Wearable and flexible sensor based on laser-induced graphene/Kevlar as smart gloves for the table tennis gesture recognition and training

Abstract

The development of conformable and robust flexible electronics is paramount for the advancement of wearable systems capable of sophisticated human motion analysis. In this work, a flexible sensor based on laser-induced graphene (LIG) on a polyimide/knitted Kevlar (PI/Kevlar) composite fabric is presented. Exceptional performance was exhibited by the resulting LIG(PI)/Kevlar sensor, including high piezoresistive sensitivity (up to 1.4 kPa⁻¹), a large gauge factor (up to 47.83), and robust stability over 1000 cycles. The primary innovation of this research is demonstrated through the application of this sensor within a smart glove system, by which high-precision recognition of athletic movements is enabled when coupled with a one-dimensional convolutional neural network (1D-CNN). Specifically, four fundamental table tennis gestures were successfully identified with an average accuracy of 92.5%. Through this study, a powerful tool for quantitative analysis and skill assessment in athletic training is provided. Ultimately, a novel solution for technique refinement in sports such as table tennis is offered, highlighting a promising pathway for integrating advanced materials with artificial intelligence to enhance human performance.