AI-embodied Flexible Electronic Systems: From Sensors to Robots

Flexible electronic microsystems represent an advanced technological framework that integrates microelectronics, flexible materials, and intelligent structures. The primary challenge lies in achieving flexibility, miniaturization, and intelligence to extend the functionalities of conventional rigid electronic systems to novel applications characterized by sensing, deformation, and adaptation to complex dynamic environments. To address this, this study proposes a new paradigm termed " AI-embodied flexible electronic microsystems." This framework deeply integrates programmable flexible electronic skins, bio-inspired modular actuators, and on-chip neuromorphic computing cores, establishing a unified perception-actuation-cognition architecture. Through the co-design of reconfigurable structural topology, distributed oscillatory actuation, and multimodal sensing networks (e.g., vision, touch, temperature), the system achieves multimodal locomotion. Additionally, empowered by on-chip hyperdimensional computing-based embodied intelligence, it can autonomously perform tasks such as terrain recognition and obstacle avoidance, heat source tracking, and path optimization without external intervention or reliance on predefined models.

Addressing the inherent challenge of balancing structural compliance, multifunctional integration, and embedded computational intelligence, this study develops an integrated platform that merges flexible electronics, bio-inspired mechanics, and embodied artificial intelligence. This provides a viable technological pathway for the next generation of autonomous intelligent microsystems capable of operating in dynamic, confined, and unstructured extreme environments.