There has been much development of wearable robots, yet the basic understanding about how humans and machines co-adapt remains unclear. We hypothesize that understanding both how humans adapt their neuromotor control to these machines and how control algorithms can adapt these machines to humans will deliver a performance of at least 100% over current results.
We plan to use a lightweight wearable robot platorm that doesn’t restrict natural movement when worn and lends itself to studying sensorimotor adaptation alongside state-of-the-art biomechanical and physiological tools as a platorm for proving the above hypothesis. This project will formalize existing early collaborations in this area and bring together a diverse team with experts in robotics, motor learning, and biomechanics and biology. We plan to develop experimental protocols and data analysis techniques to understand how individuals adapt to assistance from a wearable robot. We also plan to develop online optmization methods that enable the robot controller to minimize biomechanical and physiological objective functions such as forward propulsion, muscle activity and energetics.
Team: Conor Walsh, Maurice Smith, Scott Kuindersma, Robert Howe, and Andy Biewener