection.

Next, 2638 papers were identified by limiting the search period to “January 2018 to January 2023.” This
was further narrowed down to 1215 after limiting the document type to “Article” and “Review” 940 after
limiting the subject area to “Engineering,” and 804 after limiting the language to “English”, 740 after
limiting publication stage “Final” and finally 418 by setting “All Open Access”.

Subsequently, a screening procedure, was performed to determine the relevance of the article by
reading the abstract, methodology and conclusion through which 68 articles were identified.

The criteria for papers exclusion from the screening procedure are as follows:

1. Articles presented only content related to the concept of DT with no real application cases;
2. Some of the articles did not explain the development of the DT and the architecture to support
it;
3. Articles didn’t present a complete compliance to the DT definition given in Chapter 1;
4. Display of methods to improve DT itself rather than apply DT;
5. Articles were not downloadable.

8
 A specific search was carried out for the welding application of DT through which 4 articles were
identified, in this case the same selection and screening criteria of the first search were used but with the
search string: ("digital twin" or “virtual twin”) and (“welding” or “welder”) and (“manufacturing” or
“production system”).

As a result, 72 final articles were included for further analysis, all reporting DT applications in industrial
or laboratory manufacturing environments.

The following Table 2.2. reports all the articles clustered by their application target. Articles that were
covered by more than one application target were assigned to multiple clusters as can be seen in the Venn
Diagram (Fig2.3) while articles with very specific applications for which no common features could be
identified were assigned in the cluster named “Others”. The classification and analysis of each cluster will
be explained in depth in the next chapter.

Table 2.1. Selection criteria

Figure 2.2. Selection process.

9
Table 2.2. Articles clusterized by application target.
Application Author-Year-Reference Descritpion

Production line Ashtari Talkhestani (2019) [6] An architecture for a Digital Twin and its
required components is proposed, with which
use cases such as plug and produce, self-x and
predictive maintenance are enabled.

A novel process evaluation method based on

Liu et al. (2019) [38] digital twin technology. It shows how to
evaluate the process plan with the dynamic
change of the machining condition and
uncertain available manufacturing resources.

A two-phase digital-twin-assisted fault diagnosis

method using deep transfer learning (DFDD),
Xu et al. (2019) [70]
which realizes fault diagnosis both in the
development and maintenance phases is
presented.

Bambura et al. (2020) [7] Feasibility demonstration of the DT

implementation under real conditions of a
production plant that is specializing in
manufacturing of the aluminum components for
the automotive industry.

Demonstration of how performance losses

Barni et al. (2020) [9]
induced by highly variable cycle times can be
recovered using a digital twin.

The paper proposes a way to integrate a Digital

Negri et al. (2020) [48] Shadow simulation model with the
Manufacturing Execution System (MES) in this
way creating a DT. The MES integrated DT is
used for decision making thanks to the presence
of an intelligence layer that hosts the rules and
the knowledge to choose among alternatives.

10
 Presentation of a novel methodology for
process automation design, enhanced
Perez et al. (2020) [52]
implementation, and real-time monitoring in
operation based on creating a digital twin of the
manufacturing process.

Demonstration of digital twin capability to

predict production status and provide
Abonyi & Ruppert (2020) [1]
information for monitoring of production
performance thanks to the real time position
data acceleration data and adaptive simulation
models.

Modeling method of Digital Twin process

model for manufacturing process. Acquisition
Zhao et al. (2020) [77] method of real-time data and the management
method of simulation data are discussed.

Define the necessary steps for the development

of DTs and for their integration into
Barbieri et al. (2021) [8] manufacturing systems through a DT
architecture. A methodology based on virtual
commissioning is proposed.

Presentation of a smart execution control

Bavelos et al. (2021) [10] framework for enabling the autonomous
operation of flexible mobile robot workers. A
DT is deployed.

Modeling methods for rapidly creating a virtual

model and the connection implementation
Jiang et al. (2021) [27] mechanism between a physical world production
system at a workshop level and its mirrored
virtual model.

Digital twin-based approach for designing and

Kousi et al. (2021) [29] redesigning flexible assembly systems.

11
Leng et al. (2021) [32] Presentation of a digital twins-based remote
semi-physical commissioning (DT-RSPC)
approach for open architecture flow-type smart
manufacturing systems. A digital twin system is
developed to enable the remote semi-physical
commissioning.

Complete DT structure covering all automation

Martinez at al. (2021) [42]
pyramid stages using Artificial Intelligence (AI)
to model each stage of the Automation Pyramid.

A five-step approach to planning data-driven

digital twins of manufacturing systems and their
Resman et al. (2021) [55]
processes

A decision-making framework for dynamic

Villalonga et al. (2021) [64] scheduling of cyber-physical production systems
based on digital twins.

Ward et al. (2021) [68] DT applied for real-time vision-based multiple

objects tracking of a production process.

Innovative application framework of a digital

Wu et al. (2021) [69]
twin-driven ship intelligent manufacturing
system and analysis of operation mechanism.

Approach for engineering digital twins (DTs)

that are used to train Bayesian Networks (BNs)
Ademujimi e Prabhu (2022) [2]
for fault diagnostics at equipment and factory
levels.

Highly flexible reconfigurable manufacturing

Arnarson et al. (2022) [5] system (RMS) enabled by digital twin and
wireless power transfer solution.

Digital-Twins-Driven Semi-Physical Simulation

Cheng et al. (2022) [13]
for Testing and Evaluation of Industrial
Software in a Smart Manufacturing System.

12
Ding et al. (2022) [13] Dynamic scheduling optimization of production
based on Digital Twin.

Analysis of a closed‑loop digital twin using

Eyring et al. (2022) [17]
discrete event simulation (DES) to evaluate
identification and reaction to trends in
production.

Kombaya Touckia et al. (2022) DT application for reconfigurable

[28] manufacturing systems (RMS).

Ma et al. (2022) [40] Digital Twin and big data technologies

application in a sustainable smart manufacturing
strategy based on information management
systems for energy-intensive industries from the
product lifecycle perspective.

DT application in a flexible manufacturing

Magalhaes et al. (2022) [41]
system (FMS).

How real-time monitoring and simulation of

Matsunaga et al. (2022) [45] industrial energy consumption can optimize
processes and reduce energy waste, potential
saving opportunities in manufacturing energy
consumption and costs.

Simulation, process, and validation of each

Mendi (2022) [46] phase of product life cycle to discover the
potential problems before the production of real
components. The use of digital twin technology
in the commercial production phase of the
automotive production line is introduced.

DT framework and its application in Fest Cyber

Physical smart factory, pharmaceutical
Onaji et al. (2022)[49]
continuous crystallization system and virtual X-
ray of electric motors.

13
 Qamsane et al. (2022) [53] Evaluation of a DT Framework solution for
performance monitoring in process
manufacturing systems that aims to avoid
unplanned downtime, a prevalent challenge that
pressures profitability in manufacturing.

Miriam Ugarte Querejeta et al. Implementation of a digital twin solution for

(2022) [63] design prototyping and commissioning practices
in a Cyber Physical System (CPS).

DT application in a flexible and reconfigurable

Yang et al. (2022) [71]
manufacturing system context. The paper
Discussion of the architectural design and
implementation of the application, an
information model, and an assessment model
that enable quantitatively assessment on
reconfigurations of manufacturing systems from
various aspects.

Dynamic performance analysis and vulnerability

Zhang et al. (2022) [75]
Evaluation for a smartphone Digital Twin
workshop under temporal and spatial
disruptions.

A model framework of intelligent workshop

Zhang et al. (2022) [76]
manufacturing system based on a digital twin is
proposed, driving the deep information
integration among the physical entity, data
collection, and information decision-making.

A digital twin framework is developed to detect,

Kumbhar et al. (2023) [30]
diagnose, and improve bottleneck resources
using utilization-based bottleneck analysis,
process mining, and diagnostic analytics.

Feasibility demonstration of the DT

Bambura et al. (2020) [7]
CNC Machine implementation under real conditions of a

14
 production plant that is specializing in
manufacturing of the aluminum components for
the automotive industry.

Modeling methods for rapidly creating a virtual

Jiang et al. (2021) [27] model and the connection implementation
mechanism between a physical world production
system at a workshop level and its mirrored
virtual model.

DT applied for real-time vision-based multiple

Ward et al. (2021) [67] objects tracking of a production process.

Innovative application framework of a digital

twin-driven ship intelligent manufacturing
Wu et al. (2021) [69]
system and analysis of operation mechanism.

Highly flexible reconfigurable manufacturing

Arnarson et al. (2022) [5] system (RMS) enabled by digital twin and
wireless power transfer solution.

Digital twin architecture and system based on an

interoperable data model. How to build a digital
Choi et al. (2022) [14]
twin for the integrated control monitoring using
edge devices, data analytics, and realistic 3D
visualization.

Dynamic scheduling optimization of production

Ding et al. (2022) [16]
based on Digital Twin.

DT applied to a machine tool for speed and

Guo et al. 2022 [23] accuracy simulation and monitoring.

Digital Twin-driven reconfigurable fixturing

Hu (2022) [26] optimization for trimming operation of aircraft
skins.

