Value Creation with Digital Twins E-Book

Linard Dario Barth

• Herausgeber: BoD - Books on Demand
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 319

Veröffentlichungsjahr: 2024

Beschreibung

A digital twin is a digital representation of a real-world counterpart, which can receive and provide data to create value within a use case. Digital twins create value for users by enabling new and enhanced smart services. However, ambiguous definitions and terminology coupled with a lack of shared conceptual reference frameworks complicate cross-functional discussions and hinder the widespread implementation of digital twins. This thesis proposes a new definition and presents two conceptual reference frameworks to systematically depict value creation with digital twins. A design science research approach with mixed methods was used to iteratively design and evaluate these artifacts while ensuring scientific rigor, practical relevance, and usefulness. The applied methods within the five research phases include systematic literature research, interviews, workshops with academic experts, qualitative and quantitative questionnaires, workshops with practice experts, and an in-depth case study in smart waste management. The major findings of this research are (i) the proposal of a new definition of digital twins that reflects a practical understanding by focusing on value creation; (ii) a scientific conceptual reference framework focusing on completeness by distinguishing 81 elements involved in value creation with digital twins; (iii) a second, more application-oriented conceptual reference framework focusing on the interrelations of the elements essential for the value creation in practice; and (iv) an instantiation of the application-oriented framework for the use case of the in-depth case study. All artifacts are consistent in content and include the following main dimensions, which are to be considered when creating value with digital twins: data resources, internal value creation, and external value creation. These artifacts contribute to a common understanding of value creation with digital twins in research and practice. Furthermore, they enable researchers and practitioners to structure their digital twin activities and communicate them to internal and external stakeholders.

Leseprobe

I am deeply grateful to my family, friends, and teachers for giving me the freedom to explore and for encouraging and helping me to grow along the way.

First and foremost, I would like to thank my supervisor Prof. Radim Farana for giving me the opportunity and the confidence to approach my doctorate with this topic and objective. I greatly appreciate his openness and encouragement, which always allowed me to work on my ideas. I would also like to extend my sincere appreciation to Assoc. Prof. Pavel Žufan for his continuous support, scientific insight, and valuable guidance over the last few years. In addition, I would like to thank Doc. Veronika Solilová and Prof. Petr David for helping me in the difficult final phase to keep my goal in mind and to continue to pursue it.

Heartfelt expressions of gratitude are also due to Dr. Matthias Ehrat. His valuable and thought-provoking input repeatedly sparked new ideas, and his challenging and constructive criticism caused additional iterations crucial for the final results.

I am also indebted to Prof. Rainer Fuchs. He motivated me to undertake my doctorate in the first place and has allowed me to develop myself personally and professionally within his team over the last five years. It is a great pleasure being part of this outstanding team, and this journey would not have been possible without the support of current and former colleagues.

Finally, my deepest gratitude goes to my family for all their love, unconditional support, optimism, and motivating words that have carried me. Above all, my thanks go to Vera, who has contributed so much, not only through conversations that stimulated and challenged my thinking but also through her tireless encouragement, love, and unwavering belief in me.

Statutory declaration

Herewith I declare that I have written my final thesis: Value Creation with Digital Twins: Conceptual Reference Frameworks and Case Study by myself and all sources and data used are quoted in the list of references. I agree that my work will be published in accordance with Section 47b of Act No. 111/1998 Sb. on Higher Education as amended thereafter and in accordance with the Guidelines on the Publishing of University Student Theses.

I am aware of the fact that my thesis is subject to Act. No. 121/2000 Sb., the Copyright Act and that the Mendel University in Brno is entitled to close a licence agreement and use the results of my thesis as the “School Work” under the terms of Section 60 para. 1 of the Copyright Act.

Before closing a licence agreement on the use of my thesis with another person (subject) I undertake to request for a written statement of the university that the licence agreement in question is not in conflict with the legitimate interests of the university, and undertake to pay any contribution, if eligible, to the costs associated with the creation of the thesis, up to their actual amount.

In Brno on 29th August, 2023

Abstract

A digital twin is a digital representation of a real-world counterpart, which can receive and provide data to create value within a use case. Digital twins create value for users by enabling new and enhanced smart services. However, ambiguous definitions and terminology coupled with a lack of shared conceptual reference frameworks complicate cross-functional discussions and hinder the widespread implementation of digital twins.

This thesis proposes a new definition and presents two conceptual reference frameworks to systematically depict value creation with digital twins. A design science research approach with mixed methods was used to iteratively design and evaluate these artifacts while ensuring scientific rigor, practical relevance, and usefulness. The applied methods within the five research phases include systematic literature research, interviews, workshops with academic experts, qualitative and quantitative questionnaires, workshops with practice experts, and an in-depth case study in smart waste management.

