AbstractFlexible sensors hold significant promise for wearable monitoring and rehabilitation training applications. However, current flexible strain sensors struggle to compatible their sensing performance and fabrication with textile substrate and weaving technologies, which limits the practical applications of flexible sensors in commercial markets for electronic wearables. Differing from traditional strategies that rely on sophisticated construction of functional materials, leveraging industrial braiding, and weaving technologies, an all‐textile‐based pseudo‐triaxial mechanosensory textile (PTMT) is designed by engineering the wrapping pattern, yarn twist, and fabric architecture. The switchable hierarchically‐structured morphing of the PTMT enables an ultra‐wide strain detection range (up to 140%) along with desirable sensitivity. Moreover, the PTMT shows outstanding air permeability (1915 mm s−1), moisture permeability (1922 g m−2 h−1), high washability, and electromagnetic interference shielding (20.25 dB). The potential applications of PTMT are also demonstrated, such as in simulating fetal‐movement in pregnant women, proving its effectiveness in fetal‐movement health monitoring. Furthermore, by integrating the PTMT into shoe vamps and combining it with machine learning algorithms (CNN, RF, and PSO‐SVM), it is proved that PSO‐SVM outperformed CNN and RF in accuracy and stability, achieving a combined recognition accuracy of 95.42%.
Read full abstract