Digital Twin Technologies In Transition: A PRISMA-Guided Analysis Of Architectures, Applications

Abstract

Digital twin technology refers to the use of synchronised digital replication for the modelling, management, and enhancement of large-scale cyber-physical systems. As such, digital twin systems have been simplified (in terms of integration) using artificial intelligence (AI), cloud computing, and data-driven analytics. Unfortunately, rapid technological progress has led to many unresolved issues regarding the technology, methods, and structures of digital twins. A systematic literature review of the relevant digital twin literature (2023-2026), conducted using the PRISMA methodology, served as the basis for understanding advances through analytical overviews of the findings. After a thorough review of all relevant peer-reviewed research, including quality assessments and a comparison of articles by their main goals, we found that digital twin systems, which were previously considered fixed, are now being replaced by AI-driven analytics and cloud-edge computing systems. These new systems are modular, layered, and/or distributed.  Such systems, which currently exist in the healthcare, energy, digital infrastructure and industrial sectors, each have their own rules for manufacture and operation. Research into AI-augmented digital twins is underway to determine the extent to which they can explain, verify, synchronise, or enable cross-platform interoperability and the management of large datasets. While the results of these studies are like those of earlier research, it's important for researchers to also consider standard design frameworks, useful analytics, and strict measurement methods. Moreover, the development of a modern, technically sound synthesis has greatly improved the creation, testing, and use of advanced digital twins.