The major findings of this research are (i) the proposal of a new definition of digital twins that reflects a practical understanding by focusing on value creation; (ii) a scientific conceptual reference framework focusing on completeness by distinguishing 81 elements involved in value creation with digital twins; (iii) a second, more application-oriented conceptual reference framework focusing on the interrelations of the elements essential for the value creation in practice; and (iv) an instantiation of the application-oriented framework for the use case of the in-depth case study. All artifacts are consistent in content and include the following main dimensions, which are to be considered when creating value with digital twins: data resources, internal value creation, and external value creation. These artifacts contribute to a common understanding of value creation with digital twins in research and practice. Furthermore, they enable researchers and practitioners to structure their digital twin activities and communicate them to internal and external stakeholders.

Keywords

Digital twin, value creation, ontology, conceptual reference framework, design science research

Abstrakt

Digitální dvojče je počítačová reprezentace systému v reálném světě, která může reagovat na vstupy a produkovat odpovídající výstupní data v rámci různých případů užití. Digitální dvojčata jsou pro uživatele hodnotná tím, že umožňují vytváření vylepšených nebo nových služeb. Nejednoznačné definice a terminologie spolu s nedostatkem sdílených konceptuálních referenčních frameworků však komplikují mezioborové diskuse a brání širokému zavádění digitálních dvojčat v praxi.

Tato disertační práce navrhuje novou definici a představuje dva konceptuální referenční frameworky pro systematickou prezentaci hodnot vytvořených pomocí digitálních dvojčat. K opakovanému navrhování a vyhodnocování těchto reprezentací byl nutný vědecký výzkum v oblasti jejich návrhu, využívající kombinaci různých vědeckých metod, aby byla zajištěna vědecká přesnost, praktická relevance a užitečnost dosažených výsledků. Mezi aplikované metody v rámci pěti realizovaných výzkumných fází patřil systematický průzkum literárních zdrojů, rozhovory, workshopy s akademickými odborníky, kvalitativní a kvantitativní výzkum, workshopy s odborníky z praxe a hloubková případová studie v oblasti chytrého nakládání s odpady.

Hlavní závěry tohoto výzkumu jsou (i) návrh nové definice digitálních dvojčat, která odráží praktické chápání se zaměřením na vytváření nových hodnot; ii) vědecký konceptuální referenční framework zaměřující se na kompletnost reprezentace rozlišováním 81 prvků zapojených do vytváření hodnot digitálními dvojčaty; (iii) druhý, více aplikačně orientovaný konceptuální referenční framework zaměřující se na vzájemné vztahy prvků zásadních pro tvorbu hodnot v praxi; a (iv) konkretizaci aplikačně orientovaného frameworku pro vytvoření hloubkové případové studie. Všechny části jsou obsahově konzistentní a zahrnují následující hlavní dimenze, které je třeba vzít v úvahu při vytváření hodnot pomocí digitálních dvojčat: datové zdroje, vnitřní vytváření hodnot a vytváření externích hodnot. Tyto části přispívají ke společnému porozumění vytváření hodnot digitálními dvojčaty ve výzkumu a praxi. Kromě toho umožňují výzkumným pracovníkům a odborníkům z praxe strukturovat své aktivity v oblasti digitálních dvojčat a sdělovat je interním a externím zainteresovaným stranám.

Klíčová slova

Digitální dvojče, vytváření hodnot, ontologie, konceptuální referenční framework, výzkum v oblasti návrhu systémů

List of Abbreviations

API

Application programming interface

BoL

Beginning-of-life

CAD

Computer-aided design

cf.

Latin

confer

, meaning “compare”

CHF

Confoederatio Helvetica franc, Swiss franc

CoAP

Constrained application protocol

CPS(s)

Cyber-physical system(s)

CPSS(s)

Cyber-physical service system(s)

CRM

Customer relationship management

DSR

Design science research

DT(s)

Digital twin(s)

DTF

Digital twin framework, abbreviated version of “conceptual reference framework for value creation with digital twins”

e.g.

Latin

exempli gratia

, meaning “for example”

EIH

Embedded information hardware

EIS

Embedded information system

EoL

End-of-life

ERP

Enterprise resource planning

et al.

Latin

et aliī

, meaning “and others”

H2M

Human-to-machine

HTTP

Hypertext transfer protocol

i.e.