15
 Kombaya Touckia et al. (2022) DT application for reconfigurable
[28] manufacturing systems (RMS).

Magalhaes et al. (2022) [41] DT application in a flexible manufacturing

system (FMS).

Simulation, process, and validation of each

Mendi (2022) [46]
phase of product life cycle to discover the
potential problems before the production of real
components. The use of digital twin technology
in the commercial production phase of the
automotive production line is introduced.

DT application for dynamic scheduling and

Zhang et al. (2022) [74] adaptive control in a manufacturing cell.

Bilberg e Malik (2019) [11] Discussion of an object-oriented event-driven

Human-Robot
simulation as a digital twin of a flexible assembly
Collaboration
cell coordinated with a robot to perform
assembly tasks alongside human.

DT architecture for design, simulation and

Havard et al. (2019) [25] optimizaion cyber-physical production system
and interaction with it remotely or in a
collaborative way using virtual reality.

Digital twin based synchronized control and

Kuts et al. (2019) [31] simulation of an industrial robotic cell using
virtual reality.

A novel process evaluation method based on

digital twin technology. It shows how to
Liu et al. (2019) [37] evaluate the process plan with the dynamic
change of the machining condition and
uncertain available manufacturing resources.

This paper presents the use of a Virtual Reality

Oyekan et al. (2019) [50]
digital twin of a physical layout as a mechanism

16
 to understand human reactions to both
predictable and unpredictable robot motions.

Presentation of a novel methodology for

process automation design, enhanced
Perez et al. (2020) [52] implementation, and real-time monitoring in
operation based on creating a digital twin of the
manufacturing process.

Presentation of a smart execution control

framework for enabling the autonomous
Bavelos et al. (2021) [10] operation of flexible mobile robot workers. A
DT is deployed.

Digital twin-based approach for designing and

Kousi et al. (2021) [29]
redesigning flexible assembly systems.

A Digital Twin Demonstrator to enable flexible

Martinez et al. (2021) [43] manufacturing with robotics.

Diachenko et al. (2022) [15] Industrial collaborative robot digital twin

integration and control using robot operating
system.

DT approach for human–robot interactions

Gallala et al. (2022) [19]
(HRIs) in hybrid teams using Industry 4.0
enabling technologies, such as mixed reality, the
Internet of Things, collaborative robots, and
artificial intelligence.

Robotic Arm Manipulation Based on Mixed

Mourtzis et al. (2022) [47] Reality in a collaborative manufacturing cell.

Demonstration of digital twin capability to

Transportation predict production status and provide
system information for monitoring of production
Abonyi & Ruppert (2020) [1]
performance thanks to the real time position
17
 data acceleration data and adaptive simulation
models.

Presentation of a smart execution control

Bavelos et al. (2021) [10]
framework for enabling the autonomous
operation of flexible mobile robot workers. A
DT is deployed.

Glatt et al. (2021) [21] Modeling and implementation of a digital twin

of material flows based on physics simulation.

Digital twin-based approach for designing and

Kousi et al. (2021) [29]
redesigning flexible assembly systems.

DT design concept based on external service for

Martinez-Gutierrez et al. (2021) the transportation of the Automatic Guided
[44] Vehicles (AGVs) which are being recently
introduced for the Material Requirement
Planning satisfaction in a collaborative industrial
plant.

DT applied for real-time vision-based multiple

Ward et al. (2021) [68] objects tracking of a production process.

Application of a simulation-based Digital Twin

for predicting distributed manufacturing control
Roque Rolo et al. (2021) [57]
system performance.

Digital Twins-driven semi-physical simulation

Cheng et al. (2022) [13] for testing and evaluation of Industrial Software
in a Smart Manufacturing System.

Digital Twin-Based automated guided vehicle

Han et al. (2022) [24]
scheduling: a solution for its charging problems.

DT framework and its application in Festo

Onaji et al. (2022) [49] Cyber Physical smart factory, pharmaceutical

18
 continuous crystallization system and virtual X-
ray of electric motors.

DT application in a flexible and reconfigurable

Yang et al. (2022) [71]
manufacturing system context. The paper
Discussion of the architectural design and
implementation of the application, an
information model, and an assessment model
that enable quantitatively assessment on
reconfigurations of manufacturing systems from
various aspects.

Dynamic performance analysis and vulnerability

Zhang et al. (2022) [75]
Evaluation for a smartphone Digital Twin
workshop under temporal and spatial
disruptions.

Welding Aivaliotis et al. (2019)[4] Methodology to calculate the Remaining Useful

Process Life (RUL) of machinery equipment by utilizing
physics-based simulation models and Digital
Twin concept, in order to enable predictive
maintenance for manufacturing resources using
Prognostics and health management (PHM)
techniques.

A method for identification and sequence

Tabar et al. (2019) [61] optimization of geometry spot welds in a digital
twin context.

Quality prediction and control of assembly and

Li et al. (2020) [34]
welding process for ship group product based
on Digital Twin.

Roy et al. (2020) [58] DT model developed for an advanced

manufacturing process named friction stir
welding.

19
 Wang et al. (2020) [66] Deep learning-empowered digital twin for
visualized weld joint growth monitoring and
penetration control.

The Digital Twin is employed to monitor the

Aivaliotis et al .(2021) [5] health status of the robot and ensure the
convergence of the simulated to the actual robot
behavior: The output of the simulation is used
to estimate the future behavior of the robot and
make predictions for the quality of the products
to be produced, as well as to estimate the robot’s
Remaining Useful Life.

A detection and configuration method for

Li et al. (2022) [33] welding completeness in the automotive
body‑in‑white panel based on digital twin.

Additive Yavari et al. (2021) [72] Detecting flaws in laser powder bed fusion using
Manufacturing a DT solution.

Process monitoring of economic and

Yi et al. (2021) [73]
environmental performance of a material
extrusion printer using an augmented reality-
based digital twin.

Liu et al. (2022) [36] This paper proposes a novel Digital Twin-
enabled collaborative data management
framework for metal additive manufacturing
systems, where a Cloud DT communicates with
distributed Edge DTs in different product
lifecycle stages.

Implementation of digital twin ecosystem that

Pantelidakis et al. (2022) [51] can be used for testing, process monitoring, and
remote management of an additive

20
 manufacturing–fused deposition modeling
machine in a simulated virtual environment.

DT implementation for process monitoring in

Reisch et al. (2022) [54]
Wire arc additive manufacturing.

Ergonomics monitoring during manufacturing

Others Greco et al. (2020) [54] production.

Displacement field perception method for

Liang et al. (2021) [35] component Digital Twin in aircraft assembly.

Rodriguez-Guerra et al. (2021) DT application on fault-tolerant control systems

[56] to overcome faults without human interaction.

Sun et al. (2022) [60] A performance prediction method for a high-

precision servo valve supported by Digital Twin.

Development of a Digital Twin for a robotic

Stan et al. (2022) [59]
deburring work cell allowing monitoring and
controlling.

Farhadi et al. (2022) [18] Digital Twin framework for an industrial robot
drilling process.

21
Fig 2.3. Venn Diagram of clusters intersections.

22
3. Cluster Analysis
Services offered by the DT and the technologies implemented were analyzed, with a focus on data
acquisition and transmission and simulation features of the physical production assets.

The analysis was carried out for each cluster except Welding Process and Additive Manufacturing, this
is due to too small number of articles per cluster and the difficulty of getting a general overview in DT
implementation. In any case, the classification of implemented technologies for these clusters can be
seen in the Appendix.

Regarding data acquisition and transmission, when mentioned, the following aspects were analyzed:

• Sensors and Hardware for data collection from the shopfloor;

• Communication Protocol and architecture network. They represent a critical factor in the
creation of the DT. State synchronization between a Digital Twin and its counterpart in
physical space is based on two-way communication of real-time data. In this regard, industrial
communication protocols are always used to facilitate information exchange between physical
devices, processing streams of data or communication with middleware platforms such as the
Manufacturing Execution System (MES) and Enterprise Resource Planning (ERP);
• Process Data Streams Software: Digital twin applications, such as real-time monitoring,
forecasting and control, impose a strict latency requirement for data processing due to
continuous real-time analysis and queries. To this end, middleware platforms that enable
continuous connectivity were analyzed;
• Data Storage options: the amount of data involved makes data storage and management a
crucial aspect of the DT application, for this purpose the different data storage options used
were analyzed.

As for the simulation features used in the twinning process, it was investigated:

• Modeling Software;
• Model Type, e.g. 3D, DES (Discrete Event Simulation), FEM (Finite Element Method) etc.

3.1. Production Line cluster

The information about technologies implemented in detail can be found in the Table 3.1.

