Event
Future Leaders in Robotics and AI Seminar: Yoonchang Sung
Friday, February 28, 2025
2:00 p.m.-2:30 a.m.
https://go.umd.edu/FutureLeaders
Yancy Diaz-Mercado
301 405 6506
yancy@umd.edu
Extending Current Capabilities of Task and Motion Planning
Yoonchang Sung
Postdoctoral Fellow
The University of Texas at Austin
Abstract
Solving long-horizon tasks involving the manipulation of multiple objects, such as cleaning a cluttered room or preparing a meal, has long been a grand challenge in robotics. These tasks require selecting high-level actions, such as picking up a cup, and effectively planning how to execute these actions while satisfying physical and geometric constraints. Task and motion planning (TAMP) is a general framework for solving such tasks, characterized by a bilevel structure: high-level task reasoning (determining which actions to take) and low-level motion reasoning (determining how to execute these actions), which provide complementary guidance. Most existing research focuses either on improving the efficiency of computationally expensive planning processes or on learning the models required for TAMP planners.
In this talk, I will introduce a different research perspective aimed at extending the current capabilities of TAMP planners to address more complex problem classes by introducing additional constraints. The first extension is deadline-aware TAMP, where the robot must allocate its time among multiple options to compute an executable plan that meets a pre-specified deadline. I will discuss the hardness of this problem class and present two heuristics that enable efficient policy computation. The second extension is multi-robot TAMP, where a small group of robots collaborates to achieve a task. This class inherently introduces challenges such as robot-robot collisions and asynchronous mode switches among robots. I will present an efficient asynchronous multi-robot TAMP planner. Finally, I will conclude by discussing future research challenges associated with these extensions.
Biography
Yoonchang Sung is an incoming Assistant Professor at the College of Computing and Data Science, Nanyang Technological University (NTU), Singapore, starting Summer 2025. He is currently a postdoctoral fellow at UT Austin, working with Peter Stone, and was previously a postdoctoral associate at MIT CSAIL, working with Leslie Kaelbling and Tomás Lozano-Pérez. He completed his Ph.D. at Virginia Tech, advised by Pratap Tokekar. His research interests include robot planning and learning, task and motion planning, and multi-robot systems. He was selected as an RSS Pioneer (2019). His work was nominated for the Best Cognitive Robotics Paper Award and received the Best Robocup Paper Award, both at IROS 2021. Website: https://yoonchangsung.com/
About the Seminar Series
The Maryland Robotics Center's Future Leaders in Robotics and AI Seminar Series at the University of Maryland is a nationwide online seminar series for PhD students, postdoctoral researchers, and early-career professionals. The seminar series highlights the latest research and innovation in the field of robotics and AI. The series is intended to provide exposure and mentorship opportunities to the speakers, build a network of innovators across the country, and support the speakers’ career planning.
The seminars are held on select dates during spring. There are two speakers per seminar. Each speaker gives a 20-minute research presentation followed by a Q&A segment. Immediately after the second seminar, the speakers participate in a discussion with faculty. For more information, please visit: https://robotics.umd.edu/FutureLeaders