Latin

id est

, meaning “that is”

IIC

Industrial internet consortium

IoT

Internet of things

IS

Information systems

KPI

Key performance indicator

LoRaWAN

Long range wide area network

LPN

Low power networks

LPWAN

Low power wide area networks

LTE

Long-term evolution

M2M

Machine-to-machine

MES

Manufacturing execution system

MIS

Management information system

MoL

Middle-of-life

MQTT

Message queue telemetry transport

NASA

National Aeronautics and Space Administration

n.d.

No date

NFC

Near-field communication

NB-IoT

Narrowband internet of things

OEE

Overall equipment effectiveness

OSE

Overall service effectiveness

PDM

Product data management

PLM

Product life cycle management

PSS(s)

Product-service system(s)

RAMI 4.0

Reference Architecture Model Industrie 4.0

REST

Representational state transfer

RFID

Radio-frequency identification

SCM

Supply chain management

SCP(s)

Smart connected product(s)

VDMA

Verband Deutscher Maschinen- und Anlagenbau e.V.

VS

Virtual space

ZVEI

Verband der Elektro- und Digitalindustrie e.V.

Contents

1 Introduction

1.1 Smart Connected Products and Systems

1.2 The Digital Twin Concept

1.2.1 Origin

1.2.2 Status of Definitions

1.2.3 Properties and Characteristics of Digital Twins

1.2.4 Proposal for a New Definition

1.2.5 Progression of Value Creation with Digital Twins

1.3 Relevance

1.3.1 For Research

1.3.2 For Practice

1.4 Research Gaps

1.4.1 Corporate Knowledge Gaps

1.4.2 Academic Research Gaps

1.4.3 Call for a Conceptual Reference Framework of Digital Twins

1.5 Structure of Thesis

2 Objectives and Methods

2.1 Objectives

2.1.1 Partial Objectives

2.1.2 Contribution

2.2 Design Science in Information Systems Research

2.2.1 Nature and Relevance

2.2.2 Research Phases and Methods

2.2.3 Research Principles

2.2.4 Research Guidelines

2.3 Methods

2.3.1 Overarching Research Framework and Research Phases

2.3.2 Research phase 1: Literature Review

2.3.3 Research Phase 2: Interviews and Workshops

2.3.4 Research Phase 3: Expert Reviews with Questionnaire

2.3.5 Research Phase 4: Workshops

2.3.6 Research Phase 5: Case Study

3 Ontology and Scientific Digital Twin Framework

3.1 Data Resources

3.1.1 Data Categories

3.1.2 Data Sources

3.1.3 Data Format

3.2 External Value Creation

3.2.1 Service Scope

3.2.2 Smartness Maturity

3.2.3 System Hierarchy Levels

3.3 Internal Value Creation

3.3.1 Life Cycle

3.3.2 Value Creation Hierarchy

3.3.3 Generations/Time

3.4 Scientific Digital Twin Framework

4 Evaluation of the Scientific Digital Twin Framework

4.1 Qualitative Evaluation

4.1.1 Strengths and Weaknesses

4.1.2 Terms and Elements

4.2 Quantitative Evaluation

4.2.1 Framework Quality

4.2.2 Applicability and Use in Practice

4.3 Additional Insights

4.3.1 System Hierarchy Levels

4.3.1 Service Scope

4.3.2 Smartness Maturity

4.3.3 Value Creation Hierarchy

4.3.4 Life Cycle

4.3.5 Generations/Time

4.3.6 Data Categories

4.3.7 Data Sources

4.3.8 Data Format

5 Application-Oriented Digital Twin Framework

5.1 Introductory Representations

5.2 Product Life Cycle

5.3 Data Sources

5.3.1 Things as Real-World Counterpart Data Sources

5.3.2 Internal and External Data Sources

5.4 From Data to Value

5.4.1 Data Categories

5.4.2 Data Analysis

5.4.3 From Services to Value

5.4.4 Smartness Maturity

5.5 Role of Digital Twins

5.5.1 Product Management Levels

5.5.2 Generations/Time

5.6 Application-Oriented Digital Twin Framework

6 Case Study: Smart Waste Management

6.1 External Value Creation

6.1.1 System Hierarchy Levels

6.1.2 Service Scope

6.1.3 Smartness Maturity

6.2 Internal Value Creation

6.2.1 Product Management Levels

6.2.2 Product Life Cycle

6.2.3 Generations/Time

6.3 Data Resources

6.3.1 Data Categories

6.3.2 Data Sources

6.3.3 Data Format

6.3.4 Connectivity Solutions

6.4 Instantiation of the Application-Oriented Digital Twin Framework and Conclusion of the Case Study

7 Discussion

7.1 Major Findings

7.1.1 Value Creation with Digital Twins

7.1.2 Comparison of the two Digital Twin Frameworks

7.1.3 Implications for their Application in Practice

7.2 Contribution

7.2.1 Contribution to Practice

7.2.2 Contribution to Science

7.3 Limitations

7.4 Recommendations for Further Research

8 Conclusion and Outlook

8.1 Conclusion

8.2 Outlook

9 References

10 List of figures

11 List of tables

A List of Experts Involved in Research Phase 2

B Intermediate Version of the Scientific Digital Twin Framework

C Intermediate Versions of the Application-Oriented Digital Twin Framework

D Sketch of Use Case Project Planning Using the Scientific Digital Twin Framework

