AWE USA 2026
Robots and virtual environments share a common failure: none can reliably sense the physical world the way biological systems do. In the human body, contact, tension, and slip are processed peripherally before signals ever reach conscious awareness. Achieving this in Physical AI built on inorganic systems is still one of the most formidable open problems facing human-machine interaction.
This talk addresses that challenge across three domains: inferring contact from actuation-side signals alone, translating those inferences into feedback in robotic and virtual systems, and reading the human side of the loop through biofeedback. Live demonstrations.