Digital Twin Paradigm: A Systematic Literature Review: Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti

Digital twin paradigm: A systematic literature review
Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti
To cite this version:

Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti. Digital twin paradigm:
A systematic literature review. Computers in Industry, Elsevier, 2021, 130, pp.103469.
�10.1016/j.compind.2021.103469�. �hal-03218786�
HAL Id: hal-03218786

https://hal.archives-ouvertes.fr/hal-03218786
Submitted on 5 May 2021
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
Digital Twin Paradigm: A Systematic Literature Review
Concetta Semeraro*,1,2, Mario Lezoche2, Hervé Panetto2,

Michele Dassisti1
1
Department of Mechanical, Mathematics and Management (DMMM), Politecnico di Bari, Bari- Italy (e-mail:
{concetta.semeraro, michele.dassisti} @poliba.it).
2
Université de Lorraine, CNRS, CRAN, Nancy, France (e-mail: {concetta.semeraro, mario.lezoche,
herve.panetto}@univ-lorraine.fr)
Highlights
● A systematic literature review is conducted to explore the main features, research and technical challenges
in conceiving and building Digital Twins.
● Topic Modelling Analysis has been implemented to provide an up-to-date picture of the digital twin.
● Formal Concept Analysis (FCA) has been applied to understand the digital twin trends and strategies.
Abstract:
Manufacturing enterprises are facing the need to align themselves to the new information
technologies (IT) and respond to the new challenges of variable market demand. One of the key
enablers of this IT revolution toward Smart Manufacturing is the digital twin (DT). It embeds a
“virtual” image of the reality constantly synchronized with the real operating scenario to provide
sound information (knowledge model) to reality interpretation model to draw sound decisions. The
paper aims at providing an up-to date picture of the main DT components, their features and
interaction problems. The paper aims at clearly tracing the ongoing research and technical
challenges in conceiving and building DTs as well, according to different application domains and
related technologies. To this purpose, the main questions answered here are: ‘What is a Digital
Twin?’; ‘Where is appropriate to use a Digital Twin?’; ‘When has a Digital Twin to be developed?’;
‘Why should a Digital Twin be used?’; ‘How to design and implement a Digital Twin?’; ‘What are
the main challenges of implementing a Digital Twin?’. This study tries to answer to the previous
questions funding on a wide systematic literature review of scientific research, tools, and
technicalities in different application domains.
Keywords: Digital Twin; Industry 4.0; Cyber-Physical Systems; Predictive manufacturing.
Introduction
In the past, due to the lack of information technologies, the physical space played the main role in
controlling the production in shop floors, leading to low efficiency, accuracy, and transparency. Until
the 20th century, technologies such as computers, simulation tools, Internet, and wireless networks
introduced a parallel virtual space to virtualize physical assets and to enable the cooperation with
assets remotely. This has provided a possibility to conduct plans and operations more efficiently and
1
effectively (Tao and Zhang, 2017). Nowadays, with the developments of new generation information
technologies (New IT), the integration and the interaction between the physical and virtual spaces is
becoming increasingly important. This will create new potentialities for improving the current
operating situations and technologies in the fields of design, manufacturing, and service (Büchi et al.,
2020). Various countries are converging on this trend as the next industrial revolution (Suh, 1984),
(Prasad, 1989), and have proposed related national strategies, such as the “Industry 4.0” in Germany
(Kagermann et al., 2013); the “Advanced Manufacturing” or “Smart Manufacturing” in the United
States (Yao et al., 2019); the “Society 5.0” in Japan; the “Made in China 2025” in China; the “Industry
of the Future” in France; the “Intelligent Factory” in Italy (Osterrieder et al., 2019) and more generally
“The Factory of Future” in Europe (Drath and Horch, 2014). Although the strategies are proposed
under different environments, their common objective is to capture the opportunity brought by the
integration of the physical and virtual spaces (Hermann et al., 2016). The fusion of the physical and
virtual spaces is motivated to ensure a better flexibility and scalability of manufacturing systems
through information technologies (Dassisti and De Nicolò, 2012), (Pirola et al., 2020). The current
digital transformation of enterprises requires the design and application of digital models, called
digital twins, which represent a set of knowledge of the real processes (Panetto et al., 2019), (Dassisti
et al., 2019a). The digital twin (DT) aims at creating high-fidelity virtual models for each physical
entity to emulate their states and behaviours with abilities of evaluating, optimizing, and predicting
(Graessler and Poehler, 2017). The concept of using “twins” dated back to NASA’s Apollo program,
where two identical space vehicles were built to allow mirroring the conditions of the space vehicle
during the mission. Professor Grieves at the University of Michigan firstly put forward the concept
of ‘Digital Twin’ in Product Life cycle Management (PLM) courses in 2003 (Grieves and Vickers,
2017). The digital twin refers to a holistic, digital engineering view from the product design and
development to production planning, production engineering, production, and associated services
(Product Life cycle Management). The DT can be developed for each phase of the product life cycle
absolving different functions (Dassisti and Semeraro, 2018). The digital twin in the design stage can
help designers to configure and validate more quickly the future scenarios (Brettel et al., 2014). The
DT can help decision maker to accurately interpret the market demands and the customer preferences
(Semeraro et al., 2019b). At the manufacturing phase, DT may enable the simulation, and thus the
decision maker, to analyse the interactive behaviours among production factors by collecting data
from order, design, purchase, production planning, manufacturing, and product usage stage. The DT
can help optimizing and evaluating in real time the production planning and the behaviour of the
production process. At the service stage, DT relies on real time state monitoring and virtual operations
such as maintenance to predict the remaining life of components or products (Lee and Kim, 2018).
The virtual replication of a physical system is a rather complex task and therefore it requires the
availability of a large amount of data and models that represent the modelled system, even though
there are not specific criteria to follow (Park et al., 2019).
In the scientific literature reviewed in this paper, several studies have been devoted to analysing the
DT concept, which results different as the context of application changes (aerospace, manufacturing,
city management). In each context, digital twins have their own specificity within the life cycle phase
of the product: namely design, manufacturing, and service. As a result, each application of DT varies
depending on a different perspective and needs accordingly. In this context, the paper aims at
2
providing an advanced and up-to-date picture of the state-of-the-art considering the main features and
challenges of existing scientific research on DT’s, focussing on the different application domains and
their related technologies. The paper funds its scientific basis on information, principles and hints
derived from a systematic scientific literature review employing text mining techniques to identify
textual patterns, topic modelling, and new insights. Hereof, seven main research questions are raised
and discussed: (1) ‘What is a Digital Twin?’; (2) ‘Where is appropriate to use a Digital Twin?’; (3)
‘Who is doing Digital Twins?’; (4) ‘When has a Digital Twin to be developed?’; (5) ‘Why should a
Digital Twin be used?’; (6) ‘How to design and implement a Digital Twin?’; (7) ‘What are the main
challenges of implementing a Digital Twin?’. The research questions concern six main aspects
namely: the Digital Twin definition; the application contexts; the life cycle phases; the functions; the
architecture and the components, the research challenges. These aspects have been discussed in detail
to define and explore the main features, research and technical challenges in conceiving and building
Digital Twins. To serve this purpose, Formal Concept analysis (FCA) was run on to get deep into the
definition of DT life cycle phases and its functions and into the DT architecture and its components.
The outcome is then a multi-perspective picture of the Digital Twin, forming a paradigm emerging
from scientific literature.
The content is structured as the following: the literature review methodology is described in section
1, the state of art in section 2, the digital twin paradigm in section 3. The conclusion and the research
challenges are argued in section 4.
1 Literature Review Methodology
The literature review focuses on works related to DT technology. A systematic literature search was
conducted in the Scopus, Elsevier and ScienceDirect database, covering most of the peer-reviewed
interdisciplinary research papers. The methodology applied is composed by three-step approach:
Paper selection; Extraction of DT features; Knowledge representation. Each step is described below.
Step 1: Paper Selection: Publications identification and screening
The present study forms a state-of-the-art on digital twin. The focal point of the study was based on
DT representation in different scientific papers. This review was conducted based on content analysis.
The Scopus, Elsevier and ScienceDirect scientific databases were used to find the literature for this
review. In addition to ‘Digital Twin’, search terms such as ‘Factory of Future’, ‘Industry 4.0
technologies’, ‘Cyber-physical system’, ‘Predictive manufacturing’ were used to search for suitable
papers within the targets and scopes of this review paper. We found over 300 papers from our search.
The relevant literature was selected by analysing the title, abstract, keywords, paper contents and
journal's main topic of interest. Finally, we selected the papers based on impact factor, citation, and
review process. We identified and analysed 150 papers of which 35 in the fields of ‘Factory of
Future’, ‘Industry 4.0 technologies’, ‘Cyber-physical system’, ‘Predictive manufacturing’ and 115 in
the field of ‘Digital Twin’. The selected papers on the digital twin present the following parameters:
Time Span: 2002–2020; Language: English; Type = “Article”; “Journal Paper”; “Conference
Proceeding”; “Book Chapter”, as recapped in Figure 1.
3
Figure 1: Digital Twin Paper Distribution as type of Bibliographical Reference
Step 2 Extraction of DT Features: Text mining analysis
Following the publications identification and screening, a technical approach has been designed and
applied to extract all possible features and information from the selected DT items. Text mining
analysis has been selected for this specific purpose. Text mining is the process of analysing text to
extract information that can be useful for different purposes (Hearst, 2003). A set of text mining
techniques have been used employing Orange tool (Ljubljana, 2005). It allows to design and create
workflows by linking predefined or user-designed components called widgets. Two different models
were developed in Orange to this aim, as shown in figure 2 and in figure 3, for analysing the selected
papers to capture key concepts, trends, and hidden relationships in DT studies.
The first model aims to apply the text mining techniques to DT definitions to clarify what is a digital
twin and why it matters. The workflow below shows that all DT definitions were collected, listed,
and pre-processed to perform the hierarchical clustering algorithm (HCA) (Grira et al., 2004). HCA
is an unsupervised clustering technique that groups similar objects into groups called clusters. The
distance between two clusters is computed by the cosine distance because it is a good measure of
semantic relatedness (Mikolov et al., 2013). The endpoint is a hierarchy of nested clusters, called
dendrogram, where each cluster is distinct from each other cluster, and the objects within each cluster
are broadly similar to each other. In this model, the HCA groups the DT definitions in a set of clusters
that can be visualized in a data table or in a word cloud respectively named “Data Table (Clusters
DT Definitions)” and “Word Cloud of each Cluster”. The clusters and their respective word clouds
are discussed in section 2 to initiate and structure a comprehensive review on the state of art of Digital
Twins comparing the definitions provided in literature.
4
Figure 2: First model designed in Orange to automatically discover clusters in the DT definitions set
The second model has been designed for identifying which topics are the most debated and discussed
in the selected DT papers. The statistical model that has been used is the topic modelling as shown in
Figure 3. Topic modelling concerns using a text-mining tool for discovering hidden semantic
structures in a text body. In our review, for each DT paper the authors name, title, abstract, keywords
and content were collected, listed, and pre-processed for discovering the “topics” that occur in our
selection employing the Latent Dirichlet allocation (LDA). Latent Dirichlet allocation is a statistical
model that automatically detects a set of topic modelling, classifies papers, and estimate their
relevance to various topics. The outcome is reported in “Data Table (Topic Modelling)”. A paper
typically contains multiple topics in different proportions; thus, LDA also reports the topic weight
per paper and this can be visualized in “Distributions (Topic Modelling)”. The results will be
discussed in section 3 to design and compose a digital twin paradigm.
Figure 3: Second model designed in Orange for discovering Topic Modelling in DT papers
Step 3: Knowledge Representation: Formal Concept Analysis
A more detailed look through the topics discovered in step 2 was felt necessary to highlight the digital
twin trends and strategies. This further analysis is conducted by Formal Concept Analysis (FCA).
Formal Concept Analysis (FCA) is a mathematical theory oriented at applications in knowledge
representation (Agrawal et al., 1993). It provides tools to group the data and to discover formal
patterns by representing it as a hierarchy of formal concepts organised in a semi-ordered set named
lattice (Wille, 2002). In formal concept analysis (FCA), a formal context is a triple K = (O, A, R),
where O and A are non-empty sets, and R is a binary relation between O and A (R ⊆ O × A) (Ganter,
Stumme, and Wille 2005). The formal context (O, A, I) of an input matrix of n rows and m columns
5
consists of a set of objects defined as O = {𝑂𝑏𝑗1 , 𝑂𝑏𝑗2 , 𝑂𝑏𝑗3 , 𝑂𝑏𝑗𝑛 }, a set of attributes defined as
A={𝐴𝑡𝑡𝑟1 , 𝐴𝑡𝑡𝑟2, 𝐴𝑡𝑡𝑟3 … 𝐴𝑡𝑡𝑟𝑚 } and a binary relation R defined as 𝑂𝑏𝑗𝑖 , 𝐴𝑡𝑡𝑟𝑗 ∈ 𝑅 if and only if the
intersection of i-th row and j-th column is not blank (Škopljanac-Mačina and Blašković, 2014). The
FCA data table is composed by the set of objects (O) in rows and the set of attributes (A) in columns
as shown in Table 1. In our review, the objects are the papers selected in step 1, while the attributes
are all the topics identified by the model presented in step 2, Figure 3. The symbol “•” denotes that
there is a relationship (R) between the object and the attribute.
Table 1: FCA Table
O 𝑨𝒕𝒕𝒓𝟏 𝑨𝒕𝒕𝒓𝟐 𝑨𝒕𝒕𝒓𝟑 𝑨𝒕𝒕𝒓𝒎

