A self-powered hydrogel electronic skin with decoupled multimodal sensing for closed-loop human-machine interactions.

Abstract

Bridging biological and artificial systems, intelligent interfaces drive the demand for flexible electronics that emulate the skin's multifunctionality. However, achieving such multifunctionality in a compact, self-sustained form remains challenging, as multimodal sensors often rely on rigid materials, discrete components, and external power sources. Herein, this study presents a single-component poly(vinyl alcohol) hydrogel e-skin integrating thermogalvanic, piezoionic, and diffusion mechanisms for self-powered sensing of skin temperature, arterial pulsation, and sweat secretion, simultaneously. The hydrogel features high stretchability, low modulus, and a prismatic architecture synergizing ionic polarization. Moreover, a temporal machine learning model with local attention is developed to decouple multimodal signals. Of practical importance, an active multimodal signal generator wristband is developed as a multifunctional human-machine interface for physiological detection, robotic control, and haptic feedback reproduction. Hence, this hydrogel e-skin represents an efficient material platform for intelligent interactions, showing broad potential for real-time health monitoring, robotic control, and virtual reality.