Modeling and simulation of a multi-level digital twin ecosystem for the circular economy

Circular economy initiatives require new digital tools capable of translating lifecycle complexity into decisions. This paper presents a conceptual digital twin system that integrates products, fleets, and processes within a unified simulation framework. Structured through the data–model–service paradigm and guided by model-based systems engineering, the approach connects stakeholder needs with multi-level decision logic. The simulation engine, implemented in AnyLogic, combines the discrete-event simulation, agent-based modeling, and systems dynamics paradigms to capture regeneration across scales. A real-world industry use case of e-bikes demonstrates the ability of the framework to evaluate regeneration pathways by analyzing key performance indicators such as costs, inventories, and environmental impacts. The results highlight how digital twins act not only as digital representations but also as systemic enablers for multi-level scenario analysis, evaluation, and support of circular strategies across scales.