Developing Digital Twins at the Subsystem Level Using Heterogenous Modeling and Simulation (M&S) for Development and Testing of Artifical Intelligence/Machine Learning (AI/ML) and Autonomous Ground Systems

To achieve Army modernization plans, advanced approaches for testing and evaluation of autonomous ground systems and their integration with human operators should be utilized. This paper presents a framework for developing digital twins at the subsystem level using heterogeneous modeling and simulation (M&S) to address the challenges of manned-unmanned teaming (MUM-T) in operational environments. Focusing on the interplay between robotic combat vehicles (RCVs) and human operations, the framework enables evaluation of soldiers’ cognitive loads while managing tasks such as maneuvering robotic systems, interacting with aided target detection, and engaging simulated adversaries. By employing subsystem-level digital twins, we aim to isolate and control key variables, enabling a detailed assessment of both systems’ performance and operator effectiveness. Through realistic operational scenarios and human-machine interface testing, our approach may help identify optimal solutions for soldier-robot collaborations, ensuring readiness in MUM-T operations. This methodology provides a pathway for refining AI/ML capabilities, enhancing autonomy, and informing the Army’s broader testing and evaluation objectives.