A
𝑶𝒃𝒋𝟏 • •
𝑶𝒃𝒋𝟐 •
𝑶𝒃𝒋𝟑 • •
𝑶𝒃𝒋𝒏 • • •
Given a set of objects (O), a set of attributes (A), and defined the relations (R) between objects and
attributes, a formal concept represents a subset of objects sharing the same subset of attributes, as
displayed in Figure 4. Each node in the figure represents a concept. A concept is constituted by two
parts: its extension which consists of all objects belonging to the concept, and its intention which
comprises all attributes shared by those objects. This understanding allows the formal discovery of
associations among concepts and consequently recognizing which concepts are closely related based
on the set of shared attributes (Valtchev, Missaoui, and Godin 2004). The results provided by Formal
concept analysis will be discussed in section 3 to explore the trends in the combination between the
identified topics and their regularity of appearance in the literature.
6
Figure 4: Example of Concept Lattice (Lezoche and Panetto, 2018)
2 State of Art Digital Twin: Definitions and Sights
Industry and academia define a digital twin in several different ways (Trauer et al., 2020). For
example, according to some, a digital twin is a virtual representation/model that interacts with the
physical system throughout its life cycle (Grieves and Vickers, 2017), (Glaessgen and Stargel, 2012).
Other widely circulated definitions regard the need to exchange information between the two spaces
involving sensors, data, and models (Lee et al., 2013), (Negri et al., 2017). Others consider a digital
twin as the cyber part of a cyber-physical system (CPS) (Alam and El Saddik, 2017), (Graessler and
Poehler, 2017). The concept of a digital twin has been investigated employing the model presented
in Figure 2 to analyse the DT definitions listed below for understanding why it matters. The data table
presents an additional column: “Belonging Cluster” that results from the application of the
hierarchical cluster algorithm (HCA). This is a way to visualize how the DT definitions are grouped.
Table 2: Digital Twin Definitions
DIGITAL TWIN DEFINITIONS

Belonging
ID Year Authors DT Definition
Cluster
“a set of virtual information constructs that fully describes a potential or
actual physical manufactured product from the micro atomic level to the
1 2002 (Grieves,2014)
macro geometrical level. At its optimum, any information that could be C1
obtained from inspecting a physical manufactured product can be obtained
7
from its Digital Twin. The Digital Twin concept model contains three main
parts: a) physical products in Real Space, b) virtual products in Virtual Space,
and c) the connections of data and information that ties the virtual and real
products together.”
“an integrated multi-physics, multi-scale, probabilistic simulation of a
(Glaessgen and
2 2012
Stargel, 2012)
complex product and uses the best available physical models, sensor updates, C1
etc., to mirror the life of its corresponding twin.”
“a cradle-to-grave model of an aircraft structure’s ability to meet mission
3 2012 (Tuegel, 2012) requirements, including sub-models of the electronics, the flight controls, the C5
propulsion system, and other subsystems.”
“a coupled model enables a digital twin of the real machine that operates in
the cloud platform in parallel with the real process and simulates the health
(Lee et al., condition with an integrated knowledge from both data driven analytical
4 2013
2013) algorithms as well as other available physical knowledge. The coupled model C3
approach first constructs a digital image of a machine from the early design
stage.”
“a product equivalent digital counterpart that exists along the product life
(Ríos et al., cycle from conception and design to usage and servicing, knows the product
5 2015
2015) past, current and possible future states, and facilitates the development of
C1
product related intelligent services.”
(Rosen et al., “a very realistic model of the current state of the process and their own
6 2015
2015) behaviour in interaction with their environment in the real world.”
C4
“a virtual representation of the real product. It has product’s information
(G. N.
from the beginning of the life until the disposal of the product. The Digital
7 2016 Schroeder et
Twin is a counter part of the physical device, machine or product in a CPS. C1
al., 2016)
It has the information related to the whole life cycle of a product.”
(Alam and El “the cyber layer of CPS, which evolves independently and keeps close
8 2017
Saddik, 2017) integration with the physical layer.”
C2
(Brenner and “a digital copy of a real factory, machine, worker etc., that is created and can
9 2017 Hummel, be independently expanded, automatically updated as well as being globally C3
2017) available in real time.”
“a digital avatar encompassing CPS data and intelligence, representing
(Ciavotta et al.,
10 2017
2017)
structure, semantics, and behaviour of the associated CPS, and providing C4
services to mesh the virtual and physical worlds.”
“a cyber-physical device of its own, which is connected to the CPPS and tries
(Graessler and to emulate the human employee through dynamically adapted values of a
11 2017
Poehler, 2017) database, which represent for example properties, preferences, work C2
schedule and skillset.”
“a set of realistic product and production process models linking enormous
(H. Zhang et amounts of data to fast simulation and allowing the early and efficient
12 2017
al., 2017) assessment of the consequences, performance, quality of the design decisions C1
on products and production line.”
“a virtual and computerized counterpart of a physical system that can exploit
(Negri et al.,
13 2017
2017)
a real-time synchronization of the sensed data coming from the field and is C5
deeply linked with Industry 4.0.”
“a bi-directional relation between a physical artefact and the set of its virtual
(Schleich et al.,
14 2017
2017)
models, enabling the efficient execution of product design, manufacturing, C1
servicing, and various other activities throughout the product life cycle.”
“a one-to-one virtual replica of a “technical asset” (e.g., machine,
component, and part of the environment). A digital twin contains models of
its data (geometry, structure, . . .), its functionality (data processing,
behaviour, . . .), and its communication interfaces. It integrates all knowledge
(Schluse et al.,
15 2017
2017)
resulting from modelling activities in engineering (digital model) and from C3
working data captured during real-world operation (digital shadow). A
Digital Twin contains models of its “data” (geometry, structure, …), its
functionality (data processing, behaviour, …) and its communication
interfaces.”
“a digital copy of a product or a production system, going across the design,
(Söderberg et
16 2017
al., 2017)
pre-production, and production phases and performing real-time C1
optimization.”
“a unique instance of the universal Digital Master model of an asset, its
(Stark et al.,
17 2017
2017)
individual Digital Shadow and an intelligent linkage (algorithm, simulation C3
model, correlation, etc.) of the two elements above.”
8
“a digital representation that contains all the states and functions of a physical
(Weber et al.,
18 2017
2017)
asset and has the possibility to collaborate with other digital twins to achieve C3
a holistic intelligence that allows for decentralized self-control.”
“a perfect digital entity of a physical system; it accurately reflects the status
(Yun et al.,
19 2017
2017)
of the corresponding physical machine. We can tightly control the system C3
through a digital twin, that is, a cyber model of the machine.”
(Autiosalo,
20 2018
2018)
“the cyber part of a Cyber-Physical System.” C2
“a virtual replica of real physical installation, which can check the
(Asimov et al., consistency for monitoring data, perform data mining to detect existing and
21 2018
2018) forecast upcoming problems, and which uses an AI knowledge engine to C5
support effective business decisions.”
(Bao et al., “a virtual model in the virtual space, and it is used to simulate the behaviour
22 2018
2018) and characteristics of the corresponding physical object in real time.” C4
“a near real-time digital image of a physical object or process that helps
optimize business performance. Two concepts of IoT (Internet of things) and
23 2018 (Lee and Kim,
IoS (Internet of Service) are combined to realis e the smart factory based
C3
2018)
on a digital twin.”
“a comprehensive digital representation of an individual product. It includes
the properties, condition, and behaviour of the real-life object through
(Haag and models and data. The digital twin is a set of realistic models that can simulate
24 2018
Anderl, 2018) its actual behaviour in the deployed environment. The digital twin is
C3
developed alongside its physical twin and remains its virtual counterpart
across the entire product life cycle.”
“a complete virtual prototype of an entire system and a one-to-one mapping
relationship. Therefore, a multi-domain digital modelling method is needed;
(Luo et al., a consistent model between the designed and the actual environment of a
25 2018
2018) machine tool should be established, which needs the real-time and accurate
C5
data mapping method; an effective machine learning algorithm to mine the
data gathered from sensors and the control system is also necessary.”
“a digital replica of the physical environment along with the operator. This
(Nikolakis et
26 2018
al., 2018)
model constrains the behaviour of the twin towards replicating the actions of C4
the physical system’s actuators.”
“a set of virtual models. These mirror images and mapping of the physical
products in the virtual space. They could reflect the whole life cycle process,
(Tao et al., as well as simulate, monitor, diagnose, predict, and control the state and
27 2018
2018b) behaviours of the corresponding physical entities. The virtual models include
C4
not only the geometric models, but also all rules and behaviours, such as
material properties, mechanical analysis, health monitoring.”
“a living model that continually adapts to change in the environment or
(Z. Liu et al.,
28 2018
2018)
operation using real-time sensory data and can forecast the future of the C4
corresponding physical assets for predictive maintenance.”
“a dynamic model in the virtual world that is fully consistent with its
(Zhuang et al., corresponding physical entity in the real world and can simulate its physical
29 2018
2018) counterpart’s characteristics, behaviour, life, and performance in a timely
C4
fashion.”
“each physical device will have its cyber part as a digital representation of
(Leng et al., the real device, culminating in the digital twin models. So, the digital twin
30 2019
2019) can monitor and control the physical entity, while the physical entity can
C3
send data to update and synchronize its virtual model.”
HCA detects five different clusters labelled: C1, C2, C3, C4, C5 as shown in Figure 5. To assist with
the interpretation and verification of each cluster, word clouds were generated to provide additional
evaluations identifying the occurrence of words shared by the grouped definitions.
9
C1
C
1
C
1
C2
C3
C4
C5
Figure 5: Hierarchical Clustering (HCA) Results - Clusters DT definitions
The cluster C1 involves the definitions provided by: (Grieves, 2014), (Glaessgen and Stargel, 2012),
(Ríos et al., 2015), (G. N. Schroeder et al., 2016), (H. Zhang et al., 2017), (Schleich et al., 2017),
(Söderberg et al., 2017). The corresponding word cloud in Figure 6 shows that the consideration of
the life cycle phases is the core point in the definition of a Digital Twin. The concept of a Digital
Twin was first mentioned in a presentation of the University of Michigan in 2002 entitled “Conceptual
Ideal for PLM”. As the concept was emerging out of the field of Product life cycle management
(PLM), (Grieves and Vickers, 2017) referred to the connection between real space and virtual space
over all phases of the product life cycle presenting all the elements of the Digital Twin: real space,
virtual space, the link for data flow from real space to virtual space, the link for information flow
from virtual space to real space and virtual sub-spaces. (Ríos et al., 2015) and (Schleich et al., 2017)
specify the product life cycle from conception and design to usage and servicing while (G. N.
Schroeder et al., 2016) suggest the existence of a Twin from the beginning of a product’s life until its
disposal. In (H. Zhang et al., 2017) the DT can integrate data in the product life cycle to accurately
simulate and assess the performance and the quality of the design decisions on products and
production lines. According to (Söderberg et al., 2017) a Digital Twin exists over the complete life
cycle, subdivided in the phases design, pre-production, and production for performing real-time
optimization. The basic idea behind a Digital Twin, in (Glaessgen and Stargel, 2012), is a high-fidelity
virtual model of the physical entities having the scope of replicating and simulating the states and
behaviours of these latter along its life.
The Digital Twin is defined as a new paradigm in simulation (Rosen et al., 2015). It extends the use
of simulation to all phases of the product life cycle (Garetti et al., 2012), (Rodič, 2017). Simulation
is the basis for design decisions, validation, and test not only for a generic device but also for
10
monitoring complete systems (Boschert and Rosen, 2016). The digital twin is the evolution of other
research fields such as Virtual Manufacturing systems, Model-based Predictive control (MPC), and
Building Information Modelling (BIM).
• Virtual manufacturing (VM) is defined as a system aimed at generating a virtual representation of
a physical system without using real facilities/entities (Onosato and Iwata, 1993). In comparison
with the characteristics of the Digital Twin, the key differences are the lack of connection between
physical and virtual entities (Jones et al., 2020).
• The Model-based predictive control uses a model of the system to make predictions about the
system’s future behaviour (Ma et al., 2011). The digital twin and MPC simulates current state to
change the future state but the DT aims to create virtual models in sync with their physical part.
• The Building Information Modelling (BIM) is a process for creating and managing a model
containing digital information about a specific asset across its whole life cycle. The major
differences between a building’s BIM and a digital twin are that BIM is designed to improve the
efficiency of design and construction and it does not work with real-time data. Whereas the Digital
twin works specifically with real-time data to monitor a physical asset and improve its operational
efficiency enabling the predictive manufacturing (Khajavi et al. 2019).
The DT is becoming increasingly relevant to Model-based systems engineering (MBSE) (Bachelor
et al., 2019). MBSE is defined as: “formalized application of modelling to support system
requirements, design, analysis, verification, and validation activities beginning in the conceptual
design phase and continuing throughout development and later life cycle phases” (International
Council on Systems Engineering (INCOSE), 2007). The simulation in MBSE is mostly considered
to be a tool for R&D departments (Boschert and Rosen, 2016). The Digital Twin extends the concept
of MBSE from engineering and manufacturing to the operation and service phases (Estefan, 2007),
(Boschert et al., 2018). Digital twins could provide an effective implementation for realising the end-
to-end integration of a system across the entire product life cycle (Cheng et al., 2020) because it is
designed to interconnect and close the gap between product design, product development, production
planning, production, and associated services (Weyer et al., 2016).
Figure 6: Word Cloud DT Definitions of Cluster C1
The real power of a digital twin is that it can provide a near-real-time comprehensive linkage between
the physical and cyber systems. It is the core aspect reflected in the Word cloud in Figure 7, that
11
introduces the Cyber-Physical System (CPS) concept in DT definitions (cluster C2) provided by
(Alam and El Saddik, 2017), (Autiosalo, 2018), (Graessler and Poehler, 2017). A Cyber-physical
system (CPS) (Monostori et al., 2016) aims at embedding computing, communication and controlling
capabilities (3C) into physical assets to converge the physical space with the virtual space. (Alam and
El Saddik, 2017) and (Autiosalo, 2018) agree in defining the DT as the cyber part of a Cyber-Physical
System to emulate the human employee through dynamically adapted values of a database, which
represent for example properties, preferences, work schedule and skillset (Graessler and Poehler,
2017).
CPS and DTs involve the integration between the physical and the cyber space. However, DTs focus
more on virtual models, while CPS emphasize 3C capabilities (computing, communication and
controlling). In this way, the physical space can be integrated, monitored, controlled and coordinated
by the virtual space in real-time and vice versa (Blume et al., 2014). Sensors and actuators are
considered as key elements in CPS (Lee et al., 2015), while models and data are the core components
in a DT (Tao et al., 2019). A digital representation, hence, a digital twin, bridges the gap between the
physical and cyber system improving and supporting the decision making (Zhang et al., 2019).
Some authors (Lee et al., 2013), (Brenner and Hummel, 2017), (Schluse et al., 2017), (Stark et al.,
2017), (Weber et al., 2017) (Yun et al., 2017), (Lee and Kim, 2018), (Haag and Anderl, 2018), (Leng
et al., 2019) involve a comprehensive representation of all data, information, and knowledge of the
physical twin in the Digital Twin definitions. The word cloud in Figure 8 shows the results of
definitions clustered in C3. Most definitions aiming at an increased convergence between real and
virtual space. A closed loop needs to be realized, (Leng et al., 2018) between the physical and virtual
worlds through real time data connection (Brenner and Hummel, 2017), (Schluse et al., 2017), (Stark
et al., 2017) by collecting and analysing data (Weber et al., 2017), (Lee et al., 2013), (Haag and
Anderl, 2018) to respond to the changes over the time and to optimize business performance (Lee and
Kim, 2018). Even though the definitions agree on the virtual space, they differ regarding the physical
space. Some refer to machines, or more generally to products (Lee et al., 2013), (Lee and Kim, 2018),
(Haag and Anderl, 2018), (Leng et al., 2019). Others consider an asset as a physical space (Schluse
et al., 2017), (Stark et al., 2017), (Weber et al., 2017). (Yun et al., 2017) and (Lee and Kim, 2018)
extend the notion of physical space to the whole system or process. In more details the definitions
converge to the descriptions of which type of data should be transferred from the physical to the
12
virtual space. (Schluse et al., 2017) consider not only data generated by the physical product, but also
its functionality (data processing, behaviour) and its communication interfaces. (Haag and Anderl,
2018) go even further by defining the properties, condition, and behaviour of the real-life object. A
Digital Twin is not just defined by the data. It also includes data driven analytical algorithms in (Lee
et al., 2013) and in (Stark et al., 2017) to reflect the status of the corresponding physical part (Yun et
al., 2017). A Digital Twin can integrate data from multiple sources. The interaction with the physical
system should be bidirectional (Leng et al., 2019). Data collected from the physical space updates the
virtual model. The physical twin improves its performance during real time operation exploiting
knowledge acquired from the data.
Figure 8: Word Cloud Definitions of Cluster C3
Regardless of the represented physical space, it needs to be defined which aspects of the physical
space should be transferred to the virtual space. The disunity in literature on how to model the
behaviour of the physical space is even present in the definitions belonging to cluster C4 (Rosen et
al., 2015), (Ciavotta et al., 2017), (Bao et al., 2018), (Nikolakis et al., 2018), (Tao et al., 2018b), (Z.
Liu et al., 2018), (Zhuang et al., 2018), shown in Figure 9. Some try copying the physical behaviour,
its properties, and characteristics in very realistic (Rosen et al., 2015) virtual models to simulate the
behaviour of the current status of the physical space (Bao et al., 2018). The need for a set of virtual
models stems from the fact that the virtual models include not only the geometric models, but also all
rules and behaviours, such as material properties, mechanical analysis, health monitoring (Tao et al.,
2018b) to monitor, diagnose, predict, and control the state and behaviours of the corresponding
physical entities (Tao et al., 2018b), (Nikolakis et al., 2018). Data and information should also
consider all perspectives of the physical space including, structure, semantics, and behaviour to mesh
the virtual and physical worlds (Ciavotta et al., 2017). The DT is typically applied in contexts
characterized by uncertainty and complexity, where the working conditions may vary depending on
external and internal factors. For this reason, (Z. Liu et al., 2018) propose the concept of ‘living
model’ while (Zhuang et al., 2018) the concept of ‘dynamic model’ i.e., a model that continually
adapts and changes in the environment. The Digital Twin should evolve synchronously with the real
system along its whole life cycle. It should be able to modify its initial configuration and to adapt
itself to the current situation. This aspect introduces another feature debated in the literature, namely
13
the difference between the simulation capabilities and the emulation capabilities of a DT. On the one
hand, the simulation capabilities of a DT are provided by a design of its environment allowing to
approximate the behaviour of the real systems to represent how the system reacts (Law et al., 2000).
It can be thought of as a “static feature” of the DT. On the other hand, the emulation refers to the
capability of a DT to be synchronous with the real system, so as it behaves almost similarly to the
actual behaviour of the physical system (Ayani et al., 2018). Accordingly, this feature of DT can be
thought of as a “dynamic feature”. An emulation model operates in a hardware-in-the-loop
configuration to perform the same work of the physical system. It provides a closer replication with
respect to the simulation model (Lee and Park, 2014). From the simulation point of view, the digital
twin represents a new wave in modelling and simulation (Rosen et al., 2015). From the emulation
point of view, the digital twin duplicates and imitates the physical system in the virtual word. It can
thus help to proactively understand what should be done and to react to modifications in the real
world.
The virtual system concept, in Figure 10, sums up the DT definitions clustered in C5 provided by
(Tuegel, 2012), (Negri et al., 2017), (Asimov et al., 2018), (Luo et al., 2018). The virtual system
enables the replication of the physical system into its “digital twin” throughout the entire value chain,
by merging data into behaviour models (Borangiu et al., 2019). The physical twin automatically
transfers data of its behaviour, its status, and information to the virtual space over the entire life cycle.
The virtual system defined also as virtual replica in (Asimov et al., 2018), virtual prototype in (Luo
et al., 2018), and virtual counterpart in (Negri et al., 2017), provides different services such as the
control of the current situation and the prediction of the near future and sends them back to the real
space so the physical product or process adapt accordingly.
The virtual system may enable companies and organisation to solve physical issues faster by detecting
them sooner, predict outcomes, design, and build better products, and ultimately, better serve their
customers (Trauer et al., 2020).
14
The ability of simulation along product life cycle (C1), the synchronization of the cyber system with
the physical assets (C2), the integration of real time data (C3), the behavioural modelling of the
physical space (C4), the services provided by the virtual system (C5) are the main aspects that
characterize the digital twin definition. In view of above, what it is a digital twin can be summarized
as follow:
“A set of adaptive models that emulate the behaviour of a physical system in a virtual system getting
real time data to update itself along its life cycle. The digital twin replicates the physical system to
predict failures and opportunities for changing, to prescribe real time actions for optimizing
and/or mitigating unexpected events observing and evaluating the operating profile system”.
3 Digital Twin Paradigm
Existing scientific research on DT’s, are focussed on different application domains and their related
technologies. A more detailed look is needed to define where, when, why and how develop a digital
twin. The model presented in Figure 3 and explained in section 1 (step 2), has been applied to
automatically discover the main topics in the DT collection and classify them into the discovered
topics. Four different topic modelling have been detected applying the Latent Dirichlet Allocation
(LDA) algorithm. The results can be visualized in table 3 that presents the set of keywords that make
up each topic modelling. On the base of the concepts introduced and analysed also in the state of the
art it is possible to define the semantic of the topics below. The first topic modelling presents the
keywords: ‘support-production, decision-making, simulation, analysis’ indicating that different
studies in literature review have been devoted to the analysis of the digital twin functions. The
analysis of papers classified in the first topic modelling allows to evaluate why the digital twin should
be used. The second topic modelling introduces and considers the enabling components for designing
and developing a Digital Twin as: ‘data-driven and physical model’ allowing to evaluate which are
the key components and technologies employed for implementing digital twins. The words of the
third topic modelling reveal that the value of the digital twin is its ability to understand and to predict
how the product will perform in production, and to optimize its services and performance through its
life cycle. It means that the set of papers belonging to topic 3 treat where the digital twin should be
15
applied and when should be developed. The papers grouped in the fourth topic modelling are focused
on the exploration of the digital twin configuration/architecture. The architecture is the basic
principle to investigate for understanding how to design a digital twin.
Table 3: Data Table (Topic Modelling) provided by the model designed in Orange (Figure 3)
ID TOPIC MODELLING: Set of Words TOPIC LABEL
TOPIC 1 Support-production, decision-making, simulation,

