For soft robotic systems to emulate the adaptive behaviors of natural organisms, integrated systems with both high compliance and multimodal sensing are essential. However, existing soft robotic designs often struggle to simultaneously achieve large deformability, high force output, and stable real-time multimodal perception. Here, we report a precompressed flexoskeleton soft actuator integrated with a self-powered flexible bimodal sensor for enhanced actuation and multimodal perception. The developed self-powered bimodal sensor enables the simultaneous detection of distance and pressure, thereby allowing soft robots to perceive both noncontact proximity information and contact stimuli with good operational stability and durability. To further support this sensing platform, a flexoskeleton derived from a trimmed spiral surface was incorporated into the actuator and assembled in a precompressed state, thereby constraining radial expansion and promoting deformation through the release of stored elastic energy. The actuator delivered a blocking force of 17.28 N at 70 kPa while maintaining a bending angle of 30.5°, and also supported multidirectional bending and modular assembly. The integrated bimodal sensor combines noncontact triboelectric proximity sensing with contact piezoelectric pressure sensing, enabling self-powered discrimination of approach and touch events. In the proximity sensing mode, the open-circuit voltage increases from 0.24 to 0.71 V as the distance decreases from 25 to 5 mm. In the pressure sensing mode, the device produces up to 5.98 V and 325 nA under an 80 N load, together with good operational durability. Benefiting from the synergistic integration of self-powered sensing and structural actuation, the system demonstrates adaptive interaction in serial manipulators, plant-inspired predatory grasping in parallel soft grippers, and programmable gait perception in tripedal soft robots. This work provides a feasible strategy for deeply integrating self-powered multimodal sensing with soft actuation and highlights the potential of functional sensing materials for intelligent soft robotic systems.
Luan et al. (Fri,) studied this question.