AME - Control of Robotic Lower-Limb Prostheses

This project aims to evaluate the feasibility of new control strategies for robotic lower-limb prostheses. For individuals with lower-limb amputation, both below- and above-knee, robotic protheses have the potential to provide greater functionality than conventional passive prostheses that are still far more common today. One significant challenge is how to provide control of these robotics prostheses across the wide range of tasks users perform in their activities of daily living. This project seeks to determine the range of tasks that could be reliably achieved with a single controller without requiring a highly accurate activity recognition algorithm to signal switching among various controllers. Initial work will involve analysis of publicly available datasets to identify easily measurable variables, like joint angles/velocities, that exhibit common behavior across walking speeds, slopes, stairs, etc. Subsequent work may involve analysis of how such variables could be manipulated to create task-invariant behavior. In the long term, the project would ideally involve implementing a corresponding control strategy on a transtibial (for below-knee amputation) robotic prosthesis in the lab and conducting pilot studies with individuals walking on the device with the new controller.