23
Table 3.1. Data Acquisition and Transmission-Simulation Features of Production Line Cluster
Data Acquisition and Transmission Simulation features
Reference Sensors and Hardware Process Data Streams Software Data storage Communication Protocol Model Type Software Model Name
Kumbhar et al.(2023)[30] SCADA system DES FACTS Analyzer 2.0
Ma et al.(2022)[40] RFID tags MapReduce, Storm Stream XML, B2MML Not Specified
Eyring et al.(2022)[17] (PLC) PTC Inc., kepware, ThingWorx PostgreSQL EtherNet/IP 3D DES FlexSim
ReCap Autodesk(3D scanning software), Inventor Redis, MySQL,
Ding et al.(2022)[16] RFID tags Autodesk. Influx Data Not specified 3D Unreal Engine
Magalhaes et al.(2022)[41] (PLC) RFID tags HMI TCP 3D V88-113D CIMSoft Amatrol
Matsunaga et al. (2022) Fluke 434 energy anlyzer, CCK7200 Power multimeter,
[45] Beckhoff CLP Energy Platform 2D 3D DES Tecnomatix Plant Simulation
Zhang et al.(2022)[75] PLC 3D 2D DES Tecnomatix Plant Simulation
E52-Temperature, Telaire Dust density, technometer
Mendi (2022) [46] frequency Apache Kafka, Apache Flink MQTT 3D Unity
Microcontrollers, Raspberry pi, wifi wireless
Arnarson et al.(2022)[5] comunication OPC UA 3D Kinematic Visual Components
Kombaya Touckia et DES Combinatorial Simulink MATLAB (SimEvents and
al.(2022)[28] MySql, MongoDb Sequential Stateflow)
Cheng et al.(2022)[13] PLC, RFID tags SCADA system OPC UA 3D
M.Ugarte Querejeta et Visual Components, Virtual twinCAT
al.(2022)[63] TwinCAT controllers OPC, REST API 3D controllers
Ademujimi e Prabhu
(2022)[2] Proximity, temperature, and vibration sensors TCP 3D DES Simio DES , RobotStudio

Yang et al. (2022) [71] XLM TCP/IP 2D 3D DES Tecnomatix Plant Simulation
Zhang et al.(2022) [76] RFID tags, PRID camera OPC/UA Mathematical MATLAB

Qamsane et al.(2022)[53] PID controller 3D DES Simulink and Simscape(MATLAB)

Leng et al (2021) [49] RFID tags OPC UA 2D, 3D, CAD DES Siemens NX, Tecnomatix Plant Simulator

Ward et al. (2021) [68] RFID, cameras D435 Intel RealSense, FESTO PLC's CoDesys (controlling function) OPC UA, Profinet 2D 3D CAD DES Tecnomatix Plantsim(DES model)
Leng et al (2021) [32] SCADA system XML, MySQL OPC/Modbus 3D jMonkeyEngine
SCADA system, TIA portal, OPC Scout, Process
Martinez et al.(2021)[42] Cameras Simulate SQL OPC Mathematical LabVIEW

Barbieri et al.(2021)[8] Raspberry Pi, light barrier sensors OPC 3D DES Simulink (MATLAB), Experior, CoDesys

Kousi et al.(2021) [29] JSON, URDF ROS 3D DES Gazebo, WITNESS(DES MODEL)
RFID tags, inductive sensors, machine vision system,
Resman et al.(2021)[55] Rasp berry pie SCADA system SQLite OPC 3D DES Tecnomatix Plant Simulation
MongoDB, MySQL,
Wu et al. (2021) [69] RFID JSON OPC UA 3D Unity3D, PhysX, 3dsMAX

Basler camera, RealSense camera, ROBOTCEPTION-

Bavelos et al. (2021) 160 camera, SICK laser scanner, AprilTAG marker URDF ROS 3D Gazebo

Villalonga et al. (2021) [64] Siemens PLC, RFID tags, accelerometer MongoDB OPC UA Mathematical
Jiang et al. (2021) [27] RFID tags Apache Kafka Redis OPC UA CAD DES model
Barni et al. (2020) 3D DDDSimulator

Ruppert e Abonyi(2020)[1] ProM 3D DES Tecnomatix Plant Simulation, AutoCAD

Bambura et al.(2020)[7] PLC MySQL TCP/IP 2D 3D DES Tecnomatix Plant Simulation

Simulink MATLAB (Level-2 MATLAB S-
Negri et al. (2020) [48] RFID tags XML OPC UA, TCP/IP Functions)
Perez et al. (2020) [52] PLC, FARO Focus 3D scanner 3D Unity3D
RFID tags and bar codes, laser rangefinder, infrared
Zhao et al.(2020)[77] rangefinder 3D DES Tecnomatix Plant Simulation
Ashtari OPC UA, EtherCAT, CAN-Bus, Siemens NX, Open Modelica, Simulink
Talkhestani(2019)[6] XML Profibus 3D (MATLAB)

Xu et al.(2019)[70] Process Visibility System (Envision) 2D 3D DES Process Designer & Process Simulate
Liu et al.(2019)[38] RFID tags XML OPC

24
3.1.1 Services

In Table 3.1.1. can be seen all the articles of this cluster are classified by the services mentioned in the
Chapter1.

The monitoring service was provided by the totality of the articles analyzed, so it was decided to
mention “Monitoring (generic)” only for those that did not mention other specific services in addition
to it.

Table 3.1.1. Provided services by DT in the Production Line cluster.

Article Service

Mendi (2022)[46], Ward et al.(2021)[68], Magalhaes et al.(2022)[41] Monitoring

(generic)
Ma et al.(2022)[40], Eyring et al.(2022)[17], Arnarson et al.(2022)[5], Cheng et Performance
al.(2022)[13], Yang et al.(2022)[71], Onaji et al.(2022)[49], Wu et al.(2021)[69], Prediction
Villalonga et al.(2021)[64], Kumbhar et al.(2023)[30], Kombaya Touckia et al.
(2022)[28]

Villalonga et al.(2021)[64], Ding et al.(2022)[16], Barni et al.(2020)[9], Negri et al. Scheduling

(2020)[48], Zhao et al.(2020)[77], Liu et al.(2019)[38], Bavelos et al.(2021)[10], Optimization
Kumbhar et al.(2023)[30], Ruppert e Abonyi(2020)[1]

Ma et al.(2022)[40], Matsunaga et al.(2022)[45] Energy Efficiency

Kousi et al.(2021)[29], Zhang et al.(2022)[75], Qamsane et al.(2022)[53], Wu et Anomaly Detection

al.(2021)[69], Bambura et al.(2020)[7], Xu et al.(2019)[70], Ademujimi e Prabhu
(2022)[2], Kumbhar et al.(2023)[30]

Arnarson et al.(2022)[5], Kombaya Touckia et al.(2022) [28], Onaji et al.(2022) Plant layout
[49], Wu et.al[69], Bavelos et al.(2021)[10], Zhang et al.(2022)[76] Optimization

Barni et al.(2020)[9], Qamsane et al.(2022)[53] Predictive

maintenance

Arnarson et al.(2022)[5], Negri et al.(2020)[48], Leng et al.(2021)[32], Martinez Controlling

et al.(2021)[42], Kousi et al.(2021)[29], Bavelos et al.(2021)[10], Leng et al (2021)
[32]

Cheng et al. (2022)[13], Yang et al.(2022)[71], Negri et al. (2020)[48], Martinez et Exchange data
al.(2021)[42], Resman et al.(2021)[55], Jiang et al.(2021)[27], Ashtari Talkhestani between systems.
(2019)[6]

25
 Pantelidakis et al.(2022)[51], Kousi et al.(2021) [29], Bavelos et al.(2021)[10], Human Operator
Perez et al.(2020)[52] Analysis

As can be seen from the Table 3.1.1.and more easily from the Figure 3.1.1., the most common service
is Performance Prediction: data from the shop floor are analyzed and with specific algorithms and
simulations it is possible to predict the performance of the process in the future or to find alternative
solutions that optimize the given process. This is the case with services such as Plant Layout
Optimization, Scheduling Optimization and Predictive maintenance. However, as can be seen from the
Table 3.1.1., not all DT implementations that include these services also include Controlling. In fact,
this is the case where the DT implementation is incomplete: although the proposed framework also
includes automatic back action from the DT to the physical “twin”, the case study or workshop doesn’t
focus on that.

When Controlling service is provided, DT not only mimics the behavior of the actual physical asset,
but also enables autonomous, real-time two-way communication between the physical and digital parts,
thus turning two-way communication into action, triggering certain actions on the MES(Negri et al.
(2020) [48]) (Martinez at al. (2021) [42]). This capability results in the ability not only to monitor the
physical asset in real time, but also to react to events on the shop floor that might affect the supervised
production environment.

More specifically, in the articles of Production Line cluster, Controlling service applies to:
commissioning (Leng et al (2021) [32]), FMS (Flexible Manufacturing System) and RMS
(Reconfigurable Manufacturing System) in which depending on the data collected from the shopfloor
the layout automatically self-adapt (Arnarson et al.(2022)[5])(Kousi et al.(2021)[29]) or in which mobile
robots and AGV adapt to shopfloor condition and move across the plant depending on the automatic
scheduling optimization (Bavelos et al. (2021)[10]).

