Defining the Digital Twin for Industry 4.0
Abstract
This manuscript offers a comprehensive review of Digital Twin (DT) technology, first presenting the proposal of our novel DT amendment that looks to refine the conceptual understanding of DTs across various industrial use-cases. The study synthesizes findings from an extensive literature review, emphasizing the need for precise use-case-centric definitions and categorization to advance the field effectively. Chapter 1 introduces the current state of DT research, advocating for a more nuanced approach to the identification, classification, and validation of DT systems. It challenges the broad definitions prevalent in current literature and proposes a more application-specific research trajectory.
Subsequent chapters build on this foundation by curating an extensive review of Supervisory Control And Data Acquisition (SCADA) systems. This explanation illustrates that industry's current definition of DT follows suit with the capabilities of such technology when an automatic, bi-directional data transfer is required for simulation models to be classified as DTs. After this, we explore the application of DT technology in dataset creation for the training of Intrusion Detection Systems (IDS) at the industrial application layer. Lastly, we propose a novel validation method for DTs by utilizing machine learning techniques while identifying key challenges such as system complexity and the need for diverse similarity measures.
This manuscript brings to question the current literature's tendency to create self-validating information silos and highlights the discrepancy in validation approaches for different systems. It acknowledges the infancy of practical applications and calls for a standardized validation framework.