FUNCTIONS
‘Why should a Digital Twin be
(T1) analysis, approach
used?’
COMPONENTS/
TOPIC 2 Big data, data-driven, management, physical model, TECHNOLOGIES
(T2) shopfloor, system ‘How to implement a Digital
Twin?
CONTEXT and LIFE
CYCLE
TOPIC 3 Life cycle, cps, level, product, process, service, improve, ‘Where is appropriate to use a
(T3) application Digital Twin?’
‘When has a Digital Twin to
be developed?’
TOPIC 4 ARCHITECTURE
Architecture, virtual framework, smart manufacturing ‘How to design a Digital
(T4)
Twin?’
A paper typically can cover multiple topics in different proportions (%). LDA algorithm
automatically classifies the DT papers to topics and estimates their relevance to each topic as shown
in Figure 11. For example, (Abramovici et al., 2017) (Paper 1) covers the topics 1 and 4 at 23% and
76% respectively.
16
Figure 11: Topic Modelling Distribution of each Paper
The overview distribution of all topics is provided in Figure 12. It illustrates that 46,7% of papers
covers the DT functions, DT components and DT architectures (T1ΛT2ΛT4) topics while 31,4% of
papers analyse all topics. All the absent combinations are with percentage of zero.
Figure 12: Overview Distribution of Topic Modelling
The topic modelling analysis distribution allows understanding that several studies have been devoted
to analysing the Digital Twin concept and its instantiation in different application contexts (T3). At
the same time, in each context, the digital twins have their own specificity as functions in the life
17
cycle (T3) phases namely design, manufacturing and service. As a result, each application of a DT
varies depending on a different function (T1) accordingly. However, the design of a digital twin
requires the definition of an architecture (T4) and the enabling components/technologies (T2) to
implement it. As a summary of the topic modelling performed above, these can be logically sorted to
shape the digital twin paradigm as follow:
1. Application Contexts (Where is appropriate to use a Digital Twin?), TOPIC MODELLING 3
2. Life cycle (When has a Digital Twin to be developed?), TOPIC MODELLING 3
3. Functions (Why should a Digital Twin be used?), TOPIC MODELLING 1
4. Architecture (How to design a Digital Twin?), TOPIC MODELLING 4
5. Components/Technologies (How to implement a Digital Twin?), TOPIC MODELLING 2
The papers thus classified by LDA were analysed to identify and define which and how many
subtopics each topic consists of. By subtopic we mean the identification of the main categories that
characterise each topic. This allows us to compare and review the existing works to answer the main
questions posed above. Table 4 shows the results. It reports one paper on each row and the topics and
their corresponding sub-topics on columns. The “Context and Application” columns describe the
application context taken into consideration in each article. Accordingly, to the papers belonging to
this topic, the application contexts can be grouped in five categories, that are listed as follows:
Healthcare; Maritime and Shipping; Manufacturing; City Management; Aerospace. The column
under “life cycle” is split into the main product life cycle phases namely: design, production, and
service. The “Functions” column defines the DT functions/purposes specifically: Accelerating the
product development speed; Identifying customers’ needs; Performance optimization and validation;
Remote commissioning and diagnostics. The columns under “Architecture” reports the main layers
used to design a DT namely: Physical; Network; Computing. The analysis also considers the
“Components” i.e., the most discussed and applied technologies for building a Digital Twin. The
same table was used for the analysis conducted by Formal Concept Analysis (FCA) to explore the
trends and the combinations in literature on the design and the development of a digital twin. The
results are discussed in the next sections presenting respectively the DT application contexts in
section 3.1, the DT life cycle and its functions in 3.2, and the DT architecture and its
components/technologies in 3.3.
18
Table 4: FCA Table Structure: Objects and Attributes
3.1 Digital Twin Application Contexts (‘Where is appropriate to use a Digital

