Human-Inspired Robotics and Robotics-Informed Human Studies:
From SuperLimbs to Koopman Operator Theory

H. Harry Asada
Massachusetts Institute of Technology

October 16, 2025 (Thu)

2:30-3:00 Refreshments
3:00-4:00 Seminar

Gates 403 Fujitsu Conference Room

The seminar is open to Stanford faculty, students, and sponsors.

Abstract

The human is a metaphor for robotics and a rich source of inspirations for robot morphology, function, and dexterity/intelligence. In turn, advances of robotics inform human studies. In the first half of this talk, broad aspects of human-inspired robotics research of the speaker and his collaborators will be presented. These include early works on grasp stability and compliance synthesis for explaining human dexterous manipulation, direct-drive and serial elastic actuators for realizing human-like actuation, and more recent works on Supernumerary Robotics Limbs for augmenting human physical strength, and their applications to assisting astronauts on the moon and elderly people at nursing homes.

In the second half, which is the main topic of this talk, it will be demonstrated that human-like dexterity and intelligence in performing highly nonlinear dynamical tasks can be realized with a rather simple linear control in the light of Koopman operator theory. The Koopman theory is an emerging mathematical framework for representing nonlinear dynamical processes as globally-linear models in a high-dimensional embedded function space, which is effectively constructed with neural networks. Three case studies exemplify the Koopman lifting linearization that allows for simple linear controls. 1) Human sit-to-stand transition consists of multi-phase local dynamics starting off at a relaxed sitting posture to the standing posture with stabilizing feedback control. Koopman operators combine the diverse local dynamics into a globally linear model allowing for smooth, efficient control across the multi-phase local dynamics. 2). Human gait is a switched dynamical process governed by diverse local dynamics, which are switched as each leg makes and breaks contact with a ground. Koopman operators can subsume the segmented dynamics within a globally linear model amenable for control. 3). Soccer and hockey players can dribble and shoot a ball/pack dexterously and strategically. Koopman operators can represent such contact-rich handling of objects as a unified, globally linear process. All these Koopman models are globally linear in a high dimensional embedded space and reduce otherwise highly complex, nonlinear optimal control to straightforward convex optimization suitable for real-time control. It is conjectured that human-like dexterity and intelligence in performing multi-phase, segmented dynamical tasks can be elucidated as a rather simple linear control if the human/machine can learn a globally linear dynamic model with use of effective embeddings.

Bio-Sketch

H. Harry Asada is Ford Professor of Mechanical Engineering and Director of the Brit and Alex d’Arbeloff Laboratory for Information Systems and Technology in the Department of Mechanical Engineering, Massachusetts Institute of Technology (MIT), Cambridge, MA. He specializes in robotics, system dynamics and control, and biomedical engineering. His current research includes Koopman operator theory, assistive robotics for eldercare, supernumerary robotic limbs for assisting astronauts, and multi-cable manipulation. He received Best Paper Awards at the IEEE International Conference on Robotics and Automation (ICRA) in 1993, 1997, 1999, and 2010, the best application paper award at the 2017 IEEE/RSJ International Conference on Intelligent Robotics and Systems (IROS), the O. Hugo Schuck Best Paper Award from the American Control Council in 1985, and other 8 best paper awards of major journals and conferences. He was the recipient of the 2011 Rufus Oldenburger Medal from ASME, the Henry Paynter Outstanding Investigator Award from the ASME Dynamic Systems and Control Division in 1998, and the Ruth and Joel Spira Award for Distinguished Teaching from School of Engineering, MIT, in 2011. More recently he received the 2023 Pioneer in Robotics and Automation Award from the IEEE Robotics and Automation Society. He is a fellow of IEEE and ASME. 

Please visit https://stanfordasl.github.io/robotics_seminar/ for this quarter’s lineup of speakers. Although we encourage live in-person attendance, recordings of talks will be posted also.

Covid-19 related instructions: We recommend wearing a well-fitted, high-quality face covering inside the classroom.