Robotics Colloquium: Guest Speaker Koushil Sreenath

Title: Adventures in Learning Dynamic Legged Loco-Manipulation 

Abstract: This talk delves into recent advancements in learning techniques applied to complex dynamical systems, particularly focusing on bipedal robots, quadrupedal robots, and humanoids. Through the lens of reinforcement learning, we showcase the capacity to train policies enabling a range of locomotive behaviors including walking, jumping, and running. Additionally, we also conduct a preliminary study on the stability of reinforcement learning-based policies.  Next, by investigating hierarchical reinforcement learning architectures, we seggregate planning and control policies to achieve complex tasks such as soccer goal shooting and goalkeeping. Moreover, we demonstrate the direct utilization of reinforcement learning to output joint torques and exploit symmetry within the robot/task to enhance robustness and sample efficiency. Additionally, generative methods like large language models and diffusion are employed to achieve quadrupedal locomotion.  Lastly, we utilize transformer architectures for learning humanoid locomotion both through reinforcement learning and through learning from data collected from existing controllers, as well as human motion-capture and video data. 

Bio: Koushil Sreenath is an Associate Professor of Mechanical Engineering, at UC Berkeley. He received a Ph.D. degree in Electrical Engineering and Computer Science and a M.S. degree in Applied Mathematics from the University of Michigan at Ann Arbor, MI, in 2011. He was a Postdoctoral Scholar at the GRASP Lab at University of Pennsylvania from 2011 to 2013 and an Assistant Professor at Carnegie Mellon University from 2013 to 2017. His research interest lies at the intersection of highly dynamic robotics and applied nonlinear control. His work on dynamic legged locomotion was featured on The Discovery Channel, CNN, ESPN, FOX, and CBS. His work on dynamic aerial manipulation was featured on the IEEE Spectrum, New Scientist, and Huffington Post. His work on adaptive sampling with mobile sensor networks was published as a book. He received the NSF CAREER, Hellman Fellow, Google Faculty Research Award in Robotics, and Best Paper Awards at Learning for Dynamics and Control (L4DC) and Robotics: Science and Systems (RSS).