26
Figure 3.1.1. Frequency of provided services in the Production Line cluster.

3.1.2 Communication Protocol and Architecture Network

As can be seen from the Figure 3.1., a good portion of the articles analyzed do not mention the
communication protocols used, among those where it is present the most used are:

• OPC UA (Open Platform Communications Unified Architecture): standard that facilitate the
exchange of data between programmable logic controllers (PLCs), human-machine interfaces
(HMIs), servers, clients, and other machines for the purpose of interconnectivity and
information circulation. (Arnarson et al.(2022)[5]) (Cheng et al. (2022)[13]) (Zhang et
al.(2022)[76]) (Onaji et al.(2022)[49]) (Ward et al.(2021)[68]) (Leng et al (2021)[32]), Wu et al.
(2021) [69] (Villalonga et al.(2021)[64])(Jiang et al.(2021)[27]) (Negri et al. (2020)[48]) (Ashtari
Talkhestani (2019)[6]);
• OPC: predecessor of OPC UA, OPC is a series of standards and specifications for industrial
telecommunication (M. Ugarte Querejeta et al.(2022)[63]) (Martinez at al.(2021)[42]) (Barbieri et
al.(2021)[8]) (Resman et al.(2021)[55]) (Liu et al.(2019)[38]);

27
 • TCP/IP: communication protocol suite to interconnect network devices (Yang et al.(2022)[71])
(Bambura et al(2020)[7]) (Negri et al.(2020)[48]);
• TCP: main protocol for enabling two hosts to exchange data (Ademujimi e Prabhu (2022)[2])
(Magalhaes et al.(2022)[41]).

Figure 3.1.2. Frequency of Communication Protocol in the Production Line cluster.

Regarding processing streams of data collected, although few articles mention it, the following can be
cited: Storm Stream e MapReduce (Ma et al.(2022)[40]) distributed computing frameworks for cleansing
real-time and non-real-time data, respectively, Apache Kafka (Mendi (2022)[46]) (Jiang et al.(2021)[27])
an open-source event-streaming platform, Apache Flink (Mendi (2022) [46]) a framework used for
stateful computations on unbounded and bounded data streams e (ProM Ruppert e Abonyi (2020)[1]) a
framework for process mining technique.

With respect to data storage, the dataset is often described by specifying the database: these are
classified as relational databases and non-relational databases. Usually, relational databases are used for
applications that involve the management of complex database transactions and heavy data analysis,
because of referential integrity. Non-relational databases are geared towards managing large sets of
varied and frequently updated data, often in distributed systems. They avoid the rigid schemas
associated with relational databases.

28
In the article analyzed common non-relational database are (Redis Ding et al. (2022) [16] (Jiang et al.
(2021)[27]) (MongoDB Choi et al.(2022)[14]) (Kombaya Touckia et al(2022)[28]) (Wu et.al
[69]Villalonga et al.(2021)[64]) Influx data Ding et al. (2022) [16] and as relational database
MYSQL(Ding et al. (2022)[16]) (Kombaya Touckia et al.(2022)[28]) (Leng et al (2021)[32] (Wu et.al
[69])(Bambura et al(2020)[7]), SQLite (Resman et al.(2021)[55]), PostgreSQL (Eyring et al.(2022)[17]).

Several articles do not mention the database but the data format, among them XML e JSON are the
most common.

Concerning Sensors and Hardware for data collection, as can be seen in Table 3.0, it varies greatly
depending on the application, however, is common data collection by PLC (Programmable Logic
Controller) controllers, often in conjunction with SCADA (Supervisory Control and Data Acquisition)
and by RFID tags.

3.1.3. Simulation features

The Figure 3.1.3. above indicates that the most widely used software for modeling are:

• Tecnomatix Plant Simulation (Matsunaga et al.(2022)[45]) (Zhang et al.(2022)[75]) (Yang et

al.(2022)[71]) (Onaji et al.(2022)[49]) (Resman et al.(2021)[55]) (Ward et al.(2021)[68])(Ruppert e
Abonyi (2020)[1]) (Bambura et al(2020)[7])(Zhao et al.(2020)[77]).
• Matlab Simulink (Kombaya Touckia et al.(2022) [28])(Zhang et al.(2022)[76])(Qamsane et
al.(2022)[53])
(Barbieri et al. (2021)[8])(Negri et al.(2020)[48])(Ashtari Talkhestani(2019)[6]).

Both models allow di modelling, simulating, process and system optimization application.

Most articles model the physical system in 2D or 3D, however in several articles they do not use any
model, but extract information from the system experimentally, analyzing the acquired data (Villalonga
et al. (2021)[64]) (Martinez at al.(2021)[42])(Zhang et al.(2022)[76].

29
Figure 3.1.3. Frequency of Software Model in the Production Line cluster.

3.2. CNC Machine Cluster

All the information about technologies implemented in DT application can be found in the Table 3.2.

3.2.1. Services

In the following table 3.2.1. the articles of the cluster are classified by the services mentioned in
Chapter 1.

Table 3.2.1. Provided services by DT in the CNC Machine cluster.

Article Function

Mendi (2022)[46], Magalhaes et al.(2022)[41] Monitoring (generic)

Arnarson et al.(2022)[5], Kombaya Touckia et Performance prediction

al.(2022)[28], Hu (2022)[26], Guo et al. 2022[23], Ward et
al. (2021)[67], Wu et al.(2021)[69]

30
 Ding et al.(2022)[16], Zhang et al. (2022)[74] Scheduling optimization

Energy efficiency

Zhang et al. (2022)[74], Wu et al.(2021)[69], Bambura et Anomaly Detection

al.(2020)[7]

Arnarson et al.(2022)[5], Kombaya Touckia et Plant Layout Optimization

al.(2022)[28], Wu et al.(2021)[69]

Predictive Maintenance

Arnarson et al.(2022)[5] ,Zhang et al. (2022)[74], Hu Controlling

(2022) [26], Ward et al. (2021)[67]

Jiang et al.(2021)[27] Exchange Data Between

Systems.

Human Operator Analysis

As can be seen in the Figure 3.2.1., much like the Production Line cluster (with whom the cluster
shares 8 articles) the most common service is Performance Prediction, very often also complemented
by other services such as Anomaly Detection in the manufacturing process (Wu et al.(2021) [69]) and
Layout Optimization (Arnarson et al.(2022)[5])(Kombaya Touckia et al.(2022)[28])(Wu et al.(2021)[69]).

The second most common service is Controlling: (Hu (2022)[26]) in which a trimming operation by a
five-axis high-speed milling machine is displayed and, after monitoring and performance prediction, the
optimized solution is applied; (Ward R. et al.[67]) 5-axis high performance machining center for a
finishing operation, while in (Zhang et al. (2022)[74])controlling solution is implemented on a CNC
lathe and milling operations for scheduling purpose.

31
Table 3.1. Data Acquisition and Transmission-Simulation Features of CNC Machine cluster.
Data Acquisition and Transmission Data Acquisition and Transmission
Reference Sensors and hardware Process Data Streams
Data storage
Software Communication Protocol Model Type Software Model name
Ding et al.(2022)[16] RFID tags Redis, MySQL, Not specified 3D Unreal Engine, ReCap Autodesk(3D
Magalhaes et al.(2022)[41] (PLC) RFID tags TCP 3D V88-113D CIMSoft Amatrol
E52-Temperature, Telaire Apache
Dust density, Technometer Kafka,
Mendi (2022)[46] Frequency Apache Flink MQTT 3D Unity
Choi et al.(2022)[14] XML REST API 3D
Microcontrollers, Raspberry
Arnarson et al. (2022) [5] pi, wifi wireless OPC UA 3D kinematic Visual Components
Kombaya Touckia et MySql, DES, Combinatorial Simulink MATLAB (SimEvents and
al.(2022)[28] MongoDb Not specified and Sequential Stateflow)
Hu (2022)[26] OPC 3D FEM Catia Dessault Systeme

Mathematical (Real-
time systems modeled
as networks of timed
Zhang et al.(2022)[74] RFID OPC UA automata), 3D UPPAAL
Soldid Works, Unity 3D, Polygon
Guo et al.2022[23] PLC MySql OPC 3D Cruncher
NI USB 6343, Siemens
Ward et al.[67] ADAS , Kistler 9255c Profibus Matlab
MongoDB,
Wu et al.(2021)[69] RFID MySql, JSON OPCUA 3D Unity3D, PhysX, 3dsMAX
Jiang et al.(2021)[27] RFID tags Apache Kafka Redis OPC UA CAD DES model
Bambura et al.(2020)[7] PLC MySQL TCP/IP 2D, 3D, DES Tecnomatix Plant Simulation

32
Figure 2.2.1. Frequency of services in CNC Machine cluster.

3.2.2 Communication Protocol and Architecture Network

The Figure 3.2.2. shows the most used communication protocols:

• OPC UA (Aivaliotis et al .(2021)[5])(Wu et al. (2021)[69])(Jiang et al. (2021)[27])(Zhang et

al.(2022) [74]) and OPC (Guo et al.2022[23])(Hu (2022)[26]).

Among communication protocols there are also:

• REST API (Choi et al. (2022)[14]) a software architectural style for creating web services;
• Profibus, developed to support the machine-to-machine communications and the remote
terminal control of programmable logic controllers, for process and peripheral
control/automation.

For what concerns the process of data streams Apache Kafka (Mendi (2022)[46]) (Jiang et al.(2021)[27])
and Apache Flink (Mendi (2022)[46]) are mentioned.

With respect of Data Storage options MySQL is the most common database (Kombaya Touckia et
al.(2022)[28], Bambura et al.(2020)[7], Wu et al.(2021)[69], Guo et al.2022[23], Ding et al.(2022)[16]).
For the rest of the articles, the details can be found in Table 3.2. “Sensors and Hardware” column

33
where are also present the sensors used in data collection: again, PLC controllers and RFID tags stand
out.

