EASE addresses the Body Motion Problem (BMP), which involves making robots move in a way that achieves a goal while avoiding unintended consequences. Unlike simple automation tasks, everyday manipulation requires adaptation, reasoning, and interaction with humans. Current approaches struggle with real-time awareness, limiting robots' ability to anticipate, adjust, and integrate human feedback. EASE overcomes this by developing cognition-enabled robots that combine predictive AI with structured reasoning. Using Digital Mental Models (DMMs) and hybrid AI, EASE enables robots to refine actions dynamically, ensuring safety, adaptability, and explainability. This execution-aware approach allows robots to seamlessly operate in human environments.