Twin?’ and ‘Who is doing Digital Twins?’)
The main papers covering the topic modelling 3 are discussed in this subsection to define and discuss
the main DT application contexts. The review reveals that the most explored contexts are: Healthcare;
Maritime and Shipping; Manufacturing; City Management; Aerospace. Digital Twins in Healthcare
is used for capturing and visualizing a hospital system to create a safe environment and to test the
impact of potential changes on system performances. A digital twin can be used to predict the
19
outcome of specific procedures. It can determine the better therapy option for a specific patient. In
healthcare, a digital twin recording data of a person, combined with AI models, can provide answers
for clinical problems (Bruynseels et al., 2018). Digital Twins in Maritime and Shipping are used as
support for design. The design requires to invest significant amounts of time and money in preparing
analytical models to perform simulations. The digital twin allows to visualize all key components, to
perform analyses and calculations, and to improve the control of the effects of operation on the ship’s
structural and functional components (Arrichiello and Gualeni, 2019).
A Digital Twin in Manufacturing involves different applications based on the stages across the entire
lifetime of a product, such as design, production, logistics and maintenance (Dassisti et al., 2017),
(Greif et al., 2020). The digital twin can support decision makers to predict an upcoming equipment
failure, to inform an operator when an asset begins to show signs of non-optimal performance, to
improve customer experience (Tao et al., 2018a).
Cities are areas of human settlement, with high population density, complete infrastructure, and
buildings. Digital Twins, in City Management, improve the urban environment and people’s quality
of life. The digital twin can simulate people movements and emergency evacuations, modelling smart
buildings, road traffic, air quality, infrastructure, and circular urban economies. The benefits of
modelling range from preventive maintenance to operational efficiencies and cost savings. The DTs
improve services for citizens, and increase safety and security (Mohammadi and Taylor, 2017).
Aerospace companies have begun utilizing digital twins to accomplish the goal of reducing unplanned
downtime for engines and other systems. Digital Twins in Aerospace may allow receiving advance
warnings and predictions, but also preparing a plan of actions based on simulated scenarios that
consider the weather conditions, the performance of the asset, and several other variables (Tuegel et
al., 2011). With the help of digital twins, it is possible to develop and implement predictive
maintenance for increasing the platform’s operational availability and efficiency, extending its life
cycle and reducing its cost. Moreover, DTs are capable of mitigating damage or degradation by
activating self-healing mechanisms or by recommending changes in the mission profile to decrease
loadings (Mandolla et al., 2019).
Digital Twins have attracted strong interests from industries too: GE Predix Platform, SIEMENS
PLM, Microsoft Azure, IBM Watson, PTC Thing Worx, Aveva, SAP Leonardo Platform, Twin
Thread, DNV-GL, Dassault 3D Experience, Sight Machine, Oracle Cloud. Patents have been filed by
(Hershey et al., 2017) for General Electric and by (Song and Canedo, 2016) and (Fischer and Heintel,
2017) for Siemens. The Figure 13 shows the main DT platform for each application context. The
roles of digital twins along life cycle management and its functions are discussed in the next section.
20
Figure 13: Industry DT Platform in each type of DT Application
3.2 Digital Twin Life cycle (‘When has a Digital Twin to be developed?) and
its Functions (‘Why should a Digital Twin be used?’)
In principle, out of the literature review and the DT definitions analysis presented in section 2, the
digital twin finds application in the entire product life cycle management (PLM) that can be divided
into three phases: Design; Production; Service (Bao et al., 2018), (Tao et al., 2018a). Regardless of
the context domain, the DT has a series of functions in each phase of life cycle (Barricelli et al., 2019)
that can be summarized in: Accelerating the product development speed; Identifying customers’
needs; Performance optimization and validation; Remote commissioning and diagnostics. The
existing trends and associations between DT life cycle and its functions was carried out by using
Formal Concept Analysis (FCA). The lattice, illustrated in Figure 14, represents the hierarchy of
concepts that group the papers according to their common life cycle phase and/or functions.
21
Figure 14: Lattice of Life Cycle phases and Functions
From Table 5, we can deduct that the FCA graph detects 16 different concepts. As explained in
section 1 (step 3), a concept is constituted by two parts: its extension which consists of all objects
belonging to the concept, and its intention which comprises all attributes shared by those objects.
Table 5: Formal Concepts of Life Cycle phases and Functions
FORMAL CONCEPTS
ID CONCEPT <{CONCEPT EXTENTS},{CONCEPT INTENTS}>
C1 <{115 Papers}>
C2 <{34 Papers}, {Digital Twin in the design phase}>
C3 <{23 Papers}, {Identify customers’ needs}>
C4 <{41 Papers}, {Performance optimization and validation}>
C5 <{54 Papers}, {Digital Twin in the production phase}>
C6 <{22 Papers}, { Digital Twin in the design phase, Identify customers’ needs}>
C7 <{31 Papers}, {Digital Twin in the service phase}>
C8 <{40 Papers}, {Digital Twin in the production phase, Performance optimization and
validation}>
C9 <{29 Papers}, {Digital Twin in the design phase, Digital Twin in the production
phase}>
C10 <{19 Papers}, {Digital Twin in the production phase, Identify customers’ needs}>
C11 <{21 Papers}, {Digital Twin in the design phase, Identify customer’s needs, Accelerate
the product development speed}>
C12 <{20 Papers}, {Digital Twin in the service phase, Remote commissioning and
diagnostics}>
C13 <{18 Papers}, {Digital Twin in the service phase, Performance optimization and
validation}>
22
C14 <{27 Papers}, {Digital Twin in the design phase, Digital Twin in the production phase,
Digital Twin in the service phase}>
Digital Twin in the service phase, Identify customer’s needs, Performance optimization
and validation}>
Digital Twin in the service phase, Identify customer’s needs, Accelerate the product
development speed, Performance optimization and validation, Remote commissioning
and diagnostics}>
The concept C2 shows, through its Extent column, the existence, in our literature review, of {34
papers} which analyse the application of the <{Digital Twin in the design phase}>. The digital twin
in the design phase can be applied to the conceptual design, detailed design, and virtual verification
(Tao et al., 2018a) of a product. The digital twin in design stage is designed to generate the digital
product design before the real execution (Q. Liu et al., 2018), (H. Zhang et al., 2017). In the
conceptual design stage, the digital twin serves to guide designers to formulate functional
requirements (Tao et al., 2018b). It can make the communication between customers and designers
more transparent and faster by using the real-time transmission data (Tao et al., 2018a). In the detailed
design phase, the digital twin enables simulation tests to ensure that the prototype can achieve the
desired performance (Wärmefjord et al., 2017). In the virtual verification phase, the digital twin
enables to simulate and predict the performance of the physical products based on virtual models
(Damiani et al., 2018), (Bohlin et al., 2017). The concept C6 demonstrates that {22 Papers} in our
selection, analyse the relation between the <{Digital Twin in the design phase} with the function
{Identify customers’ needs}>. Performances, customer usages and preferences are reflected in the
twin, and then feed into the product development process to increase the customer satisfaction and
market share (Tao et al., 2018b), (Macchi et al., 2018). The studies {21 Papers} grouped in C11
discuss the application of the < {Digital Twin in the design phase} for two different functions
{Identify customer’s needs, Accelerate the product development speed}>. Digital twins in the design
phase can guide the designers to iteratively adjust the customers’ expectations and improve the design
models, achieving personalized design (Tao et al., 2018a). The digital twin can be used for designing
products, testing them in real time situations, stipulating how the customer or the end user will use
them and how the design will complement the product’s environment (Söderberg et al., 2017). Data
from the real machine are loaded into the digital model to enable simulation and testing of ideas even
before actual manufacturing starts. The digital twin can be used to plan, reconfigure the product in
response to external changes.
The concept C5 shows the existence of {54 papers} which analyse the application of the <{Digital
Twin in the production phase}>. For example, (Leng et al., 2018) presents a Digital Twin for
manufacturing cyber-physical systems (MCPS). (Ding et al., 2019) introduces a DT-based Cyber-
Physical Production System (DT-CPPS). MCPS is used for controlling the shop floor manufacturing
while DT-CPPS for improving the flexibility, controllability, and efficiency of shop floor
manufacturing. A digital twin for production control and optimization can analyse the online data
collected from the physical line for searching the optimal solution to the physical line (Sun et al.,
2017) or to complex product assembly shopfloors (Zhuang et al., 2018). It can evaluate autonomously
23
the production real-time (Vachálek et al., 2017) and optimize the resource allocation (H. Zhang et al.,
2018) autonomously (Rosen et al., 2015). A Digital Twin reference model for rotating machinery
fault diagnosis was developed in (Wang et al., 2018), defining the requirements for constructing the
Digital Twin model. A digital twin for hydraulic supports (Xie et al., 2019) is built to simulate the
actual hydraulic and to support diagnosis and degradation analysis. The digital twin finds application
also in CNC machine tool (Luo et al., 2018) and in smart injection process (Liau et al., 2018) to
control the behaviours of the physical system in real-time. The papers {40 Papers} in C8 treat the
application of <{Digital Twin in the production phase} for {Performance optimization and
validation}>. Digital twins in the production phase aim at real time monitoring and optimization and
for predicting the future state of the physical twin, thus preventing downtime and failures (Lee et al.,
2013). The digital twin helps at determining the optimal set of parameters and actions that can help
maximizing some of the key performance, and providing forecasts for long-term planning (Vachálek
et al., 2017). The digital twin can analyse performance data collected over time and under different
conditions (Alcácer and Cruz-Machado, 2019), reducing unplanned machine downtime, the amount
of ‘scrap’ produced in each production line, and minimizing costly production quality faults. The DT
can optimize and elevate the production process to a higher level of effectiveness and flexibility
(Cimino et al., 2019).
The concept C7 indicates that {31 papers} are oriented towards the <{Digital Twin in the service
phase}>. The service phase refers to the phases after sale, including the product utilization and the
maintenance (Tao et al., 2018a). (Abramovici et al., 2017) introduce a cloud-based Smart Product
platform for the reconfiguration of Smart Products during the use phase using the concept of virtual
product twins and an Internet of Things. The conceptual approach is prototypically demonstrated by
considering a model environment for smart cars, which are temporarily reconfigured during their use
phase. The digital twin has been developed also for the waste electrical and electronic equipment
recovery to support the manufacturing/remanufacturing operations (Wang and Wang, 2018). The
<{Digital Twin in the service phase} supports the {Remote commissioning and diagnostics}> of the
operations of interconnected systems such as manufacturing systems, as presented in the studies {20
Papers} grouped in concept C12. This allows virtual monitoring systems and validation of the current
status of production systems (i.e., energy monitoring and fault monitoring) (Qi et al., 2018). In
addition, {18 Papers} analyse the application of the {Digital Twin in the service phase} for
<{Performance optimization and validation}>. The digital twin can upgrade personalized product
functions (Cheng et al., 2020) by obtaining the user's usage. In fact, in the service phase, Digital
Twins can provide value-added services support for the prognostics and health management (PHM)
(Qi et al., 2018) (Wang et al., 2018). The PHM is an engineering process of failure prevention and
predicting reliability and remaining useful lifetime (RUL) (Sutharssan et al., 2015). In this case, the
digital twin (DT) is developed for improving the accuracy and efficiency in the life cycle monitoring
of a product (Tao et al., 2018c), (M. Zhang et al., 2018). There are currently relatively few digital
twin applications {17 papers} for supporting the entire product life cycle (C16).
24
3.3 Digital Twin Architecture and Components (‘How to design and
implement a Digital Twin?’)
A general and standard architecture of a digital twin was first built by (Grieves, 2014) that presents a
physical space, a virtual space and the connection between them. There are various understandings
of the DT architectures among researchers (Dassisti et al., 2017). (Stark et al., 2017) characterizes
the DT as (1) an unique instance of the universal Digital Master model of an asset, (2) its individual
Digital Shadow and (3) an intelligent linkage (algorithm, simulation model, correlation, etc.) of the
two elements above (Kritzinger et al., 2018). An extended five-layer DT is proposed by (Tao et al.,
2018c) and it is composed by: (1) Physical entity (PE); (2) Virtual entity (VE); (3) Services (Ss) for
PE and VE; (4) Data (DD); (5) Connection (CN) among PE, VE, Ss and DD. Compared to Grieves’s
architecture, data and services layers were added. The five-layer DT architecture developed by
(Ponomarev et al., 2017) presents: (1) cyber-physical layer; (2) primary processing/store data layer;
(3) distributed computing and storage layer; (4) models and algorithms layer; (5) visualisation and
user interfaces layer. This kind of architecture highlights the data storage, the distributed computing
and management system as critical parts of the digital twin. An extended six-layer DT is presented
by (Redelinghuys et al., 2019). The layers are: (1) physical devices; (2) local controllers; (3) local
data repositories; (4) IoT gateway; (5) cloud-Based information repositories; (6) emulation and
simulation. This structure is more focused on the transmission of data flow from the physical system
(Layer 1 e 2) to the cloud (Layer 5). From the computational perspective, the key functionality of a
digital twin is the combination of physics-based models and data driven models to emulate and
simulate the physical space accurately (Kaur et al., 2020).
In view of above, a DT architecture can be thought as consisting of several components and
technologies organised into three main layers: the physical layer; the network layer; the computing
layer (Boje et al., 2020). The physical layer consists of physical entities identified based on the stage
of the product life cycle. The network layer connects the physical domain to the virtual one. It shares
data and information. The computing layer involves the virtual entities emulating the corresponding
real entities, including data-driven models and analytics, physic-based models, services, and users.
Each layer is characterised by some DT components (for example hardware or software technologies,
models, information structures) with commonalities in their scope of use and interactions, having also
complementary functionalities. FCA was run on to detect which are the main and the most studied
components/technologies for each layer. The formal concepts of the physical layer are shown in
Figure 15 and in Table 6.
25
Figure 15: Lattice of Physical layer and its Components/Technologies
Table 6: Formal Concepts of Physical layer and its Components/Technologies
FORMAL CONCEPTS
C1 <{115 Papers}>
C2 <{38 Papers}, {Physical Layer}>
C3 <{12 Papers}, {Physical Layer, Unit Level}>
C4 <{9 Papers}, {Physical Layer, System level}>
C5 <{6 Papers}, {Physical Layer, System-of-systems level}>
C6 <{6 Papers}, {Physical Layer, Unit Level, System level }>
C7 <{6 Papers}, {Physical Layer, RFID sensor networks}>
C8 <{12 Papers}, {Physical Layer, Wireless sensor networks (WSN), RFID}>
C9 <{5 Papers}, {Physical Layer, Unit Level, System level, System-of-systems level}>
C10 <{5 Papers}, {Physical Layer, Unit Level, RFID sensor networks}>
C11 <{5 Papers}, {Physical Layer, RFID sensor networks, Wireless sensor networks
(WSN)}>
C12 <{4 Papers}, { Physical Layer, Unit Level, System level, System-of-systems level,
RFID sensor networks}>
C13 <{4 Papers}, {Physical Layer, RFID sensor networks, Unit level, Wireless sensor
networks (WSN)}>
C14 <{4 Papers}, {Physical Layer, System level, Wireless sensor networks (WSN)}>
C15 <{4 Papers}, {Physical Layer, Wireless sensor networks (WSN), RFID}>
26
C16 <{3 Papers}, {Physical Layer, Unit Level, System level, System-of-systems level,
Wireless sensor networks (WSN), RFID sensor networks}>
C17 <{3 Papers}, {Physical Layer, Unit Level, Wireless sensor networks (WSN), RFID}>
C18 <{3 Papers}, {Physical Layer, System Level, Wireless sensor networks (WSN), RFID
}>
C19 <{2 Papers}, {Physical Layer, Unit Level, System level, System-of-systems level,
RFID sensor networks, RFID, Wireless sensor networks (WSN)}>
The concepts C2, C3, C4, C5 show that the literature differs regarding the definition of the physical
space, as highlighted also by the DT analysis definitions. In literature review, the digital twin has
been applied at different physical levels (Tao et al., 2019) that can be summarized in: unit level (C3),
system level (C4), system-of-systems level (C5). The unit level is a minimum but independent
individual, which cannot be further divided such as a single piece of equipment (e.g., a product, a
machine tool or robot arm). It contains a basic closed loop of data between the physical and virtual
spaces with the abilities of sensing and computing. The system level can be a production system such
as a production line, a shop floor, or a factory. It is characterized by self-organization, self-
configuration, and self-optimization. While the system-of-systems level is characterized by
enterprises’ collaborations. The application of the digital twin at the system-of systems level can
achieve the horizontal integration. It refers to the exchange of information across the supply chain
such as resources management system, logistics, marketing, or intercompany value chains (Posada et
al., 2015). For each type of level, Digital Twins can get and share data between all production factors
and information systems achieving the vertical integration i.e., the integration of various Information
Technology systems at different hierarchical levels. In manufacturing contexts, the literature concurs
that the data type and consequently the data sources depend on the selected physical levels. Typically,
a system-of-systems digital twin involve and exploit different data sources such as Internet/Users
Data from CRM, E-commerce platforms (e.g., Amazon) and social networking platforms (e.g.,
Twitter, Facebook, LinkedIn, and YouTube), to understand user preference, and behaviours (Qi and
Tao, 2018). It also involves Product data from computer-aided systems like CAD/CAM, CAE;
Management data from manufacturing information systems such as MES, PDM, SCM, ERP, etc (Luo
et al., 2018); Operational data from manufacturing equipment such as product data, quality data,
maintenance data (Dassisti et al., 2019b); Environmental data which affects the physical equipment
operations, such as environmental pressure, ambient temperature, and moisture level (Cai et al.,
2017). The papers grouped in C6-C19 treat the equipping of a physical system with sensors,
actuators, and embedded communication for recording real-time states (e.g., vibration, force, torque,
and speed) and working conditions (e.g., environment parameters, loads, and control orders) (Ruppert
et al., 2018) of the physical space. The most discussed technologies for the physical layer are RFID,
RFID Sensor networks and Wireless sensor networks (WSN). RFID allows automatic identification
and data capture using radio waves, a tag, and a reader (Lee and Lee, 2015). RFID Sensor networks,
consisting of a very large number of nodes for monitoring and recording the physical conditions of
the environment (Atzori et al., 2010). Wireless sensor networks (WSN) which consist of spatially
distributed autonomous sensor-equipped devices to monitor physical or environmental conditions
(Gubbi et al., 2013), (Tan and Wang, 2010). The components belonging to physical layer carry out
27
real-time data for the synchronization of the virtual twin with its corresponding physical twin with
the capabilities of anomaly detection, prediction, prescription, and optimization.
The network layer involves connections and interactions amongst physical elements and virtual space.
This layer connects all components together for sharing data and information with other connected
components (Da Xu et al., 2014). The key technologies discussed in the literature review are
middleware, communication protocol analysis, communication protocol/interface conversion,
wireless communication, and Application Programming Interfaces (API). FCA detects 11 possible
concepts, as illustrated in Figure 16 and in Table 7.
Figure 16: Lattice of Networking Layer and its Components/Technologies
Table 7: Formal Concepts of Networking Layer and its Components/Technologies
FORMAL CONCEPTS
C1 <{115 Papers}>
C2 <{41 Papers}, {Network Layer}>
C3 <{13 Papers}, {Network Layer, Communication protocol/interface: Automation
ML}>
C4 <{25 Papers}, {Network Layer, Communication protocols}>
C5 <{8 Papers}, {Network Layer, Communication protocol/interface: Automation ML,
Programming interface (API)}>
Middleware}>
C7 <{11 Papers}, {Network Layer, Communication protocol/interface: Automation ML
Communication protocol}>
Programming interface (API), Middleware}>
Programming interface (API), Communication protocol, Wireless Communication}>
28
Middleware}>
Middleware, Programming interface (API), Communication protocol, Wireless
Communication}>
C4 concept shows, through its Extent column, the existence, in our literature review, of {25 papers}
which analyse the <{Communication protocols} for the {Network Layer}>. The communication
protocol allows two or more entities in the DT to transmit information to each other. OPC Unified
Architecture (OPC UA) and MT-Connect protocols are the protocols more employed in digital twin
applications to access to data and to transmit them in real-time (Redelinghuys et al., 2019). The
studies {13 papers} in concept C3 discuss the <{Communication protocol/interface: Automation
ML} in the {Network Layer}>. The communication protocol/interface conversion transforms various
communication protocols/ interfaces into a unit form. AutomationML is an open standard for a data
format based on XML allowing the exchange of plant engineering information (Bao et al., 2018),
(Drath et al., 2008). The AutomationML is used in digital twin to model attributes related to the digital
twin. The goal is to interconnect the heterogeneous toolchain of digital manufacturing (Um et al.,
2017). It is used to exchange data between the Digital Twin and other systems and a methodology for
communication and exchange of data (G. N. Schroeder et al., 2016), (Talkhestani et al., 2018). The
studies belonging to concepts C5-C11 help to deepen the following technologies <{Middleware,
Wireless communication, Application Programming Interfaces (API)} in the {Network Layer}>. The
middleware is a software layer interposed between the technological and the application levels. The
middleware architecture more used in the digital twin is the Service Oriented Architecture (SOA)
approach. The adoption of the SOA principles allows for decomposing complex and monolithic
systems into applications consisting of an ecosystem of simpler and well-defined components (Gubbi
et al. 2013b). The wireless communication can connect entities in the DT wirelessly, thus improving
flexibility in data transmission. The application Programming Interfaces (API) realize the
communication between different software systems and models in the virtual space that represents
the computing layer.
The computing layer is fundamental for computing and decisional support of digital twins. FCA
detects 90 possible concepts as shown in Figure 16. A set of concepts are reported in Table 8. C2
concept demonstrates that the <{Computing Layer}> is the most addressed in our literature review
{83 papers} either for the potential innovation or for the strong impact on decision support.
29
Figure 17: Lattice of Computing Layer and its Components/Technologies
Table 8: Formal Concepts of Computing Layer and its Components/Technologies
FORMAL CONCEPTS
C1 <{115 Papers}>
C2 <{83 Papers}, {Computing Layer}>
C3 <{31 Papers}, {Computing Layer, Data-driven Methods}>
C4 <{23 Papers}, {Computing Layer, Decision-making/Rule model}>
C5 <{24 Papers}, {Computing Layer, Physical model}>
C6 <{27 Papers}, {Computing Layer, Geometric model}>
C7 <{19 Papers}, {Computing Layer, Collaborative information model}>
C8 <{28 Papers}, {Computing Layer, Behaviour model}>
C9 <{10 Papers}, {Computing Layer, Decision-making/Rule model, Physical model,
Geometric model, Collaborative information model, Behaviour model}>
C10 <{10 Papers}, {Computing Layer, Modularity}>
C11 <{11 Papers}, {Computing Layer, Interoperability}>
C12 <{3 Paper}, {Computing Layer, Dynamicity}>
C13 <{5 Paper}, {Computing Layer, Fidelity}>
C14 <{4 Paper}, {Computing Layer, Scalability}>
The computing layer can be perceived as a set of “layers” interconnected, which includes the
following components: data (C3), models (C4-C9), and modelling features (C10-C14).
The data layer includes all different types of data previously defined in the physical layer (Uhlemann
et al., 2017). This sub-layer has characteristics of heterogeneity of data and data sources, volume, and
speediness. Data preparation and data analysis are the key aspects discussed in literature. The data
30
preparation process includes data selection, data cleaning, data modelling, data integration, and data
transformation. The data analysis includes all data-driven models such as machine learning data
mining, pattern evaluation, and knowledge representation involved in DT building. The studies {31
papers} grouped in C3 examine the <{Data-driven Methods} in {Computing Layer}>. Data-driven
models are designed to extract knowledge from data (Y. Zhang et al., 2017), (Lee et al., 2014b). The
digital twin aims to integrate data across different domains into virtual models (Kusiak, 2018). The
main data-driven models used in digital twins are machine learning, neural networks, and deep
learning. The machine learning refers to the ability to give computers the possibility to learn without
being explicitly programmed (Clarke et al., 2009). It is classified in supervised, unsupervised
(Sutharssan et al., 2015) and reinforcement learning (van Otterlo and Wiering, 2012) (Jain et al.,
2017). Machine learning techniques used in digital twins are supervised and unsupervised learning
(Jain et al., 2017). The supervised learning develops models based on input and output data (Tidriri
et al., 2016). The supervised learning, across the digital twin, is applied for the system’s failures
prediction (Asimov et al., 2018), or for prediction of the remaining useful life (RUL) of the physical
twin (Z. Liu et al., 2018). The unsupervised learning, instead, discovers an internal representation
from input data only (Sutharssan et al., 2015), (Fahad et al., 2014). It enables discovering similar
groups within data, based on clustering techniques (Xu and Wunsch, 2005), (Grira et al., 2004). In
digital twins, these techniques are used for creating autonomous clusters for different working
regimes to analyse machine conditions (Lee et al., 2014a), (Banerjee et al., 2017) (Ding et al., 2019).
Artificial Neural Networks (ANN) and Deep Learning (DL) are computing systems that are inspired
by the human brain (Zhang, 2000). The main scopes in using neural networks and deep learning in
digital twin (Lee et al., 2013) are health assessment, performance prediction (Jain and Bhatnagar,
2020), fault diagnosis (Xu et al., 2019).
The digital twin requires the building and the applying digital models representing the set of resources
and processes knowledge. Different tools and technologies are available for developing high-fidelity
virtual models (Schleich et al., 2017). The most discussed components regard the model types
(concepts C4-C9) and the modelling features (concepts C10-C14). The model types define the
physics-based models and the functions of each model necessary to emulate the physical system.
Physics-based models compare simulated results with known information, represented by
mathematical models (Tidriri et al., 2016). A model represents a system in terms of logical and
quantitative relationships that are then manipulated and changed to see how the model reacts, and
thus how the system would react-if the mathematical model is a valid one (Law et al., 2000). The
physics-based models are based on a set of different models to represent the structure, the behaviour,
and the interactions of a physical system (Tidriri et al., 2016), (Tao et al., 2018c). The most studied
models for developing a digital twin (Semeraro et al., 2019a) are summarised as follow: Geometric
model {27 Papers} in concept C6; Physical model in {24 Papers}in concept C5; Behaviour model
{28 Papers} in C8; Collaborative information model {19 Papers} in C7; Decision-making model {23
Papers} in C4. A geometric model reflects the geometry, the kinematics, the logic, and the interfaces
of the real system (Ayani et al., 2018), (Xie et al., 2019). A geometric model defining shapes, sizes,
positions and assembling machine components is presented in (Tao et al., 2018c). A physical model
enables to simulate the physical properties and loads (Post et al., 2009) analysing the phenomena,
such as deformation, cracking and corrosion (Tao et al., 2018c). A behaviour model describes the
31
way the physical system is governed by driving factors (e.g., control orders) or disturbing factors
(e.g., human interferences) (Tao et al., 2018c), (Bao et al., 2018). A collaborative information model
(Bao et al., 2018) defines how different components interact and simulates the collaborative
behaviour among several assets. A decision-making model makes the model capable of evaluating,
reasoning, and validating. It consists of variable input, algorithms and a collection of constraints and
rules (Bao et al., 2018). It includes rules of constraints, associations, and deductions (Tao et al.,
2018c) and it stores and analyses the running status data, then it makes decisions using machine
learning algorithms.
The common features studied for modelling a digital twin concern: scalability, interoperability,
fidelity, dynamicity, and modularity. According to the studies {4 Papers} grouped in C14, the
scalability is the ability to provide an insight at different scales (from fine details to large systems)
(Schleich et al., 2017) (Putnik et al., 2013). The studies {11 Papers} in C11 define the interoperability
as the ability to convert, to combine, and to establish equivalence between different model
representations (H. Zhang et al., 2017). The model interoperability is a critical aspect for the exchange
of dynamic models and for Co-Simulation. Functional Mock-Up Interface (FMU) standard is
commonly used in digital twins to solve this problem (Negri et al., 2019) (Schluse et al., 2018). FMU
is an open standard for exchanging dynamical simulation models between different tools in a
standardized format and for co-simulation (Blochwitz et al., 2011). FMI standard specifies two
different kinds of FMUs: (1) Model Exchange (ME) – ME FMUs; (2) Co-Simulation (CS) – CS
FMUs. The model fidelity (concept C13) describes the closeness to the physical product (Schleich et
al., 2017) while the model dynamicity (concept C12) is the ability to reflect real time the physical
process and modify autonomously itself if the physical system changes. This crucial issue concerns
the convergence of the physical world with its digital counterpart (Weyer et al., 2016). According to
the {10 Papers} clustered in C10, the modularity is the ability to integrate, to add, or to replace models
(Guo et al., 2018). Two modular approaches have been developed in (Guo et al., 2018), (Semeraro et
al., 2019a). The idea behind this approach is to use and especially re-use predefined functional units
(Semeraro et al., 2019a), that are systematically developed and logically interlinked for the
configuration of a holistic manufacturing system (Stark et al., 2017), (Negri et al., 2019). Virtual
(VR) or augmented (AR) reality technologies can be integrated in digital twins to create interactive
and immersive environments (G. Schroeder et al., 2016) enabling direct interactions between the
digital twin and final users.
As a summary of the analysis discussed above, the digital twin paradigm is summarised in Figure 18
to depict the contexts, the phases of the life cycle (design, production, and service), the functions of
the Digital Twin for each life cycle phase, the architecture layers, and the main components of each
architecture layer.
32
Figure 18: The Digital Twin Paradigm (Semeraro, 2020)
4 Conclusion and Research Challenges
The present study forms a literature review that led to a digital twin paradigm aiming at assessing
which are the application contexts, the life cycle phases, the functionalities, the architectures, and the
components of existing digital twins. The paper aims at providing a detailed picture of the main
features of existing scientific research on DT’s, stressing on the different application domains and the
related technologies. The idea of Digital Twin as a “virtual” image of the reality constantly
synchronized with the real operating scenario is accurately presented and described in section 1 in all
its physical and logical aspects. This literature review tries to answer different research questions at
different level namely: DT definition, application contexts, life cycle phases, functions, architectures,
and components. Section 2 investigates on different DT definition provided in literature to address
the research question: ‘What is a Digital Twin?’. The application contexts in section 3.1 and life cycle
phases in section 3.2 focus on establishing the baseline of the Digital Twin paradigm by trying to
reply to the research questions: ‘Where is appropriate to use a Digital Twin?’, ‘Who is doing Digital
Twins?’, ‘When has a Digital Twin to be developed?’. A digital twin may enable companies and
organisation to predict outcomes, design, and build better products, and better serve their customers
(Madni et al., 2019). To that point, the third research question (‘Why should a Digital Twin be used?’),
in section 3.2, analyses the main functions of a digital twin for each life cycle phase. The fourth
33
research question investigated in section 3.3 (‘How to design and implement a Digital Twin?’) tries
to define the DT architecture and analyse the employed components/technologies for implementing
digital twins. The review results are summarised in the summary table below.
Table 9: Summary table of the review results (5W1H)
Research Question Results

