Movement data from prosthesis users are difficult to collect. AidWear aims to develop an AI-controlled prosthesis that helps users perform better on the challenging Cybathlon course. Our research explores learning from synthetic data and healthy participants, then transferring those models to prosthesis users.
HERMES is an output of AidWear. It connects synchronized multimodal sensing, real-time AI inference and closed-loop intervention, providing an open-source foundation for developing and testing intelligent assistive control.