Robust self-encapsulated fiber integrated with rescue robots for environmental information collection

In recent years, individuals operating in extreme environments have encountered challenges such as high accident rates and low work efficiency. The integration of embodied intelligence technology with flexible sensors is expected to provide a novel problem-solving approach. In this study, liquid metal (LM) was incorporated into polydimethylsiloxane (PDMS) to reduce the surface tension of LM, resulting in the formation of a highly conductive and flexible PDMS-LM ink (PLM ink). Utilizing coaxial printing technology, PLM was embedded within room-temperature-vulcanized single-component silicone rubber (RTVS), forming a self-encapsulated structure that enables the coaxial fibers with excellent electrical conductivity and mechanical properties. By integrating triboelectric nanogenerator (TENG) technology, the sensor demonstrated remarkable sensitivity (4.03 V$\bullet$kPa⁻¹), rapid response time (23 ms), and outstanding durability (exceeding 10,000 cycles). Furthermore, the sensor's excellent interfacial compatibility and environmental adaptability enable seamless integration into robotic systems, achieving 100 % accuracy in object shape recognition and 98.3 % accuracy in material recognition using artificial convolutional neural networks. Finally, the integration of the sensor with a miniature robotic system enabled the real-time perception of environmental information (road surface smoothness, obstacles) under harsh operating conditions, highlighting its potential for robotic systems operating in extreme environments.