publications

2026

1. 

   Simple recipe works: vision-language-action models are natural continual learners with reinforcement learning

   Jiaheng Hu^*, Jay Shim^*, Chen Tang, and 4 more authors

   Preprint, 2026

   paper abs

   TL;DR: Continual VLAs can be achieved through a simple approach that combines large models, reinforcement learning, and parameter-efficient fine-tuning.

2. 

   ROTATE: regret-driven open-ended training for ad hoc teamwork

   Caroline Wang^*, Arrasy Rahman^*, Jiaxun Cui, and 2 more authors

   Preprint, 2026

   paper abs

   TL;DR: A new perspective on ad hoc teamwork as an open-ended learning problem aimed at producing robust and generalizable agents.

3. 

   Effort allocation for deadline-aware task and motion planning: a metareasoning approach

   Yoonchang Sung^*, Shahaf S. Shperberg^*, Qi Wang^*, and 2 more authors

   Preprint, 2026

   paper abs

   TL;DR: A new class of TAMP problems is introduced by considering deadline constraints, where we prove its hardness and design efficient policy heuristics.

4. 

   Large-language-model-guided state estimation for partially observable task and motion planning

   Yoonwoo Kim, Raghav Arora, Roberto Martín-Martín, and 3 more authors

   IEEE International Conference on Robotics and Automation (ICRA), 2026

   paper project abs

   TL;DR: CoCo-TAMP uses LLM-derived commonsense knowledge to guide belief updates and enable more efficient planning for partially observable TAMP problems.

Before 2026

1. 

   PRESTO: fast motion planning using diffusion models based on key-configuration environment representation

   Mingyo Seo^*, Yoonyoung Cho^*, Yoonchang Sung, and 3 more authors

   IEEE International Conference on Robotics and Automation (ICRA), 2025

   paper project abs

   TL;DR: This work addresses the question of what constitutes the ideal representation for solving motion planning problems as sampling from a diffusion model.

2. 

   Asynchronous task plan refinement for multi-robot task and motion planning

   Yoonchang Sung, Rahul Shome, and Peter Stone

   IEEE International Conference on Robotics and Automation (ICRA), 2024

   paper talk abs

   TL;DR: A new multi-robot task and motion planning formulation that eliminates the need for pre-discretization of the space and synchronous actions among robots.

3. 

   Motion planning (in)feasibility detection using a prior roadmap via path and cut search

   Yoonchang Sung and Peter Stone

   Robotics: Science and Systems (RSS), 2023

   paper poster talk slides abs

   TL;DR: Iterative pathfinding and cut finding effectively reduce the search space, making the determination of whether a prior roadmap contains a feasible solution more efficient.

4. 

   A survey of decision-theoretic approaches for robotic environmental monitoring

   Yoonchang Sung, Zhiang Chen, Jnaneshwar Das, and 1 more author

   Foundations and Trends in Robotics, 2023

   paper abs

   TL;DR: A comprehensive survey on robotic environmental monitoring, covering environmental representations and properties, tasks, and decision-theoretic tools.

5. 

   Learning to correct mistakes: backjumping in long-horizon task and motion planning

   Yoonchang Sung^*, Zizhao Wang^*, and Peter Stone

   Conference on Robot Learning (CoRL), 2022

   paper poster abs

   TL;DR: A heuristic leveraged by supervised learning to guide task-level search, alleviating the need for consistently employing expensive backtracking.

6. 

   Towards optimal correlational object search

   Kaiyu Zheng, Rohan Chitnis, Yoonchang Sung, and 2 more authors

   IEEE International Conference on Robotics and Automation (ICRA), 2022

   paper abs

   TL;DR: Leveraging correlational information among objects allows the POMDP formulation to efficiently discover optimal solutions for object search.

7. 

   Learning when to quit: meta-reasoning for motion planning

   Yoonchang Sung, Leslie Kaelbling, and Tomás Lozano-Pérez

   IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021

   paper abs awarded

   Finalist for Best Cognitive Robotics Paper Award

   TL;DR: Being able to trade-off between plan quality and computation time lets you build a more efficient planner.