E List of Publications

1 Introduction

This chapter provides an introduction to the background and context of the thesis. First, the context of digitalization with the closely intertwined topics of the internet of things (IoT) and smart connected products (SCPs) and systems is explained. This is followed by a closer look at the central topic of the digital twin (DT) concept explaining its development to date and its current status, discussing specific characteristics, and presenting a proposal for a new definition of the DT concept. Thereafter, the relevance of DTs is demonstrated and highlighted, while the following section explains the identified research gaps that this thesis aims to help fill. The chapter concludes with an overview of the structure of the thesis.

1.1 Smart Connected Products and Systems

Digitalization is neither a new phenomenon nor a hype that can be expected to vanish soon. From the outset, the centrally acknowledged benefit of digitalization has been the automatization of manual tasks to increase efficiency. Saving money and time and improving the quality of processes are all integral components of the management of any value chain. Therefore, the production industry pioneered many digitalization concepts and approaches, which later became omnipresent in consumer markets and everyday life. In the professional sphere, digital machines such as robots support our production processes and software systems support our work processes. In the private sphere, too, everyone's life is becoming increasingly digitalized and networked – smartphones, wearables, social media, connected homes, connected cars, and smart cities are just some of the developments that will increasingly shape our lives.

A key driver of ongoing digitalization is the ever-increasing availability of the internet, meaning we can communicate in real-time, even with geographically dispersed participants. In addition, more and more products or “things” are interconnected in the IoT. Today, the IoT is a network of vehicles, phones, drones, appliances, factories, and other objects embedded with sensors and connectivity that enable them to exchange and analyze data. In a 5G world, with edge computing and intelligent networks, everything becomes a “thing” and everything produces data. A report by Juniper Research estimated the “total number of IoT connections” in 2020 at 35 billion and projected an increase to 83 billion by 2024, representing a growth of 130% (Smith, 2020). According to figures and forecasts from Statista, the number and growth of IoT devices is lower than in Juniper's report, but still impressive. According to this forecast, the number of IoT devices will almost double from 15 billion in 2020 to more than 29 billion IoT devices in 2030 (Vailshery, 2023). This is dramatically transforming our private lives as well as how businesses create and capture value by redefining internal business operations and disrupting markets with new products and services for customers. As a result, communication and the exchange of data – and thus also the storage and processing of data – have become hugely important. Accordingly, the global IoT market is already a critical economic factor and future growth driver of the digital economy. A report by Mordor Intelligence forecasts that the global market for IoT, which was valued at $761.4 billion in 2020, will surpass $1.38 trillion by 2026, representing a compound annual growth rate of 10.5% (Crane, 2021).

The further development and progress in information and communication technologies will transform more traditional products into SCPs, enabling novel smart services and ultimately changing whole industries (Dawid et al., 2017; Porter & Heppelmann, 2014, 2015; Wünderlich et al., 2015). Here, the DT concept is regarded as a key technology for creating value with smart services (Barbieri et al., 2019) and for realizing smart manufacturing and industrial digital transformation (Liu et al., 2023). In their seminal article, Porter and Heppelmann (2014) described the evolution of these digitalized products in five steps, as illustrated in Figure 1. Step 1 represents a physical product, such as a tractor; in step 2, this product becomes smart – equipped with embedded systems – and in step 3, connectivity is added to create a SCP.

Fig. 1 How smart, connected products are transforming competition Source: Porter & Heppelmann, 2014.

A SCP is defined as a physical product that possesses a unique identification code, continuously monitors its status and environment, stores data about itself, deploys a language to display its features and production requirements, and is capable of participating in or making decisions relevant to its destiny, generally by interacting with other information systems (IS) and users (Kiritsis, 2011; Wuest et al., 2018). The defining components of a SCP are shown schematically in Figure 2.

Fig. 2 Composition of a smart connected product

The core of a SCP is typically a physical product such as machinery, equipment, devices, or components. In addition, a SCP has sensors to sense its state and the environment, a connectivity module for communication, and actuators to generate a reaction to the measurement results and communicated commands. Specific, often simple embedded information hardware (EIH) and software (EIS) are required to establish these capabilities in the SCP.

