Deep Learning-Assisted Superhydrophobic LIG/MWCNT Wearable Sensor for Underwater Motion Detection

Carbon-based nanomaterials, particularly graphene, have garnered significant attention across diverse fields owing to their remarkable properties and versatile applications. Among these, laser-induced graphene (LIG) exhibits high porosity, excellent conductivity, and mechanical flexibility, which can be obtained through laser irradiation of carbon-rich precursors. This article introduces a novel flexible tactile sensor based on LIG/multiwalled carbon nanotube (MWCNT) composites. The surface is treated with spraying SiO2 and precured Ecoflex to form superhydrophobic property with water contact angle (WCA) of 153.4°, which is suitable for underwater condition. Enhanced with deep learning algorithms, the LIG/MWCNT sensors are integrated into a data glove, enabling precise recognition of Arabic sign language gestures with accuracy of 99.34%. The data glove equipped with the sensor can communicate wirelessly with a smartphone App via Bluetooth low energy (BLE) by a low-power and secure wireless protocol. The App displays the finger bending status and the recognized gestures in real time, and also provides feedback to the user. This work showcases the feasibility of the proposed sensing system for gesture recognition and the potential applications in underwater wearable electronics. Based on the excellent performance of the proposed sensing system, this article contributes to advancing the interdisciplinary research field of carbon-based flexible electronics and deep learning-based sensing applications.