Knitted self-powered sensing textiles for machine learning-assisted sitting posture monitoring and correction

Abstract

With increasing work pressure in modern society, prolonged sedentary positions with poor sitting postures can cause physical and psychological problems, including obesity, muscular disorders, and myopia. In this paper, we present a self-powered sitting position monitoring vest (SPMV) based on triboelectric nanogenerators (TENGs) to achieve accurate real-time posture recognition through an integrated machine learning algorithm. The SPMV achieves high sensitivity (0.16 mV/Pa), favorable stretchability (10%), good stability (12,000 cycles), and machine washability (10 h) by employing knitted double threads interlaced with conductive fiber and nylon yarn. Utilizing a knitted structure and sensor arrays that are stitched into different parts of the clothing, the SPMV offers a non-invasive method of recognizing different sitting postures, providing feedback, and warning users while enhancing long-term wearing comfortability. It achieves a posture recognition accuracy of 96.6% using the random forest classifier, which is higher than the logistic regression (95.5%) and decision tree (94.3%) classifiers. The TENG-based SPMV offers a reliable solution in the healthcare system for non-invasive and long-term monitoring, promoting the development of triboelectric-based wearable electronics.