“A set of adaptive models that emulate the behaviour of a physical
system in a virtual system getting real time data to update itself along
‘What is a Digital Twin?’ its life cycle. The digital twin replicates the physical system to predict
DIGITAL TWIN DEFINITION failures and opportunities for changing, to prescribe real time
actions for optimizing and/or mitigating unexpected events
observing and evaluating the operating profile system”.
1. Healthcare
• Improving operational efficiency of healthcare operations
2. Maritime and Shipping
‘Where is appropriate to use a • Design customization
Digital Twin?’ 3. Manufacturing
DIGITAL TWIN CONTEXTS and • Product development and predictive manufacturing
USE CASES 4. City Management
• Modelling and simulation of smart cities
5. Aerospace
• Predictive analytics to foresee future aircraft problems
GE Predix; SIEMENS PLM; Microsoft Azure; IBM Watson; PTC
‘Who is doing Digital Twins?’ Thing Worx; Aveva; Twin Thread; DNV-GL; Dassault 3D
DIGITAL TWIN PLATFORMS
Experience; Sight Machine; Oracle Cloud.
1. In design phase
‘When has a Digital Twin to be • The digital twin is used to help designers to configure and
developed?’ validate more quickly the product development accurately
DIGITAL TWIN LIFE CYCLE interpreting the market demands and the customer
preferences.
2. In production phase
• The digital twin shows a great potential in real-time process
‘Why should a Digital Twin be control and optimization, as well as accurate prediction.
used?’ 3. In service phase
DIGITAL TWIN FUNCTIONS
• The digital twin can monitor the health of a product, perform
diagnosis and prognosis.
‘How to design and implement a 1. The Physical layer involves various subsystems and sensory
Digital Twin?’ devices that collect data and working parameters.
2. The Network layer connects the physical to the virtual, sharing
DIGITAL TWIN
of data and information.
ARCHITECTURE AND
3. The Computing layer consists of virtual models emulating the
COMPONENTS corresponding physical entities.
34
The analysis of the digital twin definitions, features, tools, and methodologies was done based on the
text mining techniques and Formal Concept Analysis (FCA). The application contexts, life cycle
phases, functionalities, architectures, and components are discussed and organised in a unique
paradigm summarised in Figure 18. Balanced against the many advantages that the digital twin can
bring, there are several challenges to be overcome. For the grace of the results summarised in table
9, we try to define what are the main research challenges of implementing a Digital Twin.
Research Challenge #1.