By integrating parallel products (such as planters or tillers), a farm equipment system with the possibility for overarching management emerges in step 4. Finally, in step 5, complementing this farm equipment system with other systems (e.g., seed optimization system), a farm management system – or more generally, a system of systems – emerges, as seen in Figure 1 (Porter & Heppelmann, 2014). These systems are often referred to as cyber-physical system (CPS), as they consist of (i) a mechanical part, which is the physical system that is manufactured to perform a function in the real world, and (ii) the virtual part that collects the dynamic status of the product across its life cycle to enable different digital services and applications along with the mechanical part lifespan (Uhlemann et al., 2017; Al-Ali et al., 2018).

Figure 3 shows two SCPs that communicate with each other and the virtual part of the CPS, enabling additional services and applications for users, thus forming a cyber-physical service system (CPSS).

Fig. 3 Composition of a cyber-physical service system

Figure 4 shows how several CPSSs can be combined according to the same logic to form a system of systems, that enables additional services and applications. Most modern CPSSs can be considered complex systems of systems, incorporating heterogeneous, distributed, and collaborating components (Nazarenko & Camarinha-Matos, 2019).

Fig. 4 Composition of a cyber-physical service system of systems

One of the key enabling and tightly interrelated technologies for CPSSs is the concept of DTs (Nazarenko & Camarinha-Matos, 2019). Central to this thesis, this topic will be discussed in detail in the following chapter.

1.2 The Digital Twin Concept

To develop a comprehensive understanding of the DT concept, the first section summarizes the origin and developments at the beginning of the concept. Following this, the current understanding and definitions of the DT concept are reviewed. The largest part of this chapter is subsequently dedicated to the elaboration of the properties and characteristics of DTs. Finally, based on this elaboration, a new definition of the DT concept is presented along the central properties and characteristics, which serves as a basis for the additional research presented in this thesis.

1.2.1 Origin

The concept of using representations of an object under investigation to perform tests and thereby gain knowledge goes back to the Apollo program of the National Aeronautics and Space Administration (NASA) (Massonet et al., 2020). In the 1960s, NASA pioneered the development of a second, identical spacecraft that remained on Earth during the mission. This “twin”, which at that time did not exist digitally but in reality, was used extensively by astronauts for training during the pre-flight phase. While one spacecraft was in space, the second spacecraft (the twin) was used to simulate alternatives using available flight data to recreate flight conditions as realistically and accurately as possible. Thus, based on the data, the NASA team could always support the astronauts in critical situations during their mission in space.

The origin of the DT concept is attributed to Michael Grieves and John Vickers of NASA, with Grieves presenting the concept in a lecture on product life cycle management (PLM) in 2003 (Grieves & Vickers, 2017). Initially, the idea was referred to as the “mirrored spaces model” and not as the “digital twin” (Grieves & Vickers, 2017). However, it already had all the essential elements of a DT, namely the physical space, the virtual space, and the connection for the data flow or information flow between these two spaces, as seen in Figure 5.

Fig. 5 Connection of real and virtual space for product life cycle management Source: Grieves & Vickers, 2017.

A DT can therefore be identified by its three main pillars, (i) a physical product in real space, (ii) a virtual product representation in virtual space, and (iii) the connection of data and information which ties together these two spaces (Kahlen et al., 2017). Following the initial description, a DT is defined as a virtual representation of a physical product, from which it receives data and to which it provides information to optimize processes. This optimization of processes includes two central characteristics of DTs, (i) bidirectional communication between real and virtual space and (ii) the data from real space are structured and interpreted in virtual space (VS) before they are returned to real space. Virtual space consists of any number of subspaces (VS1, VS2 … VSn) that enable specific virtual operations, such as modeling, testing, and optimization (Jones et al., 2020). In summary, this core concept of the DT envisaged a system that ties physical entities to virtual counterparts, leveraging the benefits of both the virtual and physical environments to the advantage of the entire system by exchanging valuable information about the products and processes (Singh et al., 2018).

In 2012, the concept of DTs was revisited by NASA, which defined a DT as a “multiphysics, multiscale, probabilistic, ultra fidelity simulation that reflects, in a timely manner, the state of a corresponding twin based on the historical data, realtime sensor data, and physical model” (Glaessgen & Stargel, 2012). This definition of a DT was referred to very often in academic publications (e.g., Barbieri et al., 2019; Detzner & Eigner, 2018; Grieves & Vickers, 2017; Landahl et al., 2018; Qi et al., 2018; Shangguan et al., 2019; Tao et al., 2018, 2019a, 2019b, Uhlenkamp et al., 2019). Even though DTs are defined as an “ultra fidelity simulation”, it is important to note that unlike computer-aided design (CAD) (that exclusively focuses on the digital world) and IoT (that focuses on the physical world), DTs were still characterized by the previously mentioned bidirectional interaction between the digital and physical worlds to create new possibilities (Glaessgen & Stargel, 2012).

