Wearable Smart Gloves Based on Laser-Induced Graphene for Multimodal Sensing and AI-Enhanced Recognition

This study presents a comprehensive investigation of laser-induced graphene (LIG) for multimodal sensing applications. LIG, synthesized through direct laser irradiation of polyimide (PI) substrates, offers unique advantages such as high throughput, programmable patterning, and environmental friendliness. We explore the effects of laser intensity and scanning speed on the microstructure and sensing performance of LIG. Various LIG patterns are designed and tested for bending, tactile, temperature, humidity, and sweat sensing experiments. The results demonstrate that LIG-based sensors exhibit high sensitivity, linearity, and repeatability. Specifically, the ${R}2$ -LIG sensor shows optimal performance for bending sensing with 873.1% resistance change at 30 N and 97.5% accuracy for eight materials via a convolutional neural network (CNN) deep learning model, enabled by a four-stage signal preprocessing, e.g., data augmentation, resolution unification, ImageNet channel standardization, and dynamic dimension calculation. Furthermore, by coating LIG with PEDOT: PSS, we enhance its temperature-sensing properties, enabling temperature-sensitive sensing applications. Based on these findings, we developed a smart glove incorporating these LIG sensors, which can monitor hand joint movements, recognize tactile inputs, and sense temperature changes in real time. This study highlights the versatile sensing potential of LIG and its promising applications in wearable devices and the Internet of Things (IoT).