Systems of Digital Twins: A Systematic Literature Review and a formal framework

Abstract

The adoption of Digital Twins (DTs) in complex domains has motivated the formalization of systems of DTs to model complex physical entities with system-level properties. The current literature is characterized by a non-standardized and, in some cases, inconsistent use of definitions associated with systems of DTs, hindering the comparability of existing approaches and limiting the principled selection of architectures with respect to both the characteristics of the represented physical entity and the intended purpose. This work investigates systems of DTs through a systematic-narrative hybrid literature review. The systematic phase queried Scopus and, following PRISMA guidelines, resulted in 62 studies included in the final analysis. The review organizes the analyzed contributions by systematically extracting architectural and terminological features and by characterizing the main typologies of systems of DTs, namely Digital Twin Aggregate, Digital Twin Network, Federated Digital Twin, Composite Digital Twin, and Digital Twin Ecosystem. Based on the results of the literature analysis, the paper proposes a formal framework for the DT of a system of DTs, referred to as the Global DT (GDT). The framework introduces a purpose-oriented representation, a projection operator defining admissibility, and a model-based equivalence relation that enables the integration of heterogeneous DTs. A conceptual example involving virtual clinical trials is provided for illustration.