Wearable Capacitive Pressure Sensors Based on Porous Sponge Dielectric Materials for Gait Analysis and Identity Recognition

Abstract Gait often reveals valuable information about personal movements in daily life, and traditional gait monitoring often relies on inertial sensors, which are limited by high manufacturing costs, inconvenient operation, and complex analysis methods. This study proposes a simple and cost‐effective method to manufacture a wearable capacitive sensor, which can efficiently detect different gait signals. The sensor is made by a polyurethane sponge with deposited activated carbon (C@PU sponge) as the dielectric layer, and it has high sensitivity and pretty good stability. The sensor is further integrated into an insole for detecting and collecting signals of foot pressure during human movement. To overcome the limitations of traditional analysis methods, a convolutional neural network model based on residual networks is designed for detecting nine different human activities with a recognition accuracy of 98.15%. Furthermore, the residual network is optimized using a genetic algorithm, and the optimized model is able to effectively identify eight participants with a recognition accuracy of 98.75%. These results indicate that smart insoles based on wearable capacitive sensors show good application prospects in gait analysis and identity recognition, and are expected to be widely used in daily life in the future.