Park wins CAREER award to optimize trajectory of dynamic legged robots

MechSE Assistant Professor Hae-Won Park was awarded a prestigious National Science Foundation Faculty Early Career Development Program (CAREER) award. The CAREER award is given in support of faculty whose work shows great potential to excel as both professors and researchers early in their careers.

Park’s work involves the design and optimization of trajectory in dynamic, legged robots. Based on emulating the motion of bipedal or quadrupedal beings, designing movements of legged robots requires the difficult task of configuring different components to efficiently reach a destination.

In previous appointments, Park has worked on developing MABEL, a University of Michigan bipedal robot with the ability to walk on uneven ground, and Cheetah 2, the MIT feline-inspired robot that can change running speed as well as jump over obstacles. At Illinois, Park has focused on the design of trajectories that mimic something smaller – squirrels and small cats.

“As I see small nimble mammals move it looks like they are fully utilizing their dynamics, fully utilizing their force capability,” Park said. “So compared to a slow, quasi-static walking robot, even though they’re very big they can only handle very small obstacles because they don’t fully, efficiently utilize their dynamics.”

Not only are nimble animals very efficient in their movement, but they can traverse a myriad of obstacles and handle diverse environments. With possible applications of this research in disaster response, military situations, and exploration of inaccessible or inhospitable locations, the rodent-like versatility of the robot would be integral.

The robots are built in Park’s Dynamic Robotics Laboratory as a method for experimental research validation. In order to achieve the capability necessary to imitate movement of animals, the lab makes custom actuators and gearboxes. Currently, there are some legged robots available on the market, but without these custom components they cannot achieve the desired dynamic trajectories, nor are they robust enough to handle impact occurring when the robot lands on the ground.

The award win not only honors Park’s work, but gives him the ability to financially support a graduate student for five years to continually further this research.

“I’m happy that NSF appreciated my idea,” Park said. “I think they funded my research because my idea is important and valuable.”