Jehan Yang

Prosthetic Leg Pressure Sensing

The AMBER Lab at Caltech specializes in legged walking. Here, I worked with Rachel Gehlhar on improving the robustness of a custom prosthetic robot leg by adding a pressure sensor in order for walking on various terrain and with various users without a need for adjusting the controller used.

Here I worked on real-time pressure sensing in order to control a prosthetic robot leg. Challenges of real-time pressure sensing include reading from all the sensors at a high frequency in order to feed into the system. This time constraint caused me to look into real-time OS guarantees, multithreading, and various adjustments in order to not slow down the control loop. The results from this paper ended up being published in the Robotics and Automation Letters along with IEEE ICRA 2022.

Working with a prosthetic robot leg is challenging in that experiments require a certain amount of risk for the user. By using theoretical guarantees of nonlinear control convergence at a exponential rate using ID-CLF-QP, we were able to ensure the robot moved efficiently and follow the desired trajectory.