People are pretty smart. Prosthesis control signals are noisy. Existing control interfaces don't really take either of these facts into account. Computational motor control is a promising approach to understanding how people control their limbs that can account both for the intelligence of people and the uncertainty of their available control signals. We are applying it to prosthesis control with the hope that if we can accurately model people-in-the-loop for existing control strategies, we can then use our model to evolve better control strategies that harness the intelligence of people and minimize the effect of signal uncertainty on performance
By spending time understanding the clinical needs of people who choose not to use a prosthesis, and by digging deep into the theory of motor and transmission theory, we have developed a small yet powerful prosthetic arm that is currently being tested in field trials.