Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Learning Paradigms and Modelling Methodologies for Digital Twins in Process Industry

Michael Mayr, Georgios C. Chasparis, Josef Küng

TL;DR

The paper addresses how Digital Twins in the process industry are built using diverse modelling methodologies and learning paradigms. It systematically reviews literature to map modelling methods (e.g., CNNs, AEs, PINNs), learning strategies (supervised, unsupervised, self-supervised, transfer learning), and task categories (classification, regression, clustering). The findings show dominance of CNN/AE approaches and supervised learning, with growing interest in hybrid physics-data models; self-supervised and transfer learning remain underexplored yet promising, and transformer-like architectures are not yet prevalent in industrial DTs. The study highlights the potential of pretraining and cross-task transfer to enable scalable, generalizable DTs in high-volume IIoT environments and points to future research directions and industry adoption.

Abstract

Central to the digital transformation of the process industry are Digital Twins (DTs), virtual replicas of physical manufacturing systems that combine sensor data with sophisticated data-based or physics-based models, or a combination thereof, to tackle a variety of industrial-relevant tasks like process monitoring, predictive control or decision support. The backbone of a DT, i.e. the concrete modelling methodologies and architectural frameworks supporting these models, are complex, diverse and evolve fast, necessitating a thorough understanding of the latest state-of-the-art methods and trends to stay on top of a highly competitive market. From a research perspective, despite the high research interest in reviewing various aspects of DTs, structured literature reports specifically focusing on unravelling the utilized learning paradigms (e.g. self-supervised learning) for DT-creation in the process industry are a novel contribution in this field. This study aims to address these gaps by (1) systematically analyzing the modelling methodologies (e.g. Convolutional Neural Network, Encoder-Decoder, Hidden Markov Model) and paradigms (e.g. data-driven, physics-based, hybrid) used for DT-creation; (2) assessing the utilized learning strategies (e.g. supervised, unsupervised, self-supervised); (3) analyzing the type of modelling task (e.g. regression, classification, clustering); and (4) identifying the challenges and research gaps, as well as, discuss potential resolutions provided.

Learning Paradigms and Modelling Methodologies for Digital Twins in Process Industry

TL;DR

The paper addresses how Digital Twins in the process industry are built using diverse modelling methodologies and learning paradigms. It systematically reviews literature to map modelling methods (e.g., CNNs, AEs, PINNs), learning strategies (supervised, unsupervised, self-supervised, transfer learning), and task categories (classification, regression, clustering). The findings show dominance of CNN/AE approaches and supervised learning, with growing interest in hybrid physics-data models; self-supervised and transfer learning remain underexplored yet promising, and transformer-like architectures are not yet prevalent in industrial DTs. The study highlights the potential of pretraining and cross-task transfer to enable scalable, generalizable DTs in high-volume IIoT environments and points to future research directions and industry adoption.

Abstract

Central to the digital transformation of the process industry are Digital Twins (DTs), virtual replicas of physical manufacturing systems that combine sensor data with sophisticated data-based or physics-based models, or a combination thereof, to tackle a variety of industrial-relevant tasks like process monitoring, predictive control or decision support. The backbone of a DT, i.e. the concrete modelling methodologies and architectural frameworks supporting these models, are complex, diverse and evolve fast, necessitating a thorough understanding of the latest state-of-the-art methods and trends to stay on top of a highly competitive market. From a research perspective, despite the high research interest in reviewing various aspects of DTs, structured literature reports specifically focusing on unravelling the utilized learning paradigms (e.g. self-supervised learning) for DT-creation in the process industry are a novel contribution in this field. This study aims to address these gaps by (1) systematically analyzing the modelling methodologies (e.g. Convolutional Neural Network, Encoder-Decoder, Hidden Markov Model) and paradigms (e.g. data-driven, physics-based, hybrid) used for DT-creation; (2) assessing the utilized learning strategies (e.g. supervised, unsupervised, self-supervised); (3) analyzing the type of modelling task (e.g. regression, classification, clustering); and (4) identifying the challenges and research gaps, as well as, discuss potential resolutions provided.
Paper Structure (17 sections, 3 figures, 6 tables)

This paper contains 17 sections, 3 figures, 6 tables.

Table of Contents

  1. Introduction and Motivation
  2. Research Questions (RQs)
  3. Structure of Review
  4. Literature Search Strategy
  5. Quality Assessment Checks
  6. Selection of Primary Studies
  7. Data Synthesis and Analysis Approach
  8. Reporting the Review
  9. Overview of All Studies
  10. Overview of All Primary Studies
  11. Evaluating the Research Questions
  12. Research Question 1:
  13. Research Question 2:
  14. Research Question 3:
  15. Discussion and Conclusion
  16. ...and 2 more sections

Figures (3)

  • Figure 1: (Left) The number of publications matching the search query over time. (Right) The number of publications matching the search query and passing the selection criteria over time.
  • Figure 2: Proportions of tasks, types and learning paradigms for the analyzed primary studies.
  • Figure 3: Temporal development of the different modelling methodologies for DTs.