Foundation for Digital Twins

ROBERTO MINERVA is an Associate Professor with the Service Architecture Laboratory, Institut Mines Telecom—Telecom Sud Paris, Institute Polytechnique de Paris, France. From 2016 to 2018, he was the Technical Project Leader of SoftFIRE, a European Project devoted to the experimentation of NFV, SDN, and edge computing. He was Chairperson of the IEEE IoT Initiative from 2014 to 2016. NOEL CRESPI is a Professor and MSc Programme Director, leading the Data Intelligence and Communication Engineering laboratory (DICE) at the Institut Mines Telecom—Telecom Sud Paris, Institute Polytechnique de Paris, France, where he has been since 2002. He coordinates the standardisation activities for Institut Mines-Telecom at ETSI, 3GPP and ITU-T.

• Foreword xiiiAbout the Authors xvPreface xviiAcknowledgments xixAcronyms xxiIntroduction xxv1 What Is a Digital Twin? 11.1 Introductory Concepts 11.1.1 Basic Definitions 11.1.2 Formalized Definitions 31.2 DT: A Rationale 41.2.1 Models and Modeling 41.2.2 DT as a Combination of Models 51.3 DT Definition: A Step Further 71.3.1 DT Properties 81.4 DT Representation, Model, and More 121.5 The Software Part of a DT 171.5.1 d Architecture and development Approach 171.5.2 An Introductory DT Architecture 181.6 Specification Methodologies 201.6.1 Designing a DT Based on Available Specifications of a PS 201.6.2 Designing a DT Without Previous System Specifications 211.6.3 Understanding the Physical System Behavior 211.7 Domain Knowledge and Operation in DTs 221.7.1 The Role of Domain-specific Knowledge and Technologies 221.7.2 Operational Context: Bridging IT and OT Domains 231.8 Clarifying Scope and Terminology for DT Architectures 231.8.1 d Architecture and Scope definition 241.8.2 Evolving Terminology in DT Research 252 Models and Modeling Aspects of a DT 272.1 The Representational Aspects of a DT 272.1.1 The Importance of Environment Representation 282.1.2 Physical System Life cycle and Development of DT 292.1.3 Modeling Aspects and Their Relationships 302.1.4 Cascade and Parallel Life cycles 312.2 Data Modeling 342.2.1 Data and Behavior Modeling 352.2.2 Data Models as Enablers of Passive DTs 382.3 Behavior Modeling 392.4 Predictive Models 412.5 Prognosis Model 432.6 Prescriptive Model 462.7 A Flexible Approach to Multi-facet Models 482.7.1 Relationships Between Different Models 482.7.2 Models and Design, the Complexity of DT Modeling 512.7.3 Modeling the Environment 522.7.4 Context and Situation 522.7.5 From Top or Bottom? 533 Foundational Data for a Digital Twin 553.1 Descriptive Models 553.2 Data-driven Architecture 583.3 Which Data to Include in a DT 603.4 Organizing Data into Data Models 623.4.1 Minimal DT Structure and Multiple Types of Data 623.4.2 Data Representations for the DT 643.4.3 Data Models and APIs 663.5 Data Models for a DT 693.5.1 Data Modeling Techniques and Semantic Enrichment in DTs 693.5.2 Existing Data Models for DTs 713.6 Extending Data Models to Fit the Stakeholders’ Needs 743.6.1 Stakeholder 1: Building Manager 743.7 Relationships Between Descriptive and Behavior Models 763.7.1 Data Exploitation Chain: From Passive to Prescriptive DT 774 Digital Twin as a Behavior Model of a Physical System 854.1 Behavior Modeling 854.1.1 An Example: The Traffic Light System Behavior 864.1.2 Behavior Modeling Challenges and Needs 884.2 Foundational Behavior Models 914.2.1 Types of Behavior Models 924.2.2 Functions of Behavior Models in a DT 924.2.3 Supporting Behavior Models with NGSI-LD 934.2.4 In the Context of Traffic Light Controller Example 934.3 Developing a Behavior Model 944.3.1 Implementing the Twin as a State Management Component 954.3.2 Environment and DT Structure 974.4 AI and Digital Twins 1014.4.1 Iterative Refinement of DT Behavior Models Using Generative AI 1044.4.2 Extraction of Behavior Rules from Data 1074.4.3 Enhancing AI Explainability in DTs Through Behavior Models 1094.4.4 Autonomic DTs and Agentic AI 1114.5 Simulation Models, Behavior Validation, and GenAI Integration in DTs 1124.5.1 Applicable Simulation Models 1124.5.2 MetaModel for Unified Execution and Simulation 1134.5.3 Evolving the Traffic Light DT with Situation-aware Behavior Modeling 1154.6 Behavior Models for Prognosis and Prescriptive DTs 1174.6.1 Behavior Models for Prognostic DTs 1174.6.2 Prescriptive Model and Integration with Prognosis Capabilities 1214.6.3 Recommendation Systems in DTs 1224.6.4 The Role of Autonomics in Prescriptive DTs 1224.6.5 Benefits of Behavior Models in Prescriptive DTs 1244.7 Behavior Modeling Recap 1255 Architecting and Implementing Digital Twin Systems: Approaches, Guidelines, and Best Practices 1275.1 Requirements, Structural Concepts, and Terminology for a Digital Twin Architecture 