8. 

   Multi-resolution POMDP planning for multi-object search in 3D

   Kaiyu Zheng, Yoonchang Sung, George Konidaris, and 1 more author

   IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021

   paper abs awarded

   Winner of Best Robocup Paper Award

   TL;DR: The challenge of intractable partially observable 3D object search can be alleviated through a multi-resolution belief space representation.

9. 

   Reactive task and motion planning under temporal logic specifications

   Shen Li^*, Daehyung Park^*, Yoonchang Sung^*, and 2 more authors

   IEEE International Conference on Robotics and Automation (ICRA), 2021

   paper abs

   TL;DR: When robots work with humans, linear temporal logic allows the modeling of human interventions, while a behavior tree enables the robot to execute reactively.

10. 

    Environmental hotspot identification in limited time with a UAV equipped with a downward-facing camera

    Yoonchang Sung, Deeksha Dixit, and Pratap Tokekar

    IEEE International Conference on Robotics and Automation (ICRA), 2021

    paper abs

    TL;DR: Unlike standard multi-armed bandit settings, the arms’ rewards are correlated, and the reward amount depends on the altitude of a UAV.

11. 

    GM-PHD filter for searching and tracking an unknown number of targets with a mobile sensor with limited FOV

    Yoonchang Sung and Pratap Tokekar

    IEEE Transactions on Automation Science and Engineering (T-ASE), 2021
    IEEE International Conference on Robotics and Automation (ICRA), 2017

    paper abs

    TL;DR: The explicit boundary of limited sensing range is incorporated into the GM-PHD filter, which estimates both the number and spatial density of targets.

12. 

    Game tree search for minimizing detectability and maximizing visibility

    Zhongshun Zhang, Jonathon M. Smereka, Joseph Lee, and 3 more authors

    Autonomous Robots (AURO), 2021
    IEEE International Conference on Robotics and Automation (ICRA), 2019

    paper abs

    TL;DR: Minimax search tree and Monte-Carlo search tree methods for discrete, sequential, two-player, zero-sum games that involve trading off conflicting objectives.

13. 

    Distributed assignment with limited communication for multi-robot multi-target tracking

    Yoonchang Sung, Ashish Budhiraja, Ryan Williams, and 1 more author

    Autonomous Robots (AURO), 2020
    IEEE International Conference on Robotics and Automation (ICRA), 2018

    paper abs

    TL;DR: The reduction of multi-robot multi-target tracking problems to mixed packing and covering problems enables the use of local algorithms for distributed problem-solving.

14. 

    A competitive algorithm for online multi-robot exploration of a translating plume

    Yoonchang Sung and Pratap Tokekar

    IEEE International Conference on Robotics and Automation (ICRA), 2019

    paper abs

    TL;DR: A recursive depth-first-search algorithm for exploring the unknown shape of a moving plume, with provable competitive ratio guarantees.

15. 

    Bayesian estimation based real-time fire-heading in smoke-filled indoor environments using thermal imagery

    Jong-Hwan Kim, Yoonchang Sung, and Brian Lattimer

    IEEE International Conference on Robotics and Automation (ICRA), 2017

    paper abs

    TL;DR: The FIRE framework plans robot paths to efficiently detect and extinguish fires by jointly considering search, sensing, and action decisions under uncertainty.

16. 

    Team VALOR’s ESCHER: A novel electromechanical biped for the DARPA Robotics Challenge

    Yoonchang Sung

    Journal of Field Robotics (JFR), 2017

    paper abs

    TL;DR: Virginia Tech’s electric series compliant humanoid participated in the finals of the DARPA Robotics Challenge.

17. 

    Hierarchical sample-based joint probabilistic data association filter for following human legs using a mobile robot in a cluttered environment

    Yoonchang Sung and Woojin Chung

    IEEE Transactions on Human-Machine Systems (T-HMS), 2016

    paper abs

    TL;DR: Correlations among targets are explicitly modeled for target tracking, a consideration that is generally treated independently in the tracking literature.