AI-Driven Wearable Robotics for Healthcare

The AidWear project aims to develop the necessary artificial intelligence frameworks that will enable Robotic Assistive Devices (active prosthetics and lower-limb exoskeletons) to give Parkinson’s patients and individuals with an amputation the ability to maintain their independence.
The project will collect data in various tasks that will allow the development of control methods for three areas of interest: intention detection, mid-level optimization, and dynamic simulation. AidWear will address these issues through the development of innovative AI-based systems that will exploit new and existing kinematic datasets with measures of intent, create a robust intention detection system through learning, automatic controller optimization using RL, automatic model simplification to allow these AI frameworks to be run on efficient hardware at fast rates, and accurate dynamic models that will allow for data generation. These systems will be tested in and outside the lab, including at CYBATHLON and in the home environment.