Fully integrated AI-enhanced flexible wearable sensor for real-time movement evaluation and table tennis training

Wearable electronic devices are crucial for continuous health monitoring and human-computer interaction, yet achieving full integration, flexibility, and high-fidelity accuracy for biomechanical assessment remains challenging. Here, we report a thin, skin-conformal flexible inertial sensing platform for accurate motion capture and spatiotemporal movement assessment. High-precision micro-electromechanical systems (MEMS) inertial sensors are integrated onto a customized flexible printed circuit and encapsulated within a medical-grade adhesive patch for direct epidermal mounting. The standalone device wirelessly transmits acceleration, angular velocity, and orientation data with low power consumption, achieving static angular errors below 0.14° and dynamic drift of 6.21° during continuous motion. We implemented a spatiotemporal assessment framework that decouples spatial posture accuracy from temporal execution order using limited dynamic time warping (LDTW) and multiple sequence alignment (MAFFT). Using 60 expert demonstrations of table tennis forehand attacks, evaluation models were constructed with maximum spatial tolerance of 7.39° and 14 consensus temporal landmarks. A sensor network enabled motion capture at key anatomical locations, achieving overall RMSE of 11.54 ± 8.46° against optical motion capture. In a training validation study (n = 20), novices using automated sensor-based guidance achieved 41.67% improvement in movement quality, comparable to expert coaching (55.75%, p = 0.175). The platform demonstrated high-accuracy action recognition (99.86%), trajectory estimation, and real-time movement assessment, supporting its potential for motion analysis in sports training, rehabilitation, and personalized healthcare.