Koopman Operators in Robotics
A full-day workshop at Robotics: Science and Systems (Hybrid)
Date: July 15th
Room: Comissiekamer 3
Zoom: https://yale.zoom.us/j/92704321915
Abstract
Recent advances in Koopman operators have opened the door to exciting opportunities within robotics that span modeling and control, human-robot interaction, and soft robotic systems. Koopman operators are linear representations of nonlinear dynamical systems in a larger, potentially infinite space. They offer a solution for modeling nonlinear dynamics as linear dynamical systems. As a result, Koopman operators enable a paradigm shift that impacts how we model complex robotic systems and perform learning and control tasks. With Koopman operators it becomes possible to leverage linear systems analysis and linear control theory on linear descriptions of robot dynamics, ultimately improving the capabilities of complex robot systems.
While Koopman operators are promising, the current theory is still in its infancy. The purpose of this workshop is to bring together experts in robotics, machine learning, and Koopman operator theory to discuss recent developments, challenges, and future research directions. Furthermore, the workshop will facilitate discussion on alternative formulations using Koopman-like structures, applicability of Koopman operators in unconventional areas in robotics, and identify promising directions of Koopman operators in robotics.
Topics of Discussion
What is the Koopman operator and how does it help with interpretability and data?
Learning Capabilities and Analytical Limits
Limitations and Control in Lifted Spaces
Future Open Areas: Lifted Dynamic Constraints
Schedule
Intro and Theory
8:45 - 9:00 Welcome and Introduction: Ian Abraham
9:00 - 9:30 Invited Talk 1: Harry Asada (MIT)
Control-Linear Koopman Operator: A Physical Modeling Approach to Koopman Modeling that is Linear in Control
9:30 - 10:00 Invited Talk 2: Yunzhu Li (Columbia University)
Learning Compositional Koopman Operators for Model-Based Control
Coffee Break: 10:00-10:30
Theory
10:30 - 11:00 Invited Talk 3: Naoya Takeishi (Tokyo University)
Learning Neural Observables for Koopman Operators
11:00 - 11:30 Invited Talk 4: Petar Bevanda (TUM)
Kernelized Koopmanism for Prediction and Optimal Control
11:30 - 12:00 Invited Talk 5: Lu Shi (Tsinghua University)
Analytical Construction of Lifting Functions for Nonlinear Robot Systems
12:00 - 12:30 General Discussion I
Title: Practicalities of Theory in Application : Moderator -- Ian Abraham
Lunch Break: 12:30 - 14:00
Application
14:00 - 14:30 Invited Talk 6: Itta Nozawa (Sumitomo Heavy Industries)
A Monte Carlo Approach to Koopman Direct Encoding and Its Application to Crane Sway Control
14:30 - 15:00 Invited Talk 7: Ervin Kamenar (University of Rijeka)
Soft-Robot Control via Koopman Operators
15:00 - 15:30 Invited Talk 8: Emily Kamienski (MIT)
Human Gait Modeling and Analysis Using Koopman Operators
Coffee Break: 15:30 - 16:00
16:00 - 16:30 Lightning Round Spotlight Papers
16:00 - 16:10 Shahram Khorshidi Koopman-Based Centroidal State Estimation for Dynamic Legged Locomotion
16:10 - 16:20 Xubo Lyu Task-Oriented Koopman-Based Control with Contrastive Encoder
16:20 - 16:30 Yuan Zhang Latent Linear Quadratic Regulator for Robotic Control Tasks
16:30 - 17:00 General Discussion II -- Lu Shi
Title: Future Direction of Koopman Theory
Confirmed Distinguished Speakers
Organizers
Important Dates
Submission Deadline: July 1st
Notification of Acceptance: July 8th
Workshop: July 15th
Submission: Click Here
We invite interested workshop participants to submit an abstract (1-page excluding references) of their work as a contribution to our lightning talks (5-minute oral presentations).
A maximum of 4 submissions will be accepted and published on the workshop website after the conference. Contributions must adhere to the double-column IEEE ICRA template. If interested, please fill in the form, before the submission deadline.
All accepted papers will also be included in our poster sessions, which will be held in person at the workshop.
Spotlight papers will be given the chance to present their work alongside the line of speakers.
Technical Committee Support
IEEE RAS Technical Committee on Model-Based Optimization for Robotics
Venue
Please visit RSS Website for further information.