KneeSense: A Smart-Textile Knee Brace for In Situ Form Assessment in Remote Training and Rehabilitation
Year: 2025
Type: conference-paper
Abstract: Remote and home-based coaching expand access to expert guidance, yet their safety and effectiveness depend on in situ, joint-level assessment. We introduce KneeSense, a smart-textile knee brace that integrates dual inertial measurement units (IMUs) and a miniature time-of-flight (ToF) range sensor to produce actionable, per-repetition analytics for remote training and rehabilitation. The IMUs provide 3D orientation and acceleration, while the ToF sensor captures millimeter-scale distance changes during flexion/extension to disambiguate depth/phase and mitigate drift. From IMU-ToF signals, KneeSense computes knee flexion angle, angular speed, knee-to-knee range, and an acceleration-based instability index that flags technique issues associated with injury risk. Embedded in a washable textile sleeve, the system processes data locally to detect form deviations in real time. Bluetooth transmits summary metrics to a companion app for user feedback and remote coach/therapist review, without continuous raw-data streaming. Together, these capabilities provide a camera-free, on-body approach to joint-level form assessment that supports safer technique, improved adherence, and scalable remote care, while laying the groundwork for future closed-loop feedback.
Authors: Rongrong Gong, Haosong Zhong, Mitch Guijun Li
Institution: Hong Kong University of Science and Technology
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4416927487