DT APPLICATION CONTEXTS
According to section 3.1, the existing digital twin applications refer to specific and traditional
contexts: Healthcare; Maritime and Shipping; Manufacturing; City Management; Aerospace.
Humans play an important role in the Digital Twin applications especially in manufacturing contexts.
Human interaction is the one key challenge in the development and implementation of DT in the
manufacturing application. While some low-level operations can be autonomously achieved without
human intervention, many decision-making activities are still sustained by many manual operations
based upon human interactions. The research challenge is to form and design cognitive digital twin
able to interoperate with other digital twins and humans in a seamless way whatever happens during
their interactions. Furthermore, the digital twins could virtually help look beyond the current
industrial model of extracting, producing, consuming, and disposing. The digital twin can support
companies to move from a linear system to a 'circular economy' system that considers almost zero
waste production and pollution, keeps products and materials in the recycling loop longer and helps
regenerate natural systems. While scientific literature has analysed the adoption of DT in the
optimization of products life cycle, few contributions have yet focused on the exploitation of DT to
assess and improve the sustainability performances of whole value chains (Barni et al. 2018).
Significant research efforts need to be made on the application of a digital twin for improving the
sustainability performances in each application context.

DT LIFE CYCLE AND FUNCTIONS
The DT presents an exciting possibility of real-time simulation for product life cycle that can be
divided into three phases: design, production, and service. The digital twin can be developed in each
life cycle phase fulfilling different functions. The digital twin can potentially help to integrate even
the entire supply chain, throughout all phases of product life cycle. FCA results in section 3.2
demonstrates that most digital twin applications refer to a single phase of a product life cycle. There
are currently relatively few applications of digital twin for supporting the entire supply chain and the
network enterprises. Digital Twin applications are mainly developed for prediction purposes and used
for decision-making support. The digital twin could connect products, persons, machines, and
enterprises within the virtual space. However, this aspect holds its own challenges at the present state.
The ability to collect, aggregate and exchange data and information between different suppliers,
manufacturers and customers could present interoperability issues. As the DT can potentially
integrate data from the lifetime of a product seamlessly, several key research challenges concern the
definition of standards and communication protocols to ensure interoperability of multiple digital
35
twins with each other (Platenius-Mohr et al. 2020). The development of standards-based
interoperability is important and challenging at the same time for the evolution of digital twin
applications along the entire life cycle.

