Agentic AI and Digital Twin: A Good Marriage?

Abstract

With the rapid advancement of Agentic Artificial Intelligence (AAI) and the increasing maturity of Digital Twin (DT) technologies, a new paradigm is emerging at the intersection of cognitive autonomy and cyber physical system (CPS) representation. In fact, DT provides high-fidelity virtual models of physical assets and processes, enabling monitoring, simulation, and scenario analysis; in parallel, AAI introduces autonomous reasoning, planning, coordination, and workflow orchestration capabilities that allow systems to operate with minimal human intervention. The convergence of these two paradigms opens the door to a powerful yet complex synergy. The aim of this paper is exactly to systematically investigate the marriage between DT and AAI by examining their definitions, architectural roles, and integration patterns. To achieve this objective, we conduct a systematic mapping following the PRISMA methodology: by clarifying the conditions under which their convergence becomes meaningful, this work provides a structured foundation for understanding the emerging DT-AAI systems and their application domains, limitations, and open challenges.