Cognitive digital twin architectures for monitoring, real-time assessment and anticipation of recycling chain dynamics

This thesis aims to develop cognitive digital twin (DT) architectures, to support the monitoring, real-time assessment and anticipation of recycling chain dynamics. It includes to:

• Simulate and anticipate transformations at the meso (inter-organizational networks) and micro (process or actor-specific) levels within recycling value chains.
• Evaluate the economic, environmental and social impacts of variations in decisions.
• Generate real-time insights to inform decision-making and policy development.

The expected scientific contributions are:

• The development of advanced simulation services for the flow of recycled products and materials, enabling “what-if” scenario analysis and decision impact forecasting (thus providing predictive insights to support agile and informed value chain management).
• The System-of-Systems (SoS) integration of the DTs with the existing collaborative information system (IS) to enable physical and virtual components alignment (by exploring the integration of multiple local DTs into a coherent SoS architecture capable of decentralized, asynchronous decision-making).
• The formalization of data flow strategies between physical entities and digital surrogates based on internal and external key performance indicators (KPIs), to enhance system resilience and responsiveness while managing computational load (by generating and adapting DT models from heterogeneous and uncertain data, as well as assessing through sensitivity analyses how data quality impacts forecasting reliability under resource constraints).