DT ARCHITECTURE AND COMPONENTS/TECHNOLOGIES
The DT consists of a set of models with complex structures and behaviour, that reflect the real-time
operations of the physical system. Modelling a digital copy of a physical system to perform real-time
validation and optimization is quite complex and thus needs a big amount of data. A digital twin can
be a model of a component, a system of components or a system of systems. The digital twin requires
the construction and application of accurate models of reality. Modelling the reality in a digital twin
is a complex task, as it involves sensors, multifunctional models, multisource data, services, etc. At
present, it is difficult to build an accurate model for a DT using traditional approaches due to the
complexity of real systems. The lack of an univocal reference architecture leads at developing Digital
Twin solutions using different technologies, interfaces, and communication protocols, models and
data as assessed in section 3.3. Standard Digital Twin solutions should be developed to provide design
criteria and design constraints where reference architectural aspects, reference information model and
communication protocols are clearly defined (Lu et al. 2020). The modularization design principle
needs to be explored to improve the modelling efficiency. This would enable to improve the flexibility
and reusability of standard DT-solutions towards different applications. Modular approaches need to
be explored for the construction of flexible DT solutions, facilitating new DT applications.
Next steps of the present research will be related to the definition of a new approach for building a
digital twin by exploring the modularity feature, which is still one of the most challenging research
issues. The idea that we will explore is the definition of a modelling approach that allows to derive a
criterion to self-detect modelling constructs that can be used (and re-used) to create digital models of
different systems or processes.
Authors’ contributions:
Author CS was responsible for the study conduction and assimilating the literature to select the final
sample, also defining the methodological approach; Authors ML, HP, MD assessed the quality of the
included studies; CS synthesised the literature and wrote a first draft of the manuscript; ML, HP, MD
contributed to the final version and provided several suggestions to improve the quality of the
systematic literature review and the final research challenges. All authors have read and agreed to the
paper being submitted in the present form.
Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
36
Acknowledgements
This work was supported by the Italian Ministry of Education, University and Research under the
Programme “Department of Excellence” Legge 232/2016 (Grant No. CUP - D94I18000260001)
References
Abramovici, M., Göbel, J.C., Savarino, P., 2017. Reconfiguration of smart products during their use phase based on virtual product
twins. CIRP Annals 66, 165–168.
Agrawal, R., Imieliński, T., Swami, A., 1993. Mining association rules between sets of items in large databases, in: Acm Sigmod
Record. ACM, pp. 207–216.
Alam, K.M., El Saddik, A., 2017. C2PS: A digital twin architecture reference model for the cloud-based cyber-physical systems. IEEE
Access 5, 2050–2062.
Alcácer, V., Cruz-Machado, V., 2019. Scanning the Industry 4.0: A Literature Review on Technologies for Manufacturing Systems.
Engineering Science and Technology, an International Journal.
Arrichiello, V., Gualeni, P., 2019. Systems engineering and digital twin: a vision for the future of cruise ships design, production and
operations. International Journal on Interactive Design and Manufacturing (IJIDeM) 1–8.
Asimov, R.M., Chernoshey, S.V., Kruse, I., Osipovich, V.S., 2018. Digital twin in the Analysis of a Big Data.
Atzori, L., Iera, A., Morabito, G., 2010. The internet of things: A survey. Computer networks 54, 2787–2805.
Autiosalo, J., 2018. Platform for industrial internet and digital twin focused education, research, and innovation: Ilmatar the overhead
crane, in: 2018 IEEE 4th World Forum on Internet of Things (WF-IoT). Presented at the 2018 IEEE 4th World Forum on
Internet of Things (WF-IoT), pp. 241–244. https://doi.org/10.1109/WF-IoT.2018.8355217
Ayani, M., Ganebäck, M., Ng, A.H.C., 2018. Digital Twin: Applying emulation for machine reconditioning. Procedia CIRP, 51st CIRP
Conference on Manufacturing Systems 72, 243–248. https://doi.org/10.1016/j.procir.2018.03.139
Bachelor, G., Brusa, E., Ferretto, D., Mitschke, A., 2019. Model-Based Design of Complex Aeronautical Systems Through Digital
Twin and Thread Concepts. IEEE Systems Journal.
Banerjee, A., Dalal, R., Mittal, S., Joshi, K.P., 2017. Generating digital twin models using knowledge graphs for industrial production
lines. UMBC Information Systems Department.
Bao, J., Guo, D., Li, J., Zhang, J., 2018. The modelling and operations for the digital twin in the context of manufacturing. Enterprise
Information Systems 0, 1–23. https://doi.org/10.1080/17517575.2018.1526324
Barni, Andrea, Alessandro Fontana, Silvia Menato, Marzio Sorlini, and Luca Canetta. 2018. ‘Exploiting the Digital Twin in the
Assessment and Optimization of Sustainability Performances’. In 2018 International Conference on Intelligent Systems (IS),
706–13. IEEE.
Barricelli, B.R., Casiraghi, E., Fogli, D., 2019. A Survey on Digital Twin: Definitions, Characteristics, Applications, and Design
Implications. IEEE Access 7, 167653–167671.
Blochwitz, T., Otter, M., Arnold, M., Bausch, C., Elmqvist, H., Junghanns, A., Mauß, J., Monteiro, M., Neidhold, T., Neumerkel, D.,
2011. The functional mockup interface for tool independent exchange of simulation models, in: Proceedings of the 8th
International Modelica Conference; March 20th-22nd; Technical Univeristy; Dresden; Germany. Linköping University
Electronic Press, pp. 105–114.
Blume, M., Koch, N., Imtiaz, J., Flatt, H., Jasperneite, J., Schleipen, M., Sauer, O., Dosch, S., 2014. An OPC-UA based approach for
dynamic-configuration of security credentials and integrating a vendor independent digital product memory, in:
Jahreskolloquium Kommunikation in Der Automation (KommA 2014).
Bohlin, R., Hagmar, J., Bengtsson, K., Lindkvist, L., Carlson, J.S., Söderberg, R., 2017. Data flow and communication framework
supporting digital twin for geometry assurance. Presented at the ASME International Mechanical Engineering Congress and
Exposition, Proceedings (IMECE). https://doi.org/10.1115/IMECE2017-71405
Boje, C., Guerriero, A., Kubicki, S., Rezgui, Y., 2020. Towards a semantic Construction Digital Twin: Directions for future research.
Automation in Construction 114, 103179.
Borangiu, T., Trentesaux, D., Thomas, A., Leitão, P., Barata, J., 2019. Digital transformation of manufacturing through cloud services
and resource virtualization. Computers in Industry 108, 150–162. https://doi.org/10.1016/j.compind.2019.01.006
Boschert, S., Heinrich, C., Rosen, R., 2018. Next generation digital twin, in: Proc. TMCE. Las Palmas de Gran Canaria, Spain, pp.
209–217.
Boschert, S., Rosen, R., 2016. Digital twin—the simulation aspect, in: Mechatronic Futures. Springer, pp. 59–74.
Brenner, B., Hummel, V., 2017. Digital Twin as Enabler for an Innovative Digital Shopfloor Management System in the ESB Logistics
Learning Factory at Reutlingen - University. Procedia Manufacturing, 7th Conference on Learning Factories, CLF 2017 9,
198–205. https://doi.org/10.1016/j.promfg.2017.04.039
Brettel, M., Friederichsen, N., Keller, M., Rosenberg, M., 2014. How virtualization, decentralization and network building change the
manufacturing landscape: An Industry 4.0 Perspective. International journal of mechanical, industrial science and
engineering 8, 37–44.
Bruynseels, K., Santoni de Sio, F., van den Hoven, J., 2018. Digital twins in health care: ethical implications of an emerging engineering
paradigm. Frontiers in genetics 9, 31.
Büchi, G., Cugno, M., Castagnoli, R., 2020. Smart factory performance and Industry 4.0. Technological Forecasting and Social Change
150, 119790. https://doi.org/10.1016/j.techfore.2019.119790
Cai, Y., Starly, B., Cohen, P., Lee, Y.-S., 2017. Sensor data and information fusion to construct digital-twins virtual machine tools for
cyber-physical manufacturing. Procedia Manufacturing 10, 1031–1042.
37
Cheng, J., Zhang, H., Tao, F., Juang, C.-F., 2020. DT-II: Digital twin enhanced Industrial Internet reference framework towards smart
manufacturing. Robotics and Computer-Integrated Manufacturing 62, 101881.
Ciavotta, M., Alge, M., Menato, S., Rovere, D., Pedrazzoli, P., 2017. A microservice-based middleware for the digital factory. Procedia
Manufacturing 11, 931–938.
Cimino, C., Negri, E., Fumagalli, L., 2019. Review of digital twin applications in manufacturing. Computers in Industry 113, 103130.
Clarke, B., Fokoue, E., Zhang, H.H., 2009. Principles and theory for data mining and machine learning. Springer Science & Business
Media.
Da Xu, L., He, W., Li, S., 2014. Internet of things in industries: A survey. IEEE Transactions on industrial informatics 10, 2233–2243.
Damiani, L., Demartini, M., Giribone, P., Maggiani, M., Revetria, R., Tonelli, F., 2018. Simulation and Digital Twin Based Design of
a Production Line: A Case Study.”, in: Proceedings of the International MultiConference of Engineers and Computer
Scientists.
Dassisti, M., De Nicolò, M., 2012. Enterprise integration and economical crisis for mass craftsmanship: a case study of an Italian
furniture company, in: OTM Confederated International Conferences" On the Move to Meaningful Internet Systems".
Springer, pp. 113–123.
Dassisti, M., Giovannini, A., Merla, P., Chimienti, M., Panetto, H., 2019a. An approach to support Industry 4.0 adoption in SMEs
using a core-metamodel. Annual Reviews in Control 47, 266–274.
Dassisti, M., Giovannini, A., Merla, P., Chimienti, M., Panetto, H., 2017. Hybrid Production-System Control-Architecture for Smart
Manufacturing, in: OTM Confederated International Conferences" On the Move to Meaningful Internet Systems". Springer,
pp. 5–15.
Dassisti, M., Panetto, H., Lezoche, M., Merla, P., Semeraro, C., Giovannini, A., Chimienti, M., 2017b. Industry 4.0 paradigm: The
viewpoint of the small and medium enterprises, in: 7th International Conference on Information Society and Technology,
ICIST 2017. pp. 50–54.
Dassisti, M., Semeraro, C., 2018. Smart Sustainable Manufacturing: a new Laboratory-Factory concept to test Industry 4.0 principles,
in: 8th International Conference on Information Society and Technology, ICIST 2018.
Dassisti, M., Semeraro, C., Chimenti, M., 2019b. Hybrid Exergetic Analysis-LCA approach and the Industry 4.0 paradigm: Assessing
Manufacturing Sustainability in an Italian SME. Procedia Manufacturing 33, 655–662.
Ding, K., Chan, F.T.S., Zhang, X., Zhou, G., Zhang, F., 2019. Defining a Digital Twin-based Cyber-Physical Production System for
autonomous manufacturing in smart shop floors. International Journal of Production Research 0, 1–20.
https://doi.org/10.1080/00207543.2019.1566661
Drath, R., Horch, A., 2014. Industrie 4.0: Hit or hype?[industry forum]. IEEE industrial electronics magazine 8, 56–58.
Drath, R., Luder, A., Peschke, J., Hundt, L., 2008. AutomationML-the glue for seamless automation engineering, in: 2008 IEEE
International Conference on Emerging Technologies and Factory Automation. IEEE, pp. 616–623.
Estefan, J.A., 2007. Survey of model-based systems engineering (MBSE) methodologies. Incose MBSE Focus Group 25, 1–12.
Fahad, A., Alshatri, N., Tari, Z., Alamri, A., Khalil, I., Zomaya, A.Y., Foufou, S., Bouras, A., 2014. A survey of clustering algorithms
for big data: Taxonomy and empirical analysis. IEEE transactions on emerging topics in computing 2, 267–279.
Fischer, K., Heintel, M., 2017. Prüfung einer Konsistenz zwischen Referenzdaten eines Fertigungsobjektes und Daten eines digitalen
Zwillings des Fertigungsobjektes. DE102015217855A1.
Garetti, M., Rosa, P., Terzi, S., 2012. Life cycle simulation for the design of product–service systems. Computers in Industry 63, 361–
369.
Glaessgen, E., Stargel, D., 2012. The digital twin paradigm for future NASA and US Air Force vehicles, in: 53rd
AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference 20th AIAA/ASME/AHS
Adaptive Structures Conference 14th AIAA. p. 1818.
Graessler, I., Poehler, A., 2017. Integration of a digital twin as human representation in a scheduling procedure of a cyber-physical
production system, in: Industrial Engineering and Engineering Management (IEEM), 2017 IEEE International Conference
On. IEEE, pp. 289–293.
Greif, T., Stein, N., Flath, C.M., 2020. Peeking into the void: Digital twins for construction site logistics. Computers in Industry 121,
103264. https://doi.org/10.1016/j.compind.2020.103264
Grieves, M., 2014. Digital twin: manufacturing excellence through virtual factory replication. White paper.
Grieves, M., Vickers, J., 2017. Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems, in:
Transdisciplinary Perspectives on Complex Systems. Springer, pp. 85–113.
Grira, N., Crucianu, M., Boujemaa, N., 2004. Unsupervised and semi-supervised clustering: a brief survey. A review of machine
learning techniques for processing multimedia content 1, 9–16.
Gubbi, J., Buyya, R., Marusic, S., Palaniswami, M., 2013. Internet of Things (IoT): A vision, architectural elements, and future
directions. Future generation computer systems 29, 1645–1660.
Guo, J., Zhao, N., Sun, L., Zhang, S., 2018. Modular based flexible digital twin for factory design. Journal of Ambient Intelligence and
Humanized Computing 1–12.
Haag, S., Anderl, R., 2018. Digital twin – Proof of concept. Manufacturing Letters, Industry 4.0 and Smart Manufacturing 15, 64–66.
https://doi.org/10.1016/j.mfglet.2018.02.006
Hearst, M., 2003. What is text mining. SIMS, UC Berkeley 5.
Hermann, M., Pentek, T., Otto, B., 2016. Design principles for industrie 4.0 scenarios, in: System Sciences (HICSS), 2016 49th Hawaii
International Conference On. IEEE, pp. 3928–3937.
Hershey, J.E., Wheeler, F.W., Nielsen, M.C., Johnson, C.D., Dell’Anno, M.J., JOYKUTTI, J., 2017. Digital twin of twinned physical
system. US20170286572A1.
38
International Council on Systems Engineering (INCOSE), 2007. International Council on Systems Engineering (INCOSE), Systems
Engineering Vision 2020, Version 2.03, TP-2004-004-02, September 2007. [WWW Document]. URL
http://www.omg.org/spec/ OCL/
Jain, R., Bhatnagar, R., 2020. Applications of Machine Learning in Cyber Security - A Review and a Conceptual Framework for a
University Setup. Advances in Intelligent Systems and Computing 921, 599–608. https://doi.org/10.1007/978-3-030-14118-
9_60
Jain, S., Shao, G., Shin, S.-J., 2017. Manufacturing data analytics using a virtual factory representation. International Journal of
Production Research 55, 5450–5464.
Jones, D., Snider, C., Nassehi, A., Yon, J., Hicks, B., 2020. Characterising the Digital Twin: A systematic literature review. CIRP
Journal of Manufacturing Science and Technology.
Kagermann, H., Helbig, J., Hellinger, A., Wahlster, W., 2013. Recommendations for implementing the strategic initiative INDUSTRIE
4.0: Securing the future of German manufacturing industry; final report of the Industrie 4.0 Working Group.
Forschungsunion.
Kaur, M.J., Mishra, V.P., Maheshwari, P., 2020. The Convergence of Digital Twin, IoT, and Machine Learning: Transforming Data
into Action, in: Farsi, M., Daneshkhah, A., Hosseinian-Far, A., Jahankhani, H. (Eds.), Digital Twin Technologies and Smart
Cities, Internet of Things. Springer International Publishing, Cham, pp. 3–17. https://doi.org/10.1007/978-3-030-18732-3_1
Khajavi, Siavash H., Naser Hossein Motlagh, Alireza Jaribion, Liss C. Werner, and Jan Holmström. 2019. ‘Digital Twin: Vision,
Benefits, Boundaries, and Creation for Buildings’. IEEE Access 7: 147406–19.
Kritzinger, W., Karner, M., Traar, G., Henjes, J., Sihn, W., 2018. Digital Twin in manufacturing: A categorical literature review and
classification. IFAC-PapersOnLine 51, 1016–1022.
Kusiak, A., 2018. Smart manufacturing. International Journal of Production Research 56, 508–517.
Law, A.M., Kelton, W.D., Kelton, W.D., 2000. Simulation modeling and analysis. McGraw-Hill New York.
Lee, C.G., Park, S.C., 2014. Survey on the virtual commissioning of manufacturing systems. Journal of Computational Design and
Engineering 1, 213–222.
Lee, H., Kim, T., 2018. Smart factory use case model based on digital twin. ICIC Express Letters, Part B: Applications 9, 931–936.
https://doi.org/10.24507/icicelb.09.09.931
Lee, I., Lee, K., 2015. The Internet of Things (IoT): Applications, investments, and challenges for enterprises. Business Horizons 58,
431–440.
Lee, J., Bagheri, B., Kao, H.-A., 2015. A cyber-physical systems architecture for industry 4.0-based manufacturing systems.
Manufacturing Letters 3, 18–23.
Lee, J., Kao, H.-A., Yang, S., 2014a. Service innovation and smart analytics for industry 4.0 and big data environment. Procedia Cirp
16, 3–8.
Lee, J., Lapira, E., Yang, S., Kao, A., 2013. Predictive manufacturing system-Trends of next-generation production systems. IFAC
Proceedings Volumes 46, 150–156.
Lee, J., Wu, F., Zhao, W., Ghaffari, M., Liao, L., Siegel, D., 2014b. Prognostics and health management design for rotary machinery
systems—Reviews, methodology and applications. Mechanical systems and signal processing 42, 314–334.
Leng, J., Zhang, H., Yan, D., Liu, Q., Chen, X., Zhang, D., 2019. Digital twin-driven manufacturing cyber-physical system for parallel
controlling of smart workshop. Journal of Ambient Intelligence and Humanized Computing 10, 1155–1166.
Leng, J., Zhang, H., Yan, D., Liu, Q., Chen, X., Zhang, D., 2018. Digital twin-driven manufacturing cyber-physical system for parallel
controlling of smart workshop. Journal of Ambient Intelligence and Humanized Computing 1–12.
Lezoche, M., Panetto, H., 2018. Cyber-Physical Systems, a new formal paradigm to model redundancy and resiliency. Enterprise
Information Systems 1–22.
Liau, Y., Lee, H., Ryu, K., 2018. Digital Twin concept for smart injection molding. Presented at the IOP Conference Series: Materials
Science and Engineering. https://doi.org/10.1088/1757-899X/324/1/012077
Liu, Q., Zhang, H., Leng, J., Chen, X., 2018. Digital twin-driven rapid individualised designing of automated flow-shop manufacturing
system. International Journal of Production Research 1–17.
Liu, Z., Meyendorf, N., Mrad, N., 2018. The role of data fusion in predictive maintenance using digital twin. Presented at the AIP
Conference Proceedings. https://doi.org/10.1063/1.5031520
Ljubljana, B.L., University of, 2005. Data Mining [WWW Document]. URL https://orange.biolab.si/ (accessed 10.12.20).
Luo, W., Hu, T., Zhu, W., Tao, F., 2018. Digital twin modeling method for CNC machine tool, in: 2018 IEEE 15th International
Conference on Networking, Sensing and Control (ICNSC). Presented at the 2018 IEEE 15th International Conference on
Networking, Sensing and Control (ICNSC), pp. 1–4. https://doi.org/10.1109/ICNSC.2018.8361285
Ma, Y., Borrelli, F., Hencey, B., Coffey, B., Bengea, S., Haves, P., 2011. Model predictive control for the operation of building cooling
systems. IEEE Transactions on control systems technology 20, 796–803.
Macchi, M., Roda, I., Negri, E., Fumagalli, L., 2018. Exploring the role of Digital Twin for Asset Life cycle Management. IFAC-
PapersOnLine 51, 790–795.
Madni, A.M., Madni, C.C., Lucero, S.D., 2019. Leveraging digital twin technology in model-based systems engineering. Systems 7,
7.
Mandolla, C., Petruzzelli, A.M., Percoco, G., Urbinati, A., 2019. Building a digital twin for additive manufacturing through the
exploitation of blockchain: A case analysis of the aircraft industry. Computers in Industry 109, 134–152.
https://doi.org/10.1016/j.compind.2019.04.011
Mikolov, T., Sutskever, I., Chen, K., Corrado, G., Dean, J., 2013. Distributed representations of words and phrases and their
compositionality. arXiv preprint arXiv:1310.4546.
39
Mohammadi, N., Taylor, J.E., 2017. Smart city digital twins, in: 2017 IEEE Symposium Series on Computational Intelligence (SSCI).
IEEE, pp. 1–5.
Monostori, L., Kádár, B., Bauernhansl, T., Kondoh, S., Kumara, S., Reinhart, G., Sauer, O., Schuh, G., Sihn, W., Ueda, K., 2016.
Cyber-physical systems in manufacturing. Cirp Annals 65, 621–641.
Negri, E., Fumagalli, L., Cimino, C., Macchi, M., 2019. FMU-supported simulation for CPS Digital Twin. Procedia Manufacturing,
7th International conference on Changeable, Agile, Reconfigurable and Virtual Production (CARV2018) 28, 201–206.
https://doi.org/10.1016/j.promfg.2018.12.033
Negri, E., Fumagalli, L., Macchi, M., 2017. A Review of the Roles of Digital Twin in CPS-based Production Systems. Procedia
Manufacturing, 27th International Conference on Flexible Automation and Intelligent Manufacturing, FAIM2017, 27-30
June 2017, Modena, Italy 11, 939–948. https://doi.org/10.1016/j.promfg.2017.07.198
Nikolakis, N., Alexopoulos, K., Xanthakis, E., Chryssolouris, G., 2018. The digital twin implementation for linking the virtual
representation of human-based production tasks to their physical counterpart in the factory-floor. International Journal of
Computer Integrated Manufacturing 1–12.
Osterrieder, P., Budde, L., Friedli, T., 2019. The smart factory as a key construct of industry 4.0: A systematic literature review.
International Journal of Production Economics. https://doi.org/10.1016/j.ijpe.2019.08.011
Onosato, M., Iwata, K., 1993. Development of a virtual manufacturing system by integrating product models and factory models. CIRP
annals 42, 475–478.
Panetto, H., Iung, B., Ivanov, D., Weichhart, G., Wang, X., 2019. Challenges for the cyber-physical manufacturing enterprises of the
future. Annual Reviews in Control 47, 200–213.
Park, K.T., Nam, Y.W., Lee, H.S., Im, S.J., Noh, S.D., Son, J.Y., Kim, H., 2019. Design and implementation of a digital twin application
for a connected micro smart factory. International Journal of Computer Integrated Manufacturing 32, 596–614.
https://doi.org/10.1080/0951192X.2019.1599439
Pirola, F., Boucher, X., Wiesner, S., Pezzotta, G., 2020. Digital technologies in product-service systems: a literature review and a
research agenda. Computers in Industry 123, 103301. https://doi.org/10.1016/j.compind.2020.103301
Platenius-Mohr, Marie, Somayeh Malakuti, Sten Grüner, Johannes Schmitt, and Thomas Goldschmidt. 2020. ‘File-and API-Based
Interoperability of Digital Twins by Model Transformation: An IIoT Case Study Using Asset Administration Shell’. Future
Generation Computer Systems 113: 94–105.
Ponomarev, K., Kudryashov, N., Popelnukha, N., Potekhin, V., 2017. Main principals and issues of digital twin development for
complex technological processes. Presented at the Annals of DAAAM and Proceedings of the International DAAAM
Symposium, pp. 523–528. https://doi.org/10.2507/28th.daaam.proceedings.074
Posada, J., Toro, C., Barandiaran, I., Oyarzun, D., Stricker, D., de Amicis, R., Pinto, E.B., Eisert, P., Döllner, J., Vallarino, I., 2015.
Visual computing as a key enabling technology for industrie 4.0 and industrial internet. IEEE computer graphics and
applications 35, 26–40.
Post, J., Groen, M., Klaseboer, G., 2009. PHYSICAL MODEL BASED DIGITAL TWINS IN MANUFACTURING PROCESSES.
Optics Letters 34, 1982–4.
Prasad, B., 1989. Robotics and Factories of the Future. Springer.
Putnik, G., Sluga, A., ElMaraghy, H., Teti, R., Koren, Y., Tolio, T., Hon, B., 2013. Scalability in manufacturing systems design and
operation: State-of-the-art and future developments roadmap. CIRP Annals 62, 751–774.
Qi, Q., Tao, F., 2018. Digital Twin and Big Data Towards Smart Manufacturing and Industry 4.0: 360 Degree Comparison. IEEE
Access 6, 3585–3593.
Qi, Q., Tao, F., Zuo, Y., Zhao, D., 2018. Digital Twin Service towards Smart Manufacturing. Procedia CIRP, 51st CIRP Conference
on Manufacturing Systems 72, 237–242. https://doi.org/10.1016/j.procir.2018.03.103
Redelinghuys, A., Basson, A., Kruger, K., 2019. A six-layer digital twin architecture for a manufacturing cell. Studies in Computational
Intelligence 803, 412–423. https://doi.org/10.1007/978-3-030-03003-2_32
Ríos, J., Hernández, J.C., Oliva, M., Mas, F., 2015. Product Avatar as Digital Counterpart of a Physical Individual Product: Literature
Review and Implications in an Aircraft., in: ISPE CE. pp. 657–666.
Rodič, B., 2017. Industry 4.0 and the New Simulation Modelling Paradigm. Organizacija 50, 193–207. https://doi.org/10.1515/orga-
2017-0017
Rosen, R., von Wichert, G., Lo, G., Bettenhausen, K.D., 2015. About the importance of autonomy and digital twins for the future of
manufacturing. IFAC-PapersOnLine 48, 567–572.
Ruppert, T., Jaskó, S., Holczinger, T., Abonyi, J., 2018. Enabling Technologies for Operator 4.0: A Survey. Applied Sciences 8, 1650.
Schleich, B., ANWER, N., Mathieu, L., Wartzack, S., 2017. Shaping the digital twin for design and production engineering. CIRP
Annals - Manufacturing Technology 66. https://doi.org/10.1016/j.cirp.2017.04.040
Schluse, M., Atorf, L., Rossmann, J., 2017. Experimentable digital twins for model-based systems engineering and simulation-based
development, in: Systems Conference (SysCon), 2017 Annual IEEE International. IEEE, pp. 1–8.
Schluse, M., Priggemeyer, M., Atorf, L., Rossmann, J., 2018. Experimentable Digital Twins—Streamlining Simulation-Based Systems
Engineering for Industry 4.0. IEEE Transactions on Industrial Informatics 14, 1722–1731.
Schroeder, G., Steinmetz, C., Pereira, C.E., Muller, I., Garcia, N., Espindola, D., Rodrigues, R., 2016. Visualising the digital twin using
web services and augmented reality, in: 2016 IEEE 14th International Conference on Industrial Informatics (INDIN). IEEE,
pp. 522–527.
Schroeder, G.N., Steinmetz, C., Pereira, C.E., Espindola, D.B., 2016. Digital twin data modeling with automationML and a
communication methodology for data exchange. IFAC-PapersOnLine 49, 12–17.
40
Semeraro, C., Lezoche, M., Panetto, H., Dassisti, M., Cafagna, S., 2019a. Data-driven pattern-based constructs definition for the digital
transformation modelling of collaborative networked manufacturing enterprises, in: Working Conference on Virtual
Enterprises. Springer, pp. 507–515.
Semeraro, C., Panetto, H., Lezoche, M., Dassisti, M., Cafagna, S., 2019b. A Monitoring Strategy for Industry 4.0: Master Italy srl Case
Study. INSIGHT-International Council on Systems Engineering (INCOSE) 22, 20–22.
Semeraro, C. 2020. Contribution to the Formalization of Data-Driven Invariant Modelling Constructs of Cyber-Physical Systems, PhD
thesis.
Škopljanac-Mačina, F., Blašković, B., 2014. Formal concept analysis–overview and applications. Procedia Engineering 69, 1258–
1267.
Söderberg, R., Wärmefjord, K., Carlson, J.S., Lindkvist, L., 2017. Toward a Digital Twin for real-time geometry assurance in
individualized production. CIRP Annals 66, 137–140.
Song, Z., Canedo, A.M., 2016. Digital twins for energy efficient asset maintenance. US20160247129A1.
Stark, R., Kind, S., Neumeyer, S., 2017. Innovations in digital modelling for next generation manufacturing system design. CIRP
Annals 66, 169–172.
Suh, N.P., 1984. The future of the factory. Robotics and Computer-Integrated Manufacturing 1, 39–49.
Sun, H., Li, C., Fang, X., Gu, H., 2017. Optimized throughput improvement of assembly flow line with digital twin online analytics,
in: Robotics and Biomimetics (ROBIO), 2017 IEEE International Conference On. IEEE, pp. 1833–1837.
Sutharssan, T., Stoyanov, S., Bailey, C., Yin, C., 2015. Prognostic and health management for engineering systems: a review of the
data-driven approach and algorithms. The Journal of Engineering 2015, 215–222.
Talkhestani, B.A., Jazdi, N., Schlögl, W., Weyrich, M., 2018. A concept in synchronization of virtual production system with real
factory based on anchor-point method. Presented at the Procedia CIRP, pp. 13–17.
https://doi.org/10.1016/j.procir.2017.12.168
Tan, L., Wang, N., 2010. Future internet: The internet of things, in: 2010 3rd International Conference on Advanced Computer Theory
and Engineering (ICACTE). IEEE, pp. V5-376.
Tao, F., Cheng, J., Qi, Q., Zhang, M., Zhang, H., Sui, F., 2018a. Digital twin-driven product design, manufacturing and service with
big data. The International Journal of Advanced Manufacturing Technology 94, 3563–3576.
Tao, F., Qi, Q., Wang, L., Nee, A.Y.C., 2019. Digital twins and cyber-physical systems toward smart manufacturing and Industry 4.0:
Correlation and comparison. Engineering 5, 653–661.
Tao, F., Sui, F., Liu, A., Qi, Q., Zhang, M., Song, B., Guo, Z., Lu, S.C.-Y., Nee, A.Y.C., 2018b. Digital twin-driven product design
framework. International Journal of Production Research 1–19.
Tao, F., Zhang, M., 2017. Digital twin shop-floor: a new shop-floor paradigm towards smart manufacturing. Ieee Access 5, 20418–
20427.
Tao, F., Zhang, M., Liu, Y., Nee, A.Y.C., 2018c. Digital twin driven prognostics and health management for complex equipment. CIRP
Annals.
Tidriri, K., Chatti, N., Verron, S., Tiplica, T., 2016. Bridging data-driven and model-based approaches for process fault diagnosis and
health monitoring: A review of researches and future challenges. Annual Reviews in Control 42, 63–81.
Trauer, J., Schweigert-Recksiek, S., Engel, C., Spreitzer, K., Zimmermann, M., 2020. What is a Digital Twin?–Definitions and Insights
from an Industrial Case Study in Technical Product Development, in: Proceedings of the Design Society: DESIGN
Conference. Cambridge University Press, pp. 757–766.
Tuegel, E., 2012. The airframe digital twin: some challenges to realization, in: 53rd AIAA/ASME/ASCE/AHS/ASC Structures,
Structural Dynamics and Materials Conference 20th AIAA/ASME/AHS Adaptive Structures Conference 14th AIAA. p.
1812.
Tuegel, E.J., Ingraffea, A.R., Eason, T.G., Spottswood, S.M., 2011. Reengineering aircraft structural life prediction using a digital
twin. International Journal of Aerospace Engineering 2011.
Uhlemann, T.H.-J., Lehmann, C., Steinhilper, R., 2017. The digital twin: Realizing the cyber-physical production system for industry
4.0. Procedia Cirp 61, 335–340.
Um, J., Weyer, S., Quint, F., 2017. Plug-and-Simulate within Modular Assembly Line enabled by Digital Twins and the use of
AutomationML. IFAC-PapersOnLine 50, 15904–15909.
Vachálek, J., Bartalskỳ, L., Rovnỳ, O., Šišmišová, D., Morháč, M., Lokšík, M., 2017. The digital twin of an industrial production line
within the industry 4.0 concept, in: Process Control (PC), 2017 21st International Conference On. IEEE, pp. 258–262.
van Otterlo, M., Wiering, M., 2012. Reinforcement learning and markov decision processes, in: Reinforcement Learning. Springer, pp.
3–42.
Wang, J., Ye, L., Gao, R.X., Li, C., Zhang, L., 2018. Digital Twin for rotating machinery fault diagnosis in smart manufacturing.
International Journal of Production Research 0, 1–15. https://doi.org/10.1080/00207543.2018.1552032
Wang, X.V., Wang, L., 2018. Digital twin-based WEEE recycling, recovery and remanufacturing in the background of Industry 4.0.
International Journal of Production Research 1–11.
Wärmefjord, K., Söderberg, R., Lindkvist, L., Lindau, B., Carlson, J.S., 2017. Inspection Data to Support a Digital Twin for Geometry
Assurance, in: ASME 2017 International Mechanical Engineering Congress and Exposition. American Society of Mechanical
Engineers, p. V002T02A101-V002T02A101.
Weber, C., Königsberger, J., Kassner, L., Mitschang, B., 2017. M2DDM–a maturity model for data-driven manufacturing. Procedia
CIRP 63, 173–178.
Weyer, S., Meyer, T., Ohmer, M., Gorecky, D., Zühlke, D., 2016. Future modeling and simulation of CPS-based factories: an example
from the automotive industry. IFAC-PapersOnLine 49, 97–102.
41
Wille, R., 2002. Why can concept lattices support knowledge discovery in databases? Journal of Experimental & Theoretical Artificial
Intelligence 14, 81–92. https://doi.org/10.1080/09528130210164161
Xie, J., Wang, X., Yang, Z., Hao, S., 2019. Virtual monitoring method for hydraulic supports based on digital twin theory. Mining
Technology 0, 1–11. https://doi.org/10.1080/25726668.2019.1569367
Xu, R., Wunsch, D.C., 2005. Survey of clustering algorithms.
Xu, Y., Sun, Y., Liu, X., Zheng, Y., 2019. A Digital-Twin-Assisted Fault Diagnosis Using Deep Transfer Learning. IEEE Access 7,
19990–19999.
Yao, X., Zhou, J., Lin, Y., Li, Y., Yu, H., Liu, Y., 2019. Smart manufacturing based on cyber-physical systems and beyond. Journal of
Intelligent Manufacturing 30, 2805–2817.
Yun, S., Park, J.-H., Kim, W.-T., 2017. Data-centric middleware based digital twin platform for dependable cyber-physical systems,
in: Ubiquitous and Future Networks (ICUFN), 2017 Ninth International Conference On. IEEE, pp. 922–926.
Zhang, G.P., 2000. Neural networks for classification: a survey. IEEE Transactions on Systems, Man, and Cybernetics, Part C
(Applications and Reviews) 30, 451–462.
Zhang, H., Liu, Q., Chen, X., Zhang, D., Leng, J., 2017. A Digital Twin-Based Approach for Designing and Multi-Objective
Optimization of Hollow Glass Production Line. IEEE Access 5, 26901–26911.
https://doi.org/10.1109/ACCESS.2017.2766453
Zhang, H., Zhang, G., Yan, Q., 2018. Dynamic resource allocation optimization for digital twin-driven smart shopfloor, in: 2018 IEEE
15th International Conference on Networking, Sensing and Control (ICNSC). Presented at the 2018 IEEE 15th International
Conference on Networking, Sensing and Control (ICNSC), pp. 1–5. https://doi.org/10.1109/ICNSC.2018.8361283
Zhang, L., Zhou, L., Ren, L., Laili, Y., 2019. Modeling and simulation in intelligent manufacturing. Computers in Industry 112, 103123.
Zhang, M., Zuo, Y., Tao, F., 2018. Equipment energy consumption management in digital twin shop-floor: A framework and potential
applications, in: 2018 IEEE 15th International Conference on Networking, Sensing and Control (ICNSC). Presented at the
2018 IEEE 15th International Conference on Networking, Sensing and Control (ICNSC), pp. 1–5.
https://doi.org/10.1109/ICNSC.2018.8361272
Zhang, Y., Ren, S., Liu, Y., Sakao, T., Huisingh, D., 2017. A framework for Big Data driven product life cycle management. Journal
of Cleaner Production 159, 229–240.
Zhuang, C., Liu, J., Xiong, H., 2018. Digital twin-based smart production management and control framework for the complex product
assembly shop-floor. The International Journal of Advanced Manufacturing Technology 96, 1149–1163.
42

