Using Digital Twins to Tame the Testing of AI/ML Systems

Abstract

Program test managers and test engineers should carefully consider Digital Twinning approaches for addressing training and testing challenges for Artificial Intelligence/Machine Learning (AI/ML) systems. A hybrid Hardware in the Loop (HITL) and Digital Twin (DT) architecture is discussed for a notional Cognitive EW system. This architecture may provide effective training and testing for complex AI/ML systems that incorporate extensive Cyber-Physical interactions. Considerations for generating realistic RF test environments for Cognitive EW systems are also considered.