A wearable smart glove based on flexible printed circuit boards for multi-gesture detection

Abstract

This paper presents an enhanced inductive-sensing smart glove that expands hand-motion capture from 9 to 14 degrees of freedom while reducing the total mass to 69 g and shortening the device by 13 cm. Inductive coils on a flexible printed circuit board are read out and modeled with third-order polynomials to convert oscillation frequency to joint angles. With a 30 Hz sampling rate, the system achieves median angular resolutions of 6.8° for metacarpophalangeal (MCP) flexion–extension, 2.93° for proximal interphalangeal (PIP) flexion–extension, and 1.0° for MCP adduction–abduction, with model fits of R2 = 0.972–0.998. At flexion angles beyond ~80°, the sensing region saturates and the output becomes effectively binary or ‘fully flexed’. A Python visualizer renders a real-time three-dimensional (3D) hand model and streams joint angles for downstream applications. The present prototype was custom-fitted to a single user; generalizability across hand sizes will be addressed via adjustable finger sleeves and a reconfigurable base in future designs. Owing to its modular construction, drift-resistant inductive sensing, and lightweight chassis, the glove offers a practical basis for gesture-based interaction in rehabilitation and human–machine interface (HMI) applications, while its suitability for immersive virtual reality (VR) remains a subject of future investigation.