Sensing, Control and Learning in a Powered Ankle Exoskeleton
Human Neuromechanics Laboratory
University of Michigan
Tuesday, March 31, 2015
Goldwater Research Center (GWC) 487, Tempe campus [map]
Recent advances in robotics technology have lead to a proliferation of exoskeleton and prosthetic devices designed to augment the performance of workers and soldiers and to improve enhance the quality of life of patient populations. However, the principles which guide the human-robot interface, namely the sensing of user intent and user state are not well understood. Allowing the human operator greater authority in the control of the device could reduce the variability of outcomes of current devices. In this talk, I will discuss the design of low-cost sensors and physiologically-inspired control algorithms for actuating a powered ankle exoskeleton highlighting ways in which these studies can support new knowledge in human-robot interfaces.
Daniel A. Jacobs, Ph.D, is a postdoctoral fellow in the School of Kinesiology at the University of Michigan. He received the B.S., M.S. and Ph.D degrees in Mechanical Engineering from Stanford University, investigating dynamic models of contact, and efficient, stable control for running and bounding in legged robots. He was postdoctoral fellow in Bioengineering at Stanford University developing dynamic walking models in the open source musculoskeletal modeling program OpenSim. His research interests are in the design, control and sensing of robotic systems for augmentation and rehabilitation, legged locomotion in robots, and the neuromechanics and motor learning of human locomotion.