The integration of flexible sensors into human-machine interfaces (HMIs) is in increasing demand for intuitive and effective manipulation. Traditional glove-based HMIs, constrained by nonconformal rigid structures or the need for bulky batteries, face limitations in continuous operation. Addressing this, we introduce yarn-based bend sensors in our smart glove, which are wirelessly powered and harvest energy from a fully textile 5.8 GHz WiFi-band antenna receiver. These sensors exhibit a gauge factor (GF) of 5.60 for strains ranging from 0 to 10%. They show a consistent performance regardless of the straining frequency when being stretched and released at frequencies between 0.1 and 0.7 Hz. This reliability ensures that the sensor output is solely dependent on the yarn's elongation. Accurately detecting finger-bending movements from 0° to 90° in a virtual environment, the sensors enable enhanced degrees of freedom for human finger interaction. When integrated with advanced machine-learning techniques, the system achieves a classification accuracy of 98.75% for object recognition, demonstrating its potential for precise and accurate HMI. Unlike conventional near-field energy transfer methods that rely on magnetic flux and are limited by power loss over distance, our fully textile design effectively harvests microwave energy, showing no voltage deterioration up to 1 m away. This minimalist microwave-powered smart glove represents a significant advancement, offering a viable and practical solution for developing intuitive and reliable HMIs.
Keywords: flexible electronics; graphene; human-machine interfaces; wearable smart textiles; wireless-power transfer; yarn sensors.