Figure 3.2.2. Frequency of Communication Protocol in CNC Machine cluster.

3.2.3. Simulation Features

Regarding simulation features, as can be seen in the Figure 3.2.3. below and with more details in the
"Software Mode Name" column of Table 3.2., the most used is:

• Unity Software (Mendi (2022)[46], Guo et al. 2022[23], Wu et al.(2021)[69])

For the other articles there is great homogeneity. Same situation for what concerns the model (“Model
Type” column of Table 3.2.): besides DES model, FEM (Hu (2022)[26]), Kinematic (Arnarson et
al.(2022)[5]), Combinatorial and Sequential (Kombaya Touckia et al.(2022)[28]) can be found.

34
Figure 3.2.3. Frequency of Software Model in CNC Machine cluster.

3.3. Human-Robot Collaboration

All the information about technologies implemented in DT application can be found in the Table 3.3.

3.3.1. Services

In Table 3.3.1. articles in the Human-Robot collaboration cluster were classified according to the
services in Chapter 1.

Table 3.3.1. Provided services by DT in the Human-Robot Collaboration cluster.

Article Function

Monitoring (generic)

Gallala et al. (2022)[19], Mourtzis et al(2022)[47], Martinez et Performance prediction

al.(2021)[43], Bilberg e Malik(2019)[11], Havard et al(2019)[25],

35
 Oyekan et al.(2019)[50], Kousi et al.(2021)[29], Bavelos et
al.(2021)[10]

Bavelos et al.(2021)[10], Martinez et al.(2021)[43], Kousi et Scheduling optimization

al.(2021)[29]

Energy efficiency

Kousi et al.(2021)[29] Anomaly detection

Bavelos et al.(2021)[10], Martinez et al.(2021)[43], Kousi et Plant layout Optimization

al.(2021)[29]

Predictive maintenance

Gallala et al. (2022)[19], Diachenko et al.(2022)[15], Kousi et Controlling

al.(2021)[29], Bavelos et al.(2021)[10]

Diachenko et al.(2022)[15], Mourtzis et al(2022)[47], Liu et Exchange data between

al.(2019)[37], Havard et al.(2019)[25], Kuts et al (2019) [31], systems.
Martinez et al.(2021)[43]

Bilberg e Malik (2019) [11], Havard et al (2019) [25], Kuts et al Human operator analysis
(2019) [31], Liu et al. (2019) [37], Oyekan et al. (2019) [50], Perez
et al. (2020) [52], Bavelos et al. (2021) [10], Kousi et al. (2021)
[29], Martinez et al. (2021) [43], Gallala et al. (2022) [19], Mourtzis
et al (2022) [47]

As can be seen in the Figure 3.3.1. the service Human Operator Analysis is present in all the cluster’s
articles. In second place, as in the clusters analyzed before, the most common service is Performance
Prediction, often associated with other services as Scheduling Optimization and Plant Layout
Optimization (Martinez et al.(2021)[43]) (Bavelos et al.(2021)[10]), (Kousi et al.(2021)[29]).

The Controlling service is applied to: (Diachenko et al.(2022)[15]) Omron TM5-900 robot remotely
controlled by humans through controllers, (Gallala et al. (2022) [19]) human-robot (KUKA IIWA)
interaction remotely using KUKA’s Sunrise Workbench controller and VR tools such a MS HoloLens,
(Kousi et al.(2021)[29]) collaborative assembly line where robot movements are automatic thanks to
performance prediction of human operator, (Bavelos et al.(2021)[10]) ]) mobile UR10 robotic arm
navigation on the shopfloor environment.

36
Figure 3.3.1. Frequency of provided services in the Human-Robot collaboration cluster

37
Table 3.3. Data Acquisition and Transmission-Simulation Features of Human-Robot Collaboration cluster

Reference Sensors and hardware Software Data format Protocol Model Software
Gallala et al.(2022)[19]
Torque sensors, camera, depth sensor XML, URDF 3D Unity, Hololens MS
TCP/IP, VUFORIA
Mourtzis et al(2022)[47] Mixed reality API (RESTFUL API),
toolkit, HMI XML, URDF ROS 3D Unity 3D, MS HoloLens
Diachenko et al.(2022)[15]
URDF MQTT, ROS 3D Unity3D
Gazebo, WITNESS(DES
Kousi et al.(2021)[29]
JSON, URDF ROS 3D DES MODEL)
Martinez et al.(2021)[43] Cameras, controllers TCP/IP 3D URSim, Experior
Basler camera, RealSense camera,
ROBOTCEPTION-160 camera, SICK
Bavelos et al.(2021)[10], laser scanner, AprilTAG marker URDF ROS 3D Gazebo
Perez et al.(2020)[52] PLC, FARO Focus 3D scanner 3D Unity3D
RFID tags, Dynamometer(Kistler type
9273), Piezoelectric Accelerometer(PCB
Liu et al. (2019) [37]
model 352C65). Data acquisition cardNI OPC UA, TCP,
PXI-1031 LabVIEW XML MTConnect 3D Ms Hololens
Oyekan et al.(2019)[50] Kinect 3D Unity3D
3D CAD Unity 3D, Catia, Modelica
Havard et al(2019)[25] Perception Neuron Pro suite (VR) JSON, XML CAM (Dessault systems), VR
VirtualReality Unity 3D, 3DS Max and
Kuts et al.(2019)[31]
Toolkit ROS 3D Maya (Autodesk)
Tecnomatix Process
Bilberg e Malik(2019)[11] 3D camera, Kinect sensor 3D Simuate

38
3.3.2. Communication Protocol and Architecture Network

As can be seen from graph Figure 3.3.2., the most used communication protocols are:

• ROS (Robot Operating System) (Gallala et al.(2022)[19]) (Mourtzis et al(2022)[47]) (Kousi et

al.(2021)[29]) (Bavelos et al.(2021)[10]) (Kuts et al (2019)[31]). ROS is a software architecture
which provides the drivers that handle the data of the sensors used and feed it to a digital
world model, it acts as a middleware for the translation of the 3D motions into commands for
the actual robot, it provides advanced capabilities regarding the calculation of the robotic arm
kinematics and reverse kinematics.
• TCP/IP.

Regarding Data Format, as can be seen in the Data Acquisition and Transmission section of Table 3.3.,
the most widely used are URDF (Unified Robot Description Format) (Gallala et al.(2022)[19])
(Mourtzis et al(2022)[47]) (Kousi et al.(2021)[29]) (Bavelos et al.(2021)[10]) (Diachenko et al.(2022)[15])
used to store robot physical properties and, as in the other clusters, XML, JSON.

Figure 3.3.2. Frequency of Communication Protocol in the Human-Robot collaboration cluster.

cluster

39
3.3.3. Simulation feature

Regarding the software model used in the realization of the DT:

• Unity stands out, mainly due to the wide range of VR and MR (Mixed Reality) functionalities
(Mourtzis et al (2022) [47]) (Gallala et al. (2022) [19]) (Martinez-Gutierrez et al. (2021)[44])
(Bavelos et al. (2021)[10]) (Kuts et al (2019) [31]). Infact the use of Unity is often accompained
by the use of VR or MR softwre such as MS HoloLens (Gallala et al. (2022) [19]) (Mourtzis et
al (2022) [47]).

Figure 3.3.3. Frequency of Software Model in the Human-Robot Collaboration cluster

cluster.

3.4. Transportation System

All the information about technologies implemented in DT application can be found in the Table 3.4.

3.4.1 Services

In Table 3.4.1 below the cluster’s articles are classified according to the services explained in Chapter 1.

40
Unity stands out, mainly due to the wide range of VR and MR (Mixed Reality) functionalities (Mourtzis
et al (2022) [47]) (Gallala et al. (2022) [19]) (Martinez-Gutierrez et al. (2021)[44]) (Bavelos et al.
(2021)[10]) (Kuts et al (2019) [31]). Infact the use of Unity is often accompained by the use of VR or
MR softwre such as MS HoloLens (Gallala et al. (2022) [19]) (Mourtzis et al (2022) [47]).

Controlling service is provide: (Bavelos et al.(2021)[10]) AGV transportation in a human-robot

collaborative environment, (Roque Rolo et al. (2021)[57]), distribution control system through MAS
(multi agent system) based on conveyors, (Mourtzis et al (2022) [47]AGV transportation controlled by
Arduino Nano and Arduino Mega 2560 micro crontrollers , (Leng et al.(2021)[32]) to optimize travel
time optimization of autonomous robot carriers (Robotino) in a production line.

Table 3.4.1. Provided Services by DT in Transportation System cluster

Article Function

Ward et al.(2021)[68] Monitoring (generic)

Martinez-Gutierrez et al.(2021)[44], Roque Rolo et al.(2021)[57], Performance prediction

Mourtzis et al(2022)[47], Cheng et al.(2022)[13], Yang et al.(2022)
[71], Leng et al.(2021)[32], Bavelos et al.(2021)[10], Zhang et
al.(2022)[76].