Digital Twin Paradigm: A Systematic Literature Review: Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti

Uploaded by

Copyright:

Available Formats

Digital Twin Paradigm: A Systematic Literature Review: Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Digital Twin Paradigm: A Systematic Literature Review: Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti

Uploaded by

Copyright:

Available Formats

Digital twin paradigm: A systematic literature review

Concetta Semeraro, Mario Lezoche, Hervé Panetto, Michele Dassisti

To cite this version:

HAL Id: hal-03218786

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

Concetta Semeraro*,1,2, Mario Lezoche2, Hervé Panetto2,

Keywords: Digital Twin; Industry 4.0; Cyber-Physical Systems; Predictive manufacturing.

1 Literature Review Methodology

Step 1: Paper Selection: Publications identification and screening

Step 2 Extraction of DT Features: Text mining analysis

Step 3: Knowledge Representation: Formal Concept Analysis

Table 1: FCA Table

O 𝑨𝒕𝒕𝒓𝟏 𝑨𝒕𝒕𝒓𝟐 𝑨𝒕𝒕𝒓𝟑 𝑨𝒕𝒕𝒓𝒎

2 State of Art Digital Twin: Definitions and Sights

Table 2: Digital Twin Definitions

DIGITAL TWIN DEFINITIONS

Figure 5: Hierarchical Clustering (HCA) Results - Clusters DT definitions

Figure 6: Word Cloud DT Definitions of Cluster C1

Figure 7: Word Cloud DT Definitions of Cluster C2

Figure 8: Word Cloud Definitions of Cluster C3

Figure 9: Word Cloud DT Definitions of Cluster C4

3 Digital Twin Paradigm

ID TOPIC MODELLING: Set of Words TOPIC LABEL

TOPIC 1 Support-production, decision-making, simulation,

Figure 12: Overview Distribution of Topic Modelling

3.1 Digital Twin Application Contexts (‘Where is appropriate to use a Digital

Table 5: Formal Concepts of Life Cycle phases and Functions

Table 6: Formal Concepts of Physical layer and its Components/Technologies

Figure 16: Lattice of Networking Layer and its Components/Technologies

Table 7: Formal Concepts of Networking Layer and its Components/Technologies

Table 8: Formal Concepts of Computing Layer and its Components/Technologies

4 Conclusion and Research Challenges

Table 9: Summary table of the review results (5W1H)

Research Question Results

Research Challenge #1.

Research Challenge #2.

Research Challenge #3.

Declaration of Competing Interest

You might also like