1.2.2 Status of Definitions

The definition of the DT concept has evolved in the last decade, alongside its growing popularity and adoption into different industries and use cases, which seldom meet the stringent demands of a real one-to-one copy described in NASA's definition. Many publications adopted the DT concept with some deviations from the original definition, albeit avoiding defining the DT concept explicitly themselves (Sjarov et al., 2020). Instead, they implicitly assume a particular set of abilities and properties, thus hindering the formation of an accurate definition. The nature of a DT is described in many ways, for example, as a multi-domain simulation (Jaensch et al., 2018), a computerized counterpart of a physical system (Kritzinger et al., 2018), a virtual representation of what has been produced (Grieves, 2014), a virtual substitute of real-world objects (Schluse et al., 2018), an integrated simulation and forecasting tool (Negri et al., 2017), or a linked collection of digital artifacts (Boschert et al., 2018). In summary, many definitions up until 2019 defined a DT as a virtual representation of a physical product, asset, process, or system in a CPS, and across its life cycle, capable of mirroring in realtime its static and dynamic characteristics as a result of a seamless data transmission between its digital replica and physical entity (i.e., Ashtari Talkhestani et al., 2019).

However, there was still no common understanding of the term “digital twin” (Cimino et al., 2019), as the various definitions and concepts depend strongly on the respective application context (Schleich et al., 2017). This is illustrated through the frequent entanglement of DT descriptions with specific industries, such as manufacturing (Enders & Hoßbach, 2019). For example, the Encyclopedia of Production Engineering has also published its definition of DT:

A digital twin is a digital representation of an active unique product (real device, object, machine, service, or intangible asset) or unique product-service system (a system consisting of a product and a related service) that comprises its selected characteristics, properties, conditions, and behaviors by means of models, information, and data within a single or even across multiple life cycle phases (Stark & Damerau, 2019, p. 1).

Owing to this proliferation of similar but different definitions, an increasing number of attempts have been made since 2020 to clarify what constitutes a DT and how it should be generally defined, namely independent of industries or applications (e.g., Barth et al., 2020; Jones et al., 2020; Minerva et al., 2020). To consolidate the findings, researchers began conducting meta-reviews of selected reviews in the research field of DTs (e.g., Kuehner et al., 2021; Schweiger & Barth, 2023). The aim is to satisfy the need for a definition generic enough to be used in various application areas and at the same time able to extend the differing partial views of the DT concept (Golovatchev, 2020). In the following, the reviews that took a systematic approach to both the selection of sources and their analysis are briefly summarized.

In their 2019 study, Barricelli et al. examined state-of-the-art DT definitions, assessed essential characteristics a DT should have, and investigated domains where DT applications are currently being created. Jones et al. (2020) present findings regarding the conceptual status, key terminology, and related processes of DTs to define the characteristics, a framework, and processes of operation. Sjarov et al. (2020) assorted DT related concepts such as product avatar and digital shadow by systematically examining DT definitions. They also present and discuss prominent DT models and thereof derived DT purposes from other publications.

Van der Valk et al. (2020) conducted an extensive literature review on the scope and meaning of the term “digital twin” to create a taxonomy according to the taxonomy development method introduced by Nickerson et al. (2013). Their comprehensive analysis distilled eight key dimensions and the most common characteristics of DTs from 233 papers, as seen in Table 1.

Tab. 1 Final taxonomy of digital twins with applicated definitions given by Glaessgen and Stargel (2012), Grieves (2014, 2017), and Tao et al. (2018)

Source: Van der Valk et al., 2020.

The number in brackets for each characteristic indicates their numerical distribution alongside the papers. Three widely used definitions of DTs given by Glaessgen and Stargel (2012), Grieves (2014), Grieves and Vickers (2017), respectively, and Tao et al. (2018) are depicted in color in the taxonomy. The taxonomy shows that the term and scope can differ significantly depending on the use case or goal of the DT. However, while the taxonomy is very systematic in showing how DTs have been defined to date, it does not offer a future, generally applicable definition. For example, DTs are more often created after their physical counterparts, although influential authors such as Grieves and Vickers (2017) and Boschert and Rosen (2016) stress that a DT should be developed before its physical counterpart.