Roque Rolo et al.(2021)[57], Bavelos et al.(2021)[10], Zhang et Scheduling optimization

al.(2022)[76], Han et al.(2022)[24]]

Han et al.(2022)[24] Energy efficiency

Anomaly detection

Roque Rolo et al.(2021)[57], Bavelos et al.(2021)[10], Zhang et Plant layout Optimization

al.(2022)[76], Leng et al.(2021)[32], Kousi et al.(2021)[32]

Predictive maintenance

Martinez-Gutierrez et al.(2021)[44], Roque Rolo et al. (2021)[57], Controlling

Bavelos et al.(2021)[10], Kousi et al.(2021)[29], Mourtzis et
al(2022)[47], Leng et al.(2021)[32]

Cheng et al.(2022)[13], Ma et al.(2022)[40] Exchange data between

systems.

Ruppert e Abonyi(2020)[1], Bavelos et al.(2021)[10], Kousi et Human operator analysis

al.(2021)[29]

41
Table 3.4. Data Acquisition and Transmission-Simulation Features of Transportation System cluster

Data Acquisition and transmission Simulation Feature

Reference Sensors and hardware Software Data format Protocol Model Software
Cheng et al.(2022)[13] PLC, RFID tags SCADA system OPC UA 3D Not specified
Han et al.(2022)[24] Modbus TCP 3D FlexSim

Tecnomatix Plant
Yang et al.(2022)[71 XLM TCP/IP 2D 3D DES Simulation
Zhang et al.(2022)[76] RFID, PRID camera OPC/UA Mathematical MATLAB
Siemens NX, Tecnomatix
Leng et al (2021) [32] RFID OPC UA 2D, 3D, CAD DES Plant Simulation
RFID, cameras D435 Intel CoDesys OPC UA, Tecnomatix Plant
Ward et al.(2021)[68] RealSense, FESTO PLC's (controlling Profinet 2D, 3D, CAD, DES Simulation
Gazebo, WITNESS(DES
Kousi et al.(2021)[29] JSON, URDF ROS 3D DES MODEL)

Martinez-Gutierrez et al. (2021) [44] LIDAR, RFID JSON, XML MQTT 3D Gazebo

Roque Rolo et al. (2021) [57] JSON API 3D 2D AnyLogic

Basler camera, RealSense
camera, ROBOTCEPTION-
160 camera, SICK laser
Bavelos et al.(2021)[10] scanner, AprilTAG marker URDF ROS 3D Gazebo
Arduino nano, arduino mega
Mourtzis et al (2022) [47] 2560 OPC UA 3D pyBullet
Tecnomatix Plant
Ruppert e Abonyi (2020) [1]
ProM Not specified 3D DES Simulation, AutoCAD

42
Figure 3.4.1 Frequency of services provided in the Transportation System cluster

3.4.2 Communication Protocol and Architecture Network

From the Figure 3.4.2. below communication the protocols found are:

• OPC UA protocol. It is still the most widely used.

Beside OPC UA, among communication protocols can be found:

• TPC;
• MQTT;
• Profinet (Process Field Network) protocol (Ward et al.(2021)[68]) it is an industry technical
standard for data communication over Industrial Ethernet, designed for collecting data from,
and controlling equipment in industrial systems;
• Modbus protocol (Han et al.(2022)[24]), developed to support the machine to machine
communications and the remote terminal control of programmable logic controllers, for
process and peripheral control/automation, in this case implemented with TCP in order to
facilitate inter-communication at a higher level such MES and ERP systems.

43
 • control of programmable logic controllers, for process and peripheral control/automation, in
this case implemented with TCP in order to facilitate inter-communication at a higher level
such MES and ERP systems.

Figure 3.4.2. Frequency of Communication Protocol in the Transportation System

cluster

In general, as we can see in Table 3.4. there is little information regarding the software used to process
the data streams, while in terms of data format we find the formats common to the other clusters such
as URDF (Bavelos et al.(2021)[10]) (Kousi et al.(2021)[29]), XML (Yang et al.(2022)[71]) (Martinez-
Gutierrez et al.(2021)[44]) and JSON (Kousi et al.(2021)[29])( Martinez-Gutierrez et al. (2021)
[44])(Roque Rolo et al.(2021)[57]).

Regarding the hardware used for data collection from the shopfloor, in addition to the common PLCs
e RFID tags, are very frequent cameras for motion caption of robot and environment (Zhang et
al.(2022)[76]) (Ma et al.(2022)[40]) (Bavelos et al.(2021)[10]), LIDAR (Light Detection and Ranging)
and other laser scanners (Bavelos et al.(2021)[10])( Martinez-Gutierrez et al.(2021)[44]) mostly used for
the navigation of autonomous vehicles. In the article (Rodriguez-Guerra et al.(2021)[56]) data are
collected from Arduino micro controllers.

44
3.4.3 Simulation features

As can be seen from the Figure 3.4.3 below, the most used software is:

• Tecnomatix and the transport system was mostly modeled with a DES model.

Among the other software can be found:

• WITNESS software (Kousi et al.(2021)[29]), used to model material flows in the assembly line;
• AnyLogic (Roque Rolo et al.(2021)[57])an agent-based simulation for the modeling of
transportation system;
• pyBullet (Mourtzis et al(2022)[47]) to model mechanical interaction between rigid bodies, used
to model AGV and possible disturbances to its movement

Figure 3.4.3. Frequency of Software Model in the Transportation System cluster.

45
Chapter 4. Overall findings and trends
In summary, it seems that the application of DT requires almost the same structure regardless of the
service offered and the physical asset twinned.

The schema identified is showed in Fig. 4. The System is represented with the layers mentioned in
Chapter 1, the construction of the Digital Shadow starts from the data acquisition in a virtual
environment, that can be done with different protocols. Successively, the acquired data are analyzed
and processed with the use of digital models that open the way to representing the real system in the
chosen virtual environment with a (simulation) software. These first steps are needed to give the final
shape to the Digital Shadow. When the Digital Shadow is bidirectionally connected to the main
controller of the real system, the overall system becomes a proper DT.

Below (Figure 4.0), a framework representing the trends observed in the review is proposed with the
main technologies that can meet the DT implementation requirements for each application target
considered.

Figure 4. General DT application framework.

46
 As for data collection from sensors, especially PLC and RFID tags, and communication between the
different layers of the DT, the OPC UA and TCP/IP protocols are proposed for the Production Line
cluster (Figure 4.1.), CNC Machine (Figure 4.2.), Transportation System (Figure 4.3.) and Human-
Robot Collaboration. These types of data exchange standard are a safe, reliable, manufacturer
independent, and platform-independent industrial communication. They enable secure data exchange
between hardware platforms from different vendors and across operating systems.

To meet the very strict requirements for data analysis Apache Kafka, used by many companies for
high-throughput data pipelines, flow analytics, and data analysis and Apache Flink, used for stateful
computations on unbounded and bounded data streams, are proposed. In these frameworks, signals
from the DT platform were stored on Apache Kafka and then forwarded to Apache Flink for analysis.
MySQL and SQLite, both relational databases, are proposed for the data storage.

Figure 4.1. DT Production Line application framework.

47
For Production Line and Transportation System the simulation model Siemens’ Tecnomatix Plant
Simulation (a platform for agent-based/ discrete event simulations (DES)) is the main trend, it can be
used for real-time supervision, control, and visualization of the virtual twin. Tecnomatix uses an
object-oriented modeling method. Modeling of production systems is realized by implementation of
virtual objects which represent individual production equipment. Machines, workers, transport, and
logistic systems such as conveyors, trucks, loaders, warehouse, buffers and other storage objects
occurring in manufacturing companies. It can be used for analytics to achieve logistic process
improvement, material flow optimization and efficient resource usage. For the Transportation System,
Gazebo Software is often present in the implementations: it is particularly suitable for representing
human operators and mobile workstations in a human-robot collaboration environment. The only
differentiator between the application of DT to the CNC Machine and to the Production line is the
modeling software, on the CNC Machine the trends are Unity 3D and Matlab Simulink, used for
visualization of the physical asset in the virtual environment.

Figure 4.4. DT CNC Machine application framework.

48
Figure 4.4. DT Transportation System application framework.

Remarkably different, however, is the framework for implementation in a human-robot collaboration

context (Figure 4.4.). ROS (Robot Operating System), a reliable standard for industrial robot
integration, has been proposed for data collection, especially with cameras and RFID tags,
communication, modeling and control of robots. ROS compatibility software is provided by many
manufacturers and enables the implementation of modular control units by unifying development
practices with the same libraries and methods.

49
Figure 4.3. DT Human-Robot collaboration application framework.

In terms of the virtual layer visualization model, the most widely used software are Unity and Gazebo,
both integrated with ROS, which provide a simple but powerful development environment with a
modular approach to programming and also offer integration with all commercially available VR
systems, such as MS HoloLens, ideal for human-robot collaboration, in particular robot training for
human motion recognition and human training for virtual commissioning or to learn specific task in the
industrial environment.

50
Summary
This thesis brings a contribution to the research on this topic, by exploring the DT application
methodologies and related services in the manufacturing environment, enabling DT technology
classification by application target, and finally trying to find a framework for DT implementation
following the trends in the industry.

