Robert D. Gregg IV Research Group

University of Michigan-Ann Arbor

Can wearable robotic devices move with users across variable tasks and environments without preset activity classification? The laboratory develops phase-based control for prosthetic legs that continuously follows a user's thigh motion across speeds, slopes, and stairs. Researchers create task-agnostic exoskeleton controllers that change the physics the wearer feels so assistance follows voluntary movement without activity classification. The group designs high-torque, low-impedance actuators that remain backdrivable while producing necessary joint torque for wearable devices. Measured laboratory outcomes include lower quadriceps effort and reduced compensatory hip moments in participant studies.