Events Calendar

Ahmed H. Qureshi (Purdue University) - Harnessing Physics Priors for Efficient and Scalable Robot Motion Learning

Speaker: Ahmed H. Qureshi, Purdue University

Talk Title: Harnessing Physics Priors for Efficient and Scalable Robot Motion Learning

Abstract: This talk will outline the use of physics priors towards creating efficient plug-and-play algorithms for robot motion learning. These algorithms require minimal to no expert data and achieve high efficiency in both training and inference while effectively operating in complex, high-dimensional environments under various constraints. Recent advancements in robot motion learning include methods based on imitation and offline reinforcement learning, which are known to necessitate a significant amount of expert trajectories and entail high training times. In contrast, this presentation will introduce a new class of self-supervised, physics-informed neural motion policy learners. These methods aim to directly solve Partial Differential Equations (PDEs) that govern robot motion without depending on expert data or requiring extensive training resources. Additionally, the talk will discuss how these PDEs can create a new robot-motion-friendly mapping feature. We demonstrate that this new mapping feature is better suited for fast robot motion generation than existing mapping features, such as occupancy maps or Sign Distance Fields. This talk will demonstrate that these new physics-informed approaches outperform state-of-the-art imitation learning and offline reinforcement learning methods in terms of scalability, training efficiency, data efficiency, computational planning speed, path quality, and success rates.

Biography: Ahmed Qureshi is an Assistant Professor in the Department of Computer Science at Purdue University. Dr. Qureshi is also currently serving as an Associate Editor for IEEE Transactions on Robotics (TRO) and IEEE Robotics and Automation Letters (RA-L). In 2024, he received the Outstanding Associate Editor award from IEEE RA-L. He has previously served on the program committees of RSS, ICRA, IROS, and CoRL.  At Purdue University, Dr. Qureshi directs the Cognitive Robot Autonomy and Learning (CoRAL) Lab. His group conducts fundamental and applied research in robot motion planning and control with the aim of developing robots that can understand the general laws of physics and plan their movements in real time with minimal to no expert demonstrations. His work addresses problems such as scalable and fast motion planning, dexterous manipulation, active perception, and multiagent task and motion planning. Dr. Qureshi's contributions have been recognized with spotlight and best paper awards at various academic venues. Prior to his current roles, he earned his B.S. in Electrical Engineering from NUST, Pakistan, an M.S. in Engineering from Osaka University, Japan, and a Ph.D. in Intelligent Systems, Robotics, and Control from the University of California San Diego.

Host: Daniel Seita

Location: Olin Hall of Engineering (OHE) - 132

Audiences: Everyone (USC) is invited

Contact: CS Faculty Affairs

This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.

Add to Google Calendar