A unified digital twin approach incorporating virtual, physical, and prescriptive analytical components to support adaptive real-time decision-making

Abstract

Based on an overview of the historical and rapidly expanding literature on digital twins, this paper identifies fundamental capabilities that outline a general and adaptable twin that supports system development, real-time interactions, prescribing courses of action, and actualizing them. We relate these capabilities to business analytics concepts and decision-making processes geared toward rapid adaptation to changing situations. This leads to a general digital twin architecture supporting a system throughout its lifecycle implemented with components including Internet of Things (IoT) devices, a virtual reality environment, network communications, and an analytic simulation. The success of this architecture revolves around an authoritative data source, the Highly Integrated Virtual Environment (HIVE). We demonstrate the architecture and capabilities through a transporter system example. This demonstration highlights important timing and synchronization questions critical to fulfilling the twin’s fundamental role of reacting to evolving real-world conditions. It identifies the importance of lags in decisions, relates it to prescriptive response time and the rate of evolution of the underlying system, and quantifies this impact with new metrics of effectiveness lag and relevancy decay.