Jonathan Hurst

I started the first Robotics lab at OSU, the Dynamic Robotics Lab, and after a few years of hiring trusted colleagues in Robotics, Co-Founded the Robotics Program at Oregon State University. We became only the third institution in the USA to offer Ph.D.'s and M.S. degrees in Robotics, and have grown to over a hundred students, our own building, and quite a few faculty (12-30, depending on how you count).

My research focused on understanding the fundamental science and engineering best practices for robotic legged locomotion and physical interaction.

Specific research topics have shifted over the years, identifying the current blockers to highly dynamic legged locomotion - beginning with hardware dynamics, ensuring it would be possible to control a robot for walking and running. The culmination of this work is the bipedal robot ATRIAS, the first robot to reproduce human walking gait dynamics. With simple control ATRIAS walks and runs over outdoor terrain, with no perception.

After gaining an understanding of hardware for dynamic locomotion, our focus shifted to control, and recently we have pioneered reinforcement learning approaches to highly dynamic behavior. The success of this approach is demonstrated on Cassie, where we implemented a learned policy for running, and achieved the world record for the fastest 100 meter dash by a bipedal robot.

When I interviewed for the role at OSU, I said I would start a company when we solved some of the core science problems and understood enough about how to implement walking and running robots that we could achieve biomechanically-comparable dynamic legged locomotion. The university leadership was fully in support, and has been a great partner. It has taken about ten years to get there, find the right time, the right co-founders, and solve enough of the science that a startup company would be the best path forward.

The Robotics Institute at Carnegie Mellon was the place to be for Robotics as a grad student, most especially in the 2000's. Robotics was a very young field, yet more than 300 people were at the Robotics Institute, across professors, staff, and students! I came to build legged robots, and had the support and freedom to do so, as well as advising from some of the best and most experienced people in this field in the world. During this time, I built and supported the CMU RHex robot, designed and built the Actuator with Mechanically Adjustable Series Compliance (AMASC), and designed and built MABEL, a large bipedal robot, entirely using cable drive transmissions, and delivered to collaborators at the University of Michigan. A great experience, a great kickoff to a career.

I joined the Leg Lab for a summer internship with the hope of joining for graduate school. Unfortunately, my internship was the last year that the Leg Lab existed - and I am grateful for the experience, for getting in while it was there. I still have the lab key, though the building has since been demolished!

Summer Intern