Embodied Intelligence via Instabilities

Embedding physical intelligence into structures holds great potential for developing machines that can navigate unstructured environments while minimizing computational resources and electronic components. Realizing such physically intelligent systems requires: (1) designs that convert applied inputs into programmable deformations, and (2) the ability to sense external forces and respond to them. In this talk, I will present our recent work toward creating machines capable of autonomous maze navigation and obstacle avoidance, all without relying on traditional computational architectures.