AI-embodied multi-modal flexible electronic robots with programmable sensing, actuating and self-learning

Abstract

Achieving robust environmental interaction in small-scale soft robotics remains challenging due to limitations in terrain adaptability, real-time perception, and autonomous decision-making. Here, we introduce Flexible Electronic Robots constructed from programmable flexible electronic components and setae modules. The integrated platform combines multimodal sensing/actuation with embedded computing, enabling adaptive operation in diverse environments. Applying modular design principles to configure structural topologies, actuation sequences, and circuit layouts, these robots achieve multimodal locomotion, including vertical surface traversal, directional control, and obstacle navigation. The system implements proprioception (shape and attitude) and exteroception (vision, temperature, humidity, proximity and pathway shape recognition) under dynamic conditions. Onboard computational units enable autonomous behaviors like hazard evasion and thermal gradient tracking through adaptive decision-making, supported by embodied artificial intelligence. In this work, we establish a framework for creating small-scale soft robots with enhanced environmental intelligence through tightly integrated sensing, actuation, and decision-making architectures.