This analysis identified some misalignments between the implementation of DT and its description in
the literature. First, many studies claim to implement DT without doing so completely. An example is
often the lack of technical requirements such as bidirectionality of data between the virtual model and
the physical model and thus the lack of fundamental services such as controlling. Also, in terms of the
technologies used, many studies omit software and data management methodology, which are
fundamental pillars for DT implementation. Moreover, in many cases, DT implementation is only
partially illustrated and in the early stages, making it clear that DT research is still in an embryonic stage.
The latter aspect can also be observed in many cases of articles in which, even for the same purposes,
there is a wide variety of solutions adopted, without a clear trend being identifiable.

51
Acknowledgements
I would like to offer my special thanks to my thesis advisor Giulia Bruno for patiently helping and
following me during the writing of the thesis and of course to my family: my parents Cecilia and
Remus, my siblings Ramona and Silviu. We the best, this degree is as much yours as it is mine.

52
Appendix

Table 5. Data Acquisition and Transmission-Simulation Features of Additive Manufacturing cluster.

Data acquisition and Transmission Simulation feature

Reference Sensors and hardware Software Data format
Protocol Model Software
Raoberry Pi 3, Thermocouple sensors, IR
Yi et al.(2021)[73] distance sensors Octoprint REST API 3D Unity
Reisch et al. (2022) [54] Welding camera, voltage sensor, temperature 3D CAD Siemens NX1926
BSE detector, MeltPoos system, EOSTATE EthernetIP, Profinet, TCP, HTTP, REST MANUELA(pilot line)
Liu et al. (2022) [36] ExposureOT, PLC XML API (restful api) 3D project
Yavari et al. (2021) [72] Photodetectors, X-ray tomography Netfabb
Pantelidakis et al. (2022) [51] ZMPT101B, ACS12 ProBas API 3D CAD Unity, Siemens NX

Table 6. Data Acquisition and Transmission-Simulation Features of Welding Process.

Data acquisition and Transmission Simulation feature

Reference Sensors and hardware Software Data format Protocol Model Software
Vision and proximity
Li et al. (2022)[33]
sensors JSON MQTT 3D Open Modelica
Qr codes, circular Unity3D,
Li et al.(2020)[34]
codes Vuforia, HoloLens2 HTTP 3D HoloLens2
Hall sensor, optical
code sensor,
Wang et al.(2020)[66]
acceleration sensor,
RFID tags, bar codes XML OPC UA 3D
Arc sensors, industrial
Tabar et al.(2019)[61] camera (Point Grey
FL3FW03S1C) IPC 3D Unity
Roy et al.(2020)[58] 3D, FEM RD&T
Aivaliotis et al.(2021)[5] TCP/IP Data acquisition
Labview
Matlab, Open
Aivaliotis et al.(2019)[4]
3D modelica

53
Bibliography

1. Abonyi, J & Ruppert, T. Integration of real-time locating systems into digital twins. Journal of
Industrial Information Integration 20, (2020).

2. Ademujimi, T. & Prabhu, V. Digital Twin for Training Bayesian Networks for Fault
Diagnostics of Manufacturing Systems. Sensors 22, (2022).

3. Aivaliotis, P., Arkouli, Z., Georgoulias, K. & Makris, S. Degradation curves integration in
physics-based models: Towards the predictive maintenance of industrial robots. Robotics and
Computer-Integrated Manufacturing 71, (2021).

4. Aivaliotis, P., Georgoulias, K. & Chryssolouris, G. The use of Digital Twin for predictive
maintenance in manufacturing. International Journal of Computer Integrated Manufacturing 32, 1067–
1080 (2019).57

5. Arnarson, H., Mahdi, H., Solvang, B. & Bremdal, B. A. Towards automatic configuration and
programming of a manufacturing cell. Journal of Manufacturing Systems 64, 225–235 (2022).

6. Ashtari Talkhestani, B. et al. An architecture of an Intelligent Digital Twin in a Cyber-Physical

Production System. At-Automatisierungstechnik 67, 762–782 (2019).

7. Bambura, R., Šolc, M., Dado, M. & Kotek, L. Implementation of digital twin for engine block
manufacturing processes. Applied Sciences (Switzerland) 10, (2020).

8. Barbieri, G. et al. A virtual commissioning based methodology to integrate digital twins into
manufacturing systems. Production Engineering 15, 397–412 (2021).

9. Barni, A., Pietraroia, D., Züst, S., West, S. & Stoll, O. Digital twin based optimization of a
manufacturing execution system to handle high degrees of customer specifications. Journal of
Manufacturing and Materials Processing 4, (2020).

10. Bavelos, A. C. et al. Enabling flexibility in manufacturing by integrating shopfloor and process
perception for mobile robot workers. Applied Sciences (Switzerland) 11, (2021).

11. Bilberg, A. & Malik, A. A. Digital twin driven human–robot collaborative assembly. CIRP
Annals 68, 499–502 (2019).

12. Chen, Y. Integrated and Intelligent Manufacturing: Perspectives and Enablers. Engineering 3,
588–595 (2017).

54
13. Cheng, K., Wang, Q., Yang, D., Dai, Q. & Wang, M. Digital-Twins-Driven Semi-Physical
Simulation for Testing and Evaluation of Industrial Software in a Smart Manufacturing System.
Machines 10, (2022).23

14. Choi, S., Woo, J., Kim, J. & Lee, J. Y. Digital Twin-Based Integrated Monitoring System:
Korean Application Cases. Sensors 22, (2022).

15. Diachenko, D. et al. Industrial Collaborative Robot Digital Twin integration and Control Using
Robot Operating System. Journal of Machine Engineering (2022) doi:10.36897/jme/148110.

16. Ding, G., Guo, S. & Wu, X. Dynamic Scheduling Optimization of Production Workshops
Based on Digital Twin. Applied Sciences (Switzerland) 12, (2022).

17. Eyring, A., Hoyt, N., Tenny, J., Domike, R. & Hovanski, Y. Analysis of a closed-loop digital
twin using discrete event simulation. International Journal of Advanced Manufacturing Technology 123,
245–258 (2022).

18. Farhadi, A., Lee, S. K. H., Hinchy, E. P., O’Dowd, N. P. & McCarthy, C. T. The Development
of a Digital Twin Framework for an Industrial Robotic Drilling Process. Sensors 22, 7232 (2022).

19. Gallala, A., Kumar, A. A., Hichri, B. & Plapper, P. Digital Twin for Human–Robot Interactions
by Means of Industry 4.0 Enabling Technologies. Sensors 22, 4950 (2022).

20. Glaessgen, E. H. & Stargel, D. S. The digital twin paradigm for future NASA and U.S. air force
vehicles. in (2012).
21. Glatt, M. et al. Modeling and implementation of a digital twin of material flows based on physics
simulation. Journal of Manufacturing Systems 58, 231–245 (2021).

22. Greco, A., Caterino, M., Fera, M. & Gerbino, S. Digital twin for monitoring ergonomics during
manufacturing production. Applied Sciences (Switzerland) 10, 1–20 (2020).

23. Guo, M., Fang, X., Hu, Z. & Li, Q. Design and research of digital twin machine tool simulation
and monitoring system. International Journal of Advanced Manufacturing Technology (2022)
doi:10.1007/s00170-022-09613-2.

24. Han, W. et al. Digital Twin‐Based Automated Guided Vehicle Scheduling: A Solution for Its
Charging Problems. Applied Sciences (Switzerland) 12, (2022).

25. Havard, V., Jeanne, B., Lacomblez, M. & Baudry, D. Digital twin and virtual reality: a co-
simulation environment for design and assessment of industrial workstations. Production and
Manufacturing Research 7, 472–489 (2019).

55
26. Hu, F. Digital Twin‐Driven Reconfigurable Fixturing Optimization for Trimming Operation of
Aircraft Skins. Aerospace 9, (2022).

27. Jiang, H., Qin, S., Fu, J., Zhang, J. & Ding, G. How to model and implement connections
between physical and virtual models for digital twin application. Journal of Manufacturing Systems
58, 36–51 (2021).

28. Kombaya Touckia, J., Hamani, N. & Kermad, L. Digital twin framework for reconfigurable
manufacturing systems (RMSs): design and simulation. International Journal of Advanced
Manufacturing Technology 120, 5431–5450 (2022).26

29. Kousi, N. et al. Digital Twin for Designing and Reconfiguring Human–Robot Collaborative
Assembly Lines. Applied Sciences 11, 4620 (2021).

30. Kumbhar, M., Ng, A. H. C. & Bandaru, S. A digital twin based framework for detection,
diagnosis, and improvement of throughput bottlenecks. Journal of Manufacturing Systems 66, 92–
106 (2023).

31. Kuts, V., Otto, T., Tähemaa, T. & Bondarenko, Y. DIGITAL TWIN BASED
SYNCHRONISED CONTROL AND SIMULATION OF THE INDUSTRIAL ROBOTIC
CELL USING VIRTUAL REALITY. Journal of Machine Engineering 19, 128–144 (2019).

32. Leng, J. et al. Digital twins-based remote semi-physical commissioning of flow-type smart
manufacturing systems. Journal of Cleaner Production 306, (2021).

33. Li, H. et al. A detection and configuration method for welding completeness in the automotive
body-in-white panel based on digital twin. Scientific Reports 12, (2022).

