Conceptual and formal models for design, adaptation, and control of digital twins in supply chain ecosystems

• Conceptual and formal models of digital twins in the supply chain are proposed. • Notion of digital supply chain ecosystem is defined. • We offer a dynamic framework for digital twins using control theory. • We introduce object-driven and data-driven approaches to build and adapt digital twins • We outline a generalized framework of the cyber-physical supply chain • Application scenarios are considered The design and adaptation of digital twins in supply chains are of high relevance for academia and industry alike. While numerous prototype-based use cases have been reported, the literature lacks studies revealing generalizable methodological principles. This paper elaborates on conceptual and formal models of digital twins in the supply chain. First, we define a new notion named digital supply chain ecosystem extending the recently developed intelligent digital twin framework. A digital ecosystem is a set of digital technologies, AI-based knowledge management systems, cloud spaces, and platforms that encapsulate supply chain data enabling digital twins and simulation models. Second, we elaborate on a digital twin as a complex phenomenon comprising systems, technological-organizational models, and management decision-making support perspectives. We offer a dynamic, quantitative framework for digital twins as a decision-making support and modeling environment using control theory. Third, we introduce two views of building and adapting digital twins, i.e., object-driven and data-driven approaches. Their principle schemes are defined and discussed. Finally, we outline a generalized framework of the cyber-physical supply chain comprised of a digital ecosystem, digital twin, human-AI collaboration space, and the physical supply chain. Application scenarios are considered, e.g., using digital twins for stress testing of supply chain resilience in the setting of tariff-driven shocks as well as building resilient and viable agricultural ecosystems.