Based on a comparison of the DT concept with building information modeling and CPSs, Jiang et al. (2021) propose their own definition of DTs. Furthermore, they cluster the DT research in the field of civil engineering. Kuehner et al. (2021) detect prevalent and contrasting views to clarify commonalities in terminology, conceivable benefits, and remaining research issues by comparing existing reviews of DTs in a meta-review. The review of Semeraro et al. (2021) analyzed the concept, the life cycle, and the primary functions of DTs at different stages to answer the questions of what a DT is, when it should be developed, why it should be used, how to design and implement it, and what the main challenges of implementation are. Tomczyk and van der Valk (2022) analyzed the journey of the DT definition and found a paradigm shift from the classic three-dimensional definition, with the physical and the virtual space and their bidirectional connection as dimensions, to an expanded five-dimensional definition, with data and services as additional dimensions.

In summary, these studies examine how the DT concept has been defined to date but do not offer a forward-looking, application-oriented definition. Either they focus on comparing and summarizing existing concepts and do not provide their own definition (Barricelli et al., 2019; Sjarov et al., 2020; Jones et al. 2020; van der Valk, 2020), or if they propose new definitions, they are too narrow regarding implementation in practice, for example, by only including physical counterparts or demanding real-time connectivity (Jiang et al., 2021; Kuehner et al., 2021; Semeraro et al., 2021).

1.2.3 Properties and Characteristics of Digital Twins

As shown in the previous subchapter, academic researchers have long debated a consolidated understanding of DTs without arriving at a generally accepted and unified definition (Tomczyk & van der Valk, 2022), as the understanding of the term varies between applications (Kühner et al., 2021). As a consequence, the discussion has been placed on increasingly granular components of DTs in recent years (e.g., Minerva et al., 2020). This led to the current approach which is to arrive at a generally accepted overall definition by discussing and defining DTs’ individual properties and characteristics. Schweiger and Barth (2023) conducted a meta-review of existing reviews and discussed the defining properties and characteristics of DTs to research the definitions of DTs used in the industry. Based on their meta-review of existing reviews and publications, their experience from projects, and definitions encountered in the industry they proposed an application-oriented taxonomy and a corresponding definition of DTs. This subchapter presents first their taxonomy and then discusses the included properties and characteristics of DTs. The proposal of a corresponding definition of DTs is presented in the following subchapter.

The taxonomy developed by Schweiger and Barth (2023) is shown in Table 2 and contains, next to the DT-defining properties and characteristics, an additional column indicating whether the characteristics are exclusive or not. It systematically represents ten properties with two to six characteristics and thus enables companies to compile a DT concept with the appropriate level of sophistication for their use case in the sense of a morphological box. The properties and their characteristics are explained and substantiated in the following sections.

Tab. 2 Taxonomy with properties and characteristics of digital twins

Properties

Characteristics

Exclusivity

Counterpart

Only physical

Also non-physical

Any distinct entity

Mutual

Data sources

Sensors

Internal systems

External systems

Not

Data link

Bidirectional

Unidirectional

Mutual

Interface

Human-to-machine

Machine-to-machine

Not

Fidelity

One-to-one

Sufficient

Mutual

Synchronization

Real-time

Near-real-time

Periodic

Not

Capabilities

Simulation

Optimization

Prediction

Detection

Prevention

Automation

Not

Purpose

Performance

Availability

Quality

Not

Life cycle

Beginning of life

Middle of life

End of life

Not

Creation

Independent creation of DT

Types / instances distinguished

Not

Source: Schweiger & Barth, 2023.

Counterpart

Most authors agree that the twin counterpart should at least cover physical products and components (Sjarov et al., 2020). And even the researchers who focus on physical products often at least mention intangible counterparts. For example, Stark and Damerau (2019) mention that a DT has a physical twin throughout its life cycle, but also consider directly related services as well. And although Jones et al. (2020) do not explicitly include non-physical entities to be twinned by DTs, they acknowledge that more abstract entities, such as supply chains are also twinned. Generally regarded, this means systems, subsystems, and systems of systems, which also include the corresponding processes. Minerva et al. (2020) also state that one of the entities connected by a DT is relevant in the real world and usually physical. However, they explicitly mention that software and intangible entities such as processes and activities can also be represented with DTs. Malakuti et al. (2019) understand DTs as an evolving digital profile of the historical and current behavior and all properties of an asset – where an “asset can be anything of value for an organization”, such as a physical device, subsystem, plant, or software entity. Boss et al. (2020) agree with this view when arguing that a DT can represent anything in the real world of interest to an application as long as this real-world counterpart can be defined as an item with a recognizably distinct existence. This includes any (non-physical) logical objects, such as an organization or a production process. In principle, a DT can represent anything in the real world of interest to an application (Boss et al., 2020). Some approaches even propose DTs of human workers (Nikolakis et al., 2019) or business services (Madni et al., 2019). Hence any real-world entities with (i) a recognizably distinct existence and (ii) relevance for creating value can be digitally represented by a DT.