34. Li, L., Liu, D., Liu, J., Zhou, H. & Zhou, J. Quality Prediction and Control of Assembly and
Welding Process for Ship Group Product Based on Digital Twin. Scanning 2020, 1–13 (2020).

35. Liang, B. et al. A Displacement Field Perception Method for Component Digital Twin in
Aircraft Assembly. Sensors 20, 5161 (2020).

36. Liu, C. et al. Digital Twin-enabled Collaborative Data Management for Metal Additive
Manufacturing Systems. Journal of Manufacturing Systems 62, 857–874 (2022).

37. Liu, C., Vengayil, H., Lu, Y. & Xu, X. A Cyber-Physical Machine Tools Platform using OPC
UA and MTConnect. Journal of Manufacturing Systems 51, 61–74 (2019).

38. Liu, J. et al. Dynamic Evaluation Method of Machining Process Planning Based on Digital
Twin. IEEE Access 7, 19312–19323 (2019).

56
39. Liu, Z., Meyendorf, N. & Mrad, N. The role of data fusion in predictive maintenance using
digital twin. in vol. 1949 (2018).
40. Ma, S., Ding, W., Liu, Y., Ren, S. & Yang, H. Digital twin and big data-driven sustainable smart
manufacturing based on information management systems for energy-intensive industries.
Applied Energy 326, (2022).

41. Magalhães, L. C. et al. Conceiving a Digital Twin for a Flexible Manufacturing System. Applied
Sciences 12, 9864 (2022).

42. Martinez, E. M., Ponce, P., Macias, I. & Molina, A. Automation pyramid as constructor for a
complete digital twin, case study: A didactic manufacturing system. Sensors 21, (2021).

43. Martinez, S. et al. A Digital Twin Demonstrator to enable flexible manufacturing with robotics:
a process supervision case study. Production and Manufacturing Research 9, 140–156 (2021).

44. Martínez-Gutiérrez, A. et al. Digital Twin for Automatic Transportation in Industry 4.0. Sensors
21, 3344 (2021).

45. Matsunaga, F., Zytkowski, V., Valle, P. & Deschamps, F. Optimization of Energy Efficiency in
Smart Manufacturing Through the Application of Cyber–Physical Systems and Industry 4.0
Technologies. Journal of Energy Resources Technology 144, 102104 (2022).

46. Mendi, A. F. A Digital Twin Case Study on Automotive Production Line. Sensors 22, (2022).

47. Mourtzis, D., Angelopoulos, J. & Panopoulos, N. Closed-Loop Robotic Arm Manipulation
Based on Mixed Reality. Applied Sciences (Switzerland) 12, (2022).

48. Negri, E., Berardi, S., Fumagalli, L. & Macchi, M. MES-integrated digital twin frameworks.
Journal of Manufacturing Systems 56, 58–71 (2020).

49. Onaji, I., Tiwari, D., Soulatiantork, P., Song, B. & Tiwari, A. Digital twin in manufacturing:
conceptual framework and case studies. International Journal of Computer Integrated Manufacturing 35,
831–858 (2022).

50. Oyekan, J. O. et al. The effectiveness of virtual environments in developing collaborative

strategies between industrial robots and humans. Robotics and Computer-Integrated Manufacturing 55,
41–54 (2019).

51. Pantelidakis, M., Mykoniatis, K., Liu, J. & Harris, G. A digital twin ecosystem for additive
manufacturing using a real-time development platform. International Journal of Advanced
Manufacturing Technology 120, 6547–6563 (2022).

57
52. Pérez, L., Rodríguez-Jiménez, S., Rodríguez, N., Usamentiaga, R. & García, D. F. Digital Twin
and Virtual Reality Based Methodology for Multi-Robot Manufacturing Cell Commissioning.
Applied Sciences 10, 3633 (2020).

53. Qamsane, Y. et al. Open Process Automation- and Digital Twin-Based Performance Monitoring
of a Process Manufacturing System. IEEE Access 10, 60823–60835 (2022).

54. Reisch, R. T. et al. Context awareness in process monitoring of additive manufacturing using a
digital twin. Int J Adv Manuf Technol 119, 3483–3500 (2022).

55. Resman, M., Protner, J., Simic, M. & Herakovic, N. A five-step approach to planning data-
driven digital twins for discrete manufacturing systems. Applied Sciences (Switzerland) 11, (2021).

56. Rodriguez-Guerra, J. et al. On Fault-Tolerant Control Systems: A Novel Reconfigurable and

Adaptive Solution for Industrial Machines. IEEE Access 8, 39322–39335 (2020).

57. Roque Rolo, G., Dionisio Rocha, A., Tripa, J. & Barata, J. Application of a Simulation-Based
Digital Twin for Predicting Distributed Manufacturing Control System Performance. Applied
Sciences 11, 2202 (2021).

58. Roy, R. B. et al. Digital twin: current scenario and a case study on a manufacturing process. Int J
Adv Manuf Technol 107, 3691–3714 (2020).

59. Stan, L., Nicolescu, A. F., Pupăză, C. & Jiga, G. Digital Twin and web services for robotic
deburring in intelligent manufacturing. Journal of Intelligent Manufacturing (2022)
doi:10.1007/s10845-022-01928-x.72

60. Sun, X., Liu, S., Bao, J., Li, J. & Liu, Z. A Performance Prediction Method for a High-Precision
Servo Valve Supported by Digital Twin Assembly-Commissioning. Machines 10, 11 (2021).

61. Tabar, R. S., Wärmefjord, K. & Söderberg, R. A method for identification and sequence
optimisation of geometry spot welds in a digital twin context. Proceedings of the Institution of
Mechanical Engineers, Part C: Journal of Mechanical Engineering Science 233, 5610–5621 (2019).

62. Tao, F., Zhang, H., Liu, A. & Nee, A. Y. C. Digital Twin in Industry: State-of-the-Art. IEEE
Transactions on Industrial Informatics 15, 2405–2415 (2019).
63. Ugarte Querejeta, M. et al. Implementation of a holistic digital twin solution for design
prototyping and virtual commissioning. IET Collab Intel Manufact 4, 326–335 (2022).

64. Villalonga, A. et al. A decision-making framework for dynamic scheduling of cyber-physical

production systems based on digital twins. Annual Reviews in Control 51, 357–373 (2021).

58
65. Vrabič, R., Erkoyuncu, J. A., Butala, P. & Roy, R. Digital twins: Understanding the added value
of integrated models for through-life engineering services. in vol. 16 139–146 (2018).
66. Wang, Q., Jiao, W. & Zhang, Y. Deep learning-empowered digital twin for visualized weld joint
growth monitoring and penetration control. Journal of Manufacturing Systems 57, 429–439 (2020).

67. Ward, R. et al. Machining Digital Twin using real-time model-based simulations and lookahead
function for closed loop machining control. International Journal of Advanced Manufacturing
Technology 117, 3615–3629 (2021).

68. Ward, R., Soulatiantork, P., Finneran, S., Hughes, R. & Tiwari, A. Real-time vision-based
multiple object tracking of a production process: Industrial digital twin case study. Proceedings of
the Institution of Mechanical Engineers, Part B: Journal of Engineering19Manufacture 235, 1861–1872
(2021).

69. Wu, Q., Mao, Y., Chen, J. & Wang, C. Application research of digital twin-driven ship
intelligent manufacturing system: Pipe machining production line. Journal of Marine Science and
Engineering 9, (2021).

70. Xu, Y., Sun, Y., Liu, X. & Zheng, Y. A Digital-Twin-Assisted Fault Diagnosis Using Deep
Transfer Learning. IEEE Access 7, 19990–19999 (2019).

71. Yang, J., Son, Y. H., Lee, D. & Noh, S. D. Digital Twin-Based Integrated Assessment of
Flexible and Reconfigurable Automotive Part Production Lines. Machines 10, 75 (2022).

72. Yavari, R. et al. Digitally twinned additive manufacturing: Detecting flaws in laser powder bed
fusion by combining thermal simulations with in-situ meltpool sensor data. Materials and Design
211, (2021).

73. Yi, L., Glatt, M., Ehmsen, S., Duan, W. & Aurich, J. C. Process monitoring of economic and
environmental performance of a material extrusion printer using an augmented reality-based
digital twin. Additive Manufacturing 48, 102388 (2021).

74. Zhang, C., Zhou, G., Jing, Y., Wang, R. & Chang, F. A Digital Twin-Based Automatic
Programming Method for Adaptive Control of Manufacturing Cells. IEEE Access 10, 80784–
80793 (2022).

75. Zhang, D., Pei, Y. & Liu, Q. Vulnerability Evaluation for a Smartphone Digital Twin
Workshop under Temporal and Spatial Disruptions. Machines 10, (2022).

59
76. Zhang, Z., Zhu, Z., Zhang, J. & Wang, J. Construction of intelligent integrated model
framework for the workshop manufacturing system via digital twin. International Journal of
Advanced Manufacturing Technology 118, 3119–3132 (2022).

77. Zhao, P. et al. The Modeling and Using Strategy for the Digital Twin in Process Planning.
IEEE Access 8, 41229–41245 (2020).

78. Zheng, Y., Yang, S. & Cheng, H. An application framework of digital twin and its case study.
Journal of Ambient Intelligence and Humanized Computing 10, 1141–1153 (2019).

60
61