Adaptive hybrid reasoning for agent-based digital twins of distributed multi-robot systems

Abstract

The digital twin (DT) mainly acts as a virtual exemplification of a real-world entity, system, or process via multiphysical and logical models, allowing the capture and synchronization of its functions and attributes. The bridge between the actual system and the digital realm can be utilized to optimize the system’s performance, and forecast and predict its behavior. Incorporating intelligent and adaptive reasoning mechanisms into DTs is crucial to enable them to reason, adapt, and take efficacious actions in complex and dynamic environments. To this end, we introduce an approach for deploying agent-based DTs for cyber-physical systems. The foundation pillars of this approach are (1) integrating the concepts, entities, and relations of Zeigler’s modeling and simulation framework from the perspective of agent-based DTs; (2) utilizing an expandable and scalable architecture for designing and materializing these twins, which handily enables extending and scaling physical and digital assets of the system; and finally (3) a two-tier reasoning strategy; reactive and rational models are conceptually redefined in the context of the modeling and simulation framework of agent-based DTs and technically deployed to boost the adaptive reasoning and decision-making function of DTs. As a result, an implemented simulation and control platform for a multi-robot system demonstrates the approach’s applicability and feasibility, manifesting its usability and efficiency. The platform represents physical entities as autonomous agents, creates their DTs, and assigns adequate reasoning capability to promote adaptive planning, autonomous resource management, and flexible logical decision-making to handle different situations and scenarios.