Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Systematic Mapping Study of Digital Twins for Diagnosis in Transportation

Liliana Marie Prikler, Franz Wotawa

TL;DR

The paper addresses diagnosing faults in transportation systems using digital twins, a domain with heterogeneous definitions and applications. It conducts a systematic mapping following Kitchenham guidelines to survey digital twins across vehicles, components, and infrastructure, focusing on diagnostic use, methods, and storage. It finds that most work centers on monitoring and fault detection—driven largely by AI—with few studies on fault localization or mitigation, and highlights gaps in diagnostic reasoning, data storage practices, and security/privacy considerations. The study underscores the need for more rigorous, environmentally conscious, and secure diagnostic reasoning in digital twins for transportation, offering a foundation for targeted future research.

Abstract

In recent years, digital twins have been proposed and implemented in various fields with potential applications ranging from prototyping to maintenance. Going forward, they are to enable numerous efficient and sustainable technologies, among them autonomous cars. However, despite a large body of research in many fields, academics have yet to agree on what exactly a digital twin is -- and as a result, what its capabilities and limitations might be. To further our understanding, we explore the capabilities of digital twins concerning diagnosis in the field of transportation. We conduct a systematic mapping study including digital twins of vehicles and their components, as well as transportation infrastructure. We discovered that few papers on digital twins describe any diagnostic process. Furthermore, most existing approaches appear limited to system monitoring or fault detection. These findings suggest that we need more research for diagnostic reasoning utilizing digital twins.

A Systematic Mapping Study of Digital Twins for Diagnosis in Transportation

TL;DR

The paper addresses diagnosing faults in transportation systems using digital twins, a domain with heterogeneous definitions and applications. It conducts a systematic mapping following Kitchenham guidelines to survey digital twins across vehicles, components, and infrastructure, focusing on diagnostic use, methods, and storage. It finds that most work centers on monitoring and fault detection—driven largely by AI—with few studies on fault localization or mitigation, and highlights gaps in diagnostic reasoning, data storage practices, and security/privacy considerations. The study underscores the need for more rigorous, environmentally conscious, and secure diagnostic reasoning in digital twins for transportation, offering a foundation for targeted future research.

Abstract

In recent years, digital twins have been proposed and implemented in various fields with potential applications ranging from prototyping to maintenance. Going forward, they are to enable numerous efficient and sustainable technologies, among them autonomous cars. However, despite a large body of research in many fields, academics have yet to agree on what exactly a digital twin is -- and as a result, what its capabilities and limitations might be. To further our understanding, we explore the capabilities of digital twins concerning diagnosis in the field of transportation. We conduct a systematic mapping study including digital twins of vehicles and their components, as well as transportation infrastructure. We discovered that few papers on digital twins describe any diagnostic process. Furthermore, most existing approaches appear limited to system monitoring or fault detection. These findings suggest that we need more research for diagnostic reasoning utilizing digital twins.
Paper Structure (11 sections, 3 figures, 6 tables)

This paper contains 11 sections, 3 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Search strategy
  4. Study selection
  5. Data extraction
  6. Data synthesis
  7. Limitations
  8. Diagnostic Applications of Digital Twins
  9. Types of Digital Twins
  10. Trends and Issues
  11. Conclusion

Figures (3)

  • Figure 1: Number of papers (post-merge) by year. We shortened the time axis to the span 2017-2022, as no paper published before 2017 matched the inclusion criteria.
  • Figure 2: Models of digital twins split by application area.
  • Figure 3: Methods used to perform diagnosis on digital twins. Note that "artificial intelligence" subsumes machine learning, but also contains other forms of artificial intelligence (for example, rule-based systems). Some machine learning models additionally count towards signal processing by their usage.