Capacitive BaTiO3-PDMS hand-gesture sensor: Insights into sensing mechanisms and signal classification with machine learning

Abstract

Flexible sensors have gained extensive interest because of their versatile applications in healthcare, robotics, and wearable devices. This study introduces a capacitive sensor utilizing barium titanate oxide (BaTiO3)-polydimethylsiloxane (PDMS) for bending sensing and addresses crucial performance parameters including sensitivity, repeatability, response time, and durability. The sensor exhibited a notable capacitance change of 42.85% in conjunction with fast (1 s) responses and recovery times, and minimal hysteresis (<2%). Its reliable performance across varying bending rates and durability through extensive cyclic tests underscore its applicability in real-world scenarios. Importantly, the sensor's capabilities were enhanced by integrating machine learning (ML), achieving an impressive accuracy of 97.11% in recognizing hand-sign language gestures. Furthermore, finite element analysis was employed to validate the correlation between the increase in compression-induced packing density and capacitance enhancement. This holistic integration of advanced materials, computational simulations, and ML not only extends the boundaries of sensor technology but also holds promise for revolutionizing human–machine interactions, aiding speech-impaired individuals, and enriching virtual reality experiences. This study represents a pivotal advancement in the field of flexible sensors and the unlocking of new dimensions of their applications.