Intelligent Foot Perception System Based on Flexible MXene Triboelectric Sensing and Deep Learning for Motion Recognition and Health Monitoring

ABSTRACT

Wearable health monitoring technology places high demands on flexible sensors in terms of high performance, high comfort, and multi‐functional integration. This study presents an intelligent foot perception system for wearable health monitoring. It integrates a high‐performance, flexible triboelectric sensor and a hybrid deep‐learning model. The sensor uses a MXene‐Ecoflex composite dielectric layer with a microcolumn array, which triples its output power to 0.91 mW. It demonstrates excellent linearity (R ²  > 0.99), high sensitivity (1.3 V/kPa), a wide detection range (0–200 kPa), and a low detection limit (15.6 Pa). Furthermore, we integrated the flexible sensor with a three‐axis accelerometer on a flexible printed circuit board. This constructed a wearable perception system. The system can simultaneously acquire fused pressure and acceleration signals. The system uses a CNN‐LSTM hybrid deep learning algorithm. It achieves an accuracy of 97.6% and an F1‐score of 0.98 on various motion classification tasks. Moreover, it can effectively distinguish between healthy and disabled states. This study confirms that the proposed flexible sensing system has great potential for application. It can be used for real‐time and accurate fall warning and mobile health monitoring.