Abstract Muscle contraction and relaxation encode crucial signals for monitoring physical states, rehabilitation, and injury prevention. However, flexible wearable devices fail to deliver customizable sensors that combine high sensitivity, durability, and stability, limiting the performance of biofeedback. Herein, the machine learning‐assisted smart motion is demonstrated for full‐body motion recognition and rehabilitation assessment using a designed triboelectric nanogenerator (TENG) with an unsymmetrical structure. Unsymmetrical ZnSnO 3 quantum dots (QDs) deposited in a hydrogel made of polyacrylamide as a multimodal TENG are rationally fabricated. The non‐centrosymmetric crystal structure of ZnSnO 3 QDs generates pronounced piezoelectricity, originating from lattice polarization induced by mechanical deformation at the nanoscale. Piezoelectricity enhances the density of transferred charge, adjusting the deposition amount and spatial distribution to improve the performance of TENG. Chemical crosslinking enabled by ZnSnO 3 QDs significantly enhances the tensile capacity (1082%) and high sensitivity (gauge factor = 7.38). Thus, the TENG harvests biological energy and converts it into electricity, while integrating triboelectric sensing with deep learning enables a breakthrough from simple motion perception to accurate motion pattern recognition. The unsymmetrical film offers a promising strategy to enhance the output performance and durability, offering new opportunities for proximal prediction and tactile identification functions in advanced human‐machine interfaces.
Yin et al. (2025) studied this question.