Data Sources

An important feature of a DT is the ability to collect, store, and represent all relevant present and past data of its real-world counterpart (Minerva et al., 2020). Consolidating this data in a DT lays the foundation for the data-based innovation of services and processes. A DT can collect data from various sources, such as onboard product systems, internal enterprise systems, and third-party sources (Barth et al., 2020). Onboard product systems in IoT devices enable the collection of sensor data from various subcomponents of physical assets and edge devices (Sharma et al., 2022). Even in the case of non-physical entities there are usually SCPs with sensors and embedded systems that send relevant data about the nonphysical entities, such as processes, from the real world to the DT. According to Detzner and Eigner (2018) redundant data and potentially conflicting information from heterogeneous, non-embedded systems can be minimized by utilizing a DT as a single source of truth for instance related data. These non-embedded systems can be divided into internal and external IS connected to the DT. Most DTs use data from internal IS such as authoring, product data management (PDM), enterprise resource planning (ERP), and customer relationship management (CRM) systems (Holler et al., 2016). Additionally, DTs can use data from external systems, such as other companies’ interconnected systems and third-party data provider offering data via application programming interfaces (API) or IoT platforms.

Data Link

In order to use data from these different sources, DTs need data links to access them. The basic idea behind the concept of DTs is that they are able to send the processed data back to their real-world counterparts as well. This requires bidirectional connections between the DT and its real-world counterpart, which are even considered a mandatory part of a DT by influential authors, such as Grieves (2014) and Tao et al. (2018). Kuehner et al. (2021) also noted that a bidirectional connection is one of only four predominant definition elements in DT literature. That a DT must, by definition, have a bidirectional data link is also stated by van der Valk et al. (2022), who take the concept a step further when mentioning that DTs are embedded into networks with many bidirectional data links and, therefore, the term multi-directional data links would be more precise.

However, if analyses are considered which examined DT applications or whose concepts are oriented towards application in practice, a more differentiated picture emerges. For example, Enders and Hoßbach (2019) identified three types of connections in self-proclaimed DT applications in their review and just over a quarter had a bidirectional data link (amount in parenthesis): no connection (23), one-directional connection (39), and bidirectional connection (25). Kritzinger et al. (2018) made an influential argument regarding data links in the sense of the level of integration between physical and digital. They propose to distinguish three subcategories of DTs depending on whether the connections between the objects are manual or automatic: digital model (both manual), digital shadow (automatic from physical to digital), and DT (both automatic). In their review, they also noted that the term DT is often used interchangeably with the other terms, with only eight of 43 publications describing DTs that explicitly have automatic connections in both directions.

It follows, that from a practice-oriented viewpoint it is also appropriate to speak of a DT if not both connections are strictly digital and automated. Research consequently started to elaborate a more differentiated view, investigating not only the mere presence of data links but also how they are implemented. For example, the model of Stark and Damerau (2019) includes a dimension called connectivity mode that specifies unidirectional and bidirectional connectivity as well as a third level of automatic, context-aware, and self-directed communication capability. Such a view is also supported by the review of Jones et al. (2020), who have stated that conceptually it is possible to generate a DT with just a one-way physical-to-virtual connection and that the role of human in-the-loop is not frequently discussed in literature. To illustrate, they explain a brief example in which a human technician is sent out by a DT to carry out a maintenance task determined by a predictive model. The elaborations of Minerva et al. (2020) regarding the entanglement of DTs with the real world seem to go into a similar direction, as they also acknowledge unidirectional and bidirectional connections between the DT and the real-world counterpart. They furthermore argue that this connection can be direct or indirect, with the two communicating objects relying on a third party to send and receive information in the indirect case.

It can be concluded that a DT must in any case have a bidirectional connection to its real counterpart. Regarding the implementation of DTs, it may be sufficient that this connection is digital and automated in only one direction (from the real counterpart to the DT) and that the processed data (in the form of information or decisions) are not fed back directly via the DT, but manually via human actors.

Interfaces

The interface property as defined by Schweiger and Barth (2023) classifies the gateways that can be used to access the data and information provided by a DT. Following on from the elaborations regarding data links, two possible types of interfaces can be distinguished: the DT communicates directly with the real-world counterpart or other DTs via machine-to-machine (M2M) interfaces or with human users via human-to-machine (H2M) interfaces (van der Valk et al., 2022).