Leveraging Data and the Koopman Operator to Make Soft Robots More Capable

Daniel Bruder
Assistant Professor, Mechanical Engineering
University of Michigan

Abstract: Soft robots are able to safely interact with delicate objects, absorb impacts without damage, and adapt to the shape of their environment, making them ideal for applications that require safe robot-human interaction. However, their use in real-world applications has been limited due to the difficulty involved in modeling and controlling them. In this talk, I’ll describe a data-driven modeling approach aimed at overcoming the limitations of previous methods. This approach leverages Koopman operator theory to construct linear representations of nonlinear dynamical systems, enabling the use of linear control techniques. Using this Koopman-based approach, a pneumatically actuated soft arm was able to autonomously perform manipulation tasks such as trajectory following, pick-and-place with a variable payload, and writing on a dry-erase board without undergoing any task-specific training. In the future, this approach could offer a paradigm for designing and controlling all soft robotic systems, leading to their more widespread adoption in real-world applications.

Bio: Daniel Bruder is an Assistant Professor of Mechanical Engineering at the University of Michigan. Prior to this appointment, he was a postdoctoral fellow at Harvard University in the Microrobotics Lab supervised by Prof. Robert Wood. He received a Ph.D. in mechanical engineering from the University of Michigan in 2020 and a B.S. degree in engineering sciences from Harvard University in 2013. He is a recipient of the NSF Graduate Research Fellowship and the Richard and Eleanor Towner Prize for Outstanding Ph.D. Research. His research interests include the design, modeling, and control of robotic systems, especially soft robots.