1275.1.1 Requirements and Properties of DTs 1285.1.2 State of the Art in DT Architectures 1285.1.3 Reimagining a Versatile DT Architecture 1295.2 Software Interaction Paradigms for DT Implementation 1325.2.1 Client-server Paradigm 1335.2.2 Event-driven Paradigm 1335.2.3 Agent-based Paradigm 1335.2.4 Data-sharing Paradigm 1335.3 Componentization of DT Architecture 1355.3.1 System Engine 1355.3.2 Data Management 1365.3.3 DT Engine 1375.3.4 DT Life-cycle Management 1415.4 Model First 1445.4.1 Alignment with the Reference Architecture 1455.5 From Modules to Components and Microservices: A DT Perspective 1495.5.1 Rationale for Further Decomposition: DT Specifics 1505.5.2 Microservice Design Representation 1515.5.3 Rationale for Microservice Decomposition: DT Advantages 1545.5.4 Preparing the Components for Deployment 1555.5.5 Comparison of the Architecture with ISO 23247 Standard 1575.6 Testing and Validation of DTs 1585.6.1 Importance of Testing and Validation 1595.6.2 Continuous Impact Within the Enterprise 1595.6.3 Techniques and Approaches for Testing and Validation 1595.6.4 Value of a Validated DT 1606 Deploying and Operating a Digital Twin 1616.1 Distribution of DT Components and Functions 1616.1.1 Centralized or Distributed DT 1616.1.2 A Deployment Scenario 1646.2 Operating the DT 1656.2.1 Product and DT Life-cycle Management and Phase Transitions 1656.2.2 Management of DT Functionalities 1676.2.3 AI for DT Management and Operation 1696.2.4 Data Management for the DT 1706.2.5 Management of the System Infrastructure 1726.2.6 Additional Relevant Topics 1736.3 Example: Traffic Light Service DT 1736.3.1 Deployment Mapping 1746.3.2 Operational Workflow 1756.3.3 Benefits 1766.4 DT Impact on Organization Processes 1776.4.1 Breaking Down Information Silos 1776.4.2 Cross-departmental Cooperation and Commitment 1776.4.3 Integration of Diverse Technologies 1776.4.4 Organizational Change and Digital Maturity 1786.4.5 Additional Organizational Challenges 1786.5 Life-cycle Insights from Industrial Experiences 1796.5.1 Life-cycle Phases 1796.5.2 Exemplary Industrial DT Projects 1816.5.3 Enablers and Barriers to Industrial Exploitation 1826.5.4 Advantages and Enterprise Effort 1847 Some Examples of Applicability of the Digital Twin Architecture 1857.1 Introduction 1857.2 Developing DTs for Smart Cities: A Bottom-up Approach 1867.2.1 From Simple DTs to Specialized Behavior 1867.2.2 Microservices and Component Flexibility 1897.2.3 Integrating Heterogeneous Data Streams for Holistic Urban Insights 1897.3 NDT: The Edge-cloud Continuum Representation 1907.3.1 A Top-down Approach for NDT Design 1907.3.2 NDT Template: Monitoring and Optimization 1917.3.3 Stakeholder Views and Insights 1917.3.4 NDT Architecture Overview 1927.3.5 Optimization and Prognosis: Example Workflows 1927.4 The Challenge of DTs for Cultural Heritage 1947.4.1 A Hybrid Approach: Bottom-up and Top-down 1957.4.2 Multi-view D for Artifacts 1957.4.3 Web of Related DTs 1967.4.4 Personalized and Adaptive Experiences 1977.4.5 Case Study: Egyptian Scarabs and the Power of DTs 1977.4.6 Context DT Architecture Support for Cultural Heritage 1987.5 DT as an Integral Part of the Metaverse 1997.5.1 State of the Art of Metaverse Platforms and Mapping to the Context Digital Architecture 2007.5.2 Integration Points and Mutual Enhancement 2017.5.3 Use Cases: Education, Tourism, and Factory Management 2027.6 Implementing Services with the Envisaged Architecture 2058 The Digital Twin of the Future 2078.1 The Evolution of DT 2078.2 Evaluating DTs as a General Solution 2098.3 Promising Application Domains 2108.3.1 Initial Criteria for DT Suitability 2118.4 Anticipated Evolution of DT Technologies 2118.4.1 DT Platforms 2128.4.2 DT Creation and Development 2138.4.3 Decomposition of Modules 2158.4.4 Extensible Architecture 2168.4.5 Integrated Methodologies 2168.4.6 AI Integration 2178.4.7 Testing and Assessment 2198.5 Improved Operations 2198.5.1 Life-cycle Management 2208.5.2 Managing the Switch of States 2208.5.3 Operations Tools 2218.6 Interoperability and Standards 2218.6.1 Standardization Efforts 2228.6.2 Open APIs and Data Models 2228.7 Ethical, Privacy, and Trust Considerations in DT Systems 2238.8 Human-in-the-loop and User Experience 2248.8.1 User-centric Design 2258.8.2 Visualization and Immersive Interfaces 2258.9 Sustainability and Societal Impact 2258.9.1 DTs for Sustainable Development 2258.9.2 Societal Impact and Digital Inclusion 2268.10 Future Steps 226A Listings and Details 229A.1 Descriptive Modeling 229A.1.1 Data Model and Application Programming Interfaces 229A.1.2 Extending the Data Model to Fit Stakeholder Needs 235A.2 Behavior Modeling 236A.2.1 NGSI-LD Data Model for Traffic Light Behavior 236References 239Index 267