Towards Robust Execution of Long-Horizon Whole-Body Control Tasks

RSS 2026 Workshop
July 13, Morning, Sydney, Australia

Introduction

Recent progress in robot learning has significantly advanced robotic capabilities on short-horizon skills and well-defined tasks. However, despite these advances, robots, such as humanoids and robotic arms, continue to struggle when deployed on long-horizon, complex tasks in the real world. Topics of interest of this workshop include on-policy progress estimation for multi-stage tasks, run-time failure recovery, hierarchical and memory-augmented policies trained under long-horizon rollouts, sim-to-real transfer with execution-time fine-tuning, and leveraging foundation models for task planning, perception, and feedback within closed-loop, on-policy execution.

Join Discord to chat with the organizers.

Call for Contributions

We invite researchers to share their work with the community through submissions to the workshop in a variety of formats beyond traditional papers, including reports, demos, video, and etc. Submissions may include research papers or reports, but we equally welcome alternative formats such as videos demonstrating systems in action, demos, interactive artifacts, or other creative presentations of research ideas. We particularly welcome ongoing, preliminary, or exploratory work.

Topics of Interest

We welcome works on a wide range of topics, including but not limited to:

Long-horizon robot learning and multi-stage task execution.
Whole-body control for high-dimensional, multi-contact robotic systems.
Loco-manipulation and mobile manipulation in unstructured environments.
On-policy progress estimation under partial observability.
Task decomposition and sub-goal discovery for complex behaviors.
Run-time failure detection, recovery, and robustness in real-world deployment.
Closed-loop decision-making with feedback-driven adaptation.
Integration of learning, planning, and control for sequential decision-making.
Data efficiency in robot learning, including on-policy data collection and hybrid offline-online training.
Metrics and evaluation protocols for long-horizon performance, robustness, and generalization.

Guidelines

All contributions must be submitted through OpenReview.
Manuscripts are required to use the LaTex or Word template.
No strict page length requirements on submissions.
To facilitate double-blind review, all submissions must be fully anonymized.
All accepted contributions will be made available online on this workshop website as non-archival reports.
Selected contributions will be invited for presentations.

Timeline

For any potential ambiguities, please refer to OpenReview.

Submission start: May 10, 2026
Submission end: June 21, 2026
Notification: July 01, 2026

Schedule

Hongyang Li
The University of Hong Kong, China
08:30 AM
Opening Remarks
Tianyu Li
Archon Robotics, China
08:35 AM
TBD
Biography
Tianyu Li is the co-founder of Archon Robotics and a PhD researcher affiliated with Fudan University, Shanghai Innovation Institute, and OpenDriveLab. His research focuses on Physical AI, autonomous driving, end-to-end autonomous driving systems, 3D reconstruction, and generative models.
Yi Li
Tsinghua University, China
09:05 AM
Beyond Imitation: Executable, Correctable, and Adaptable Skills for Humanoid Robots
Biography
Li Yi is a tenure-track assistant professor at the Institute for Interdisciplinary Information Sciences (IIIS), Tsinghua University, and Chief Scientist at Beijing Galbot Co., Ltd. He received his Ph.D. from Stanford University, advised by Professor Leonidas J. Guibas. And he was previously a Research Scientist at Google. Before joining Stanford, he got his B.E. in Electronic Engineering from Tsinghua University. His recent research focuses on 3D computer vision, humanoid robot learning, and dexterous manipulation, and his mission is to equip robotic agents with the ability of understanding and interacting with the 3D world. He has published papers at top-tier computer vision, computer graphics, and machine learning conferences with more than 35000 citations. And he has served as an Area Chair for CVPR, IJCAI, and NeurIPS. His representative work includes ShapeNet, PointNet++, and HOI4D.
Selected Presenters
09:35 AM
Spotlight Presentations
All Participants
09:50 AM
Breakout Discussion
Coffee Break
10:00 AM
Fan Shi
National University of Singapore, Singapore
11:00 AM
Simulation Enabled Robust Locomotion Learning. Can It Do the Same for Manipulation?
Biography
Fan Shi is an Assistant Professor at the National University of Singapore, where he holds the prestigious NUS Presidential Young Professorship. His research lies at the intersection of artificial intelligence and robotics, with a focus on physical simulation, robot learning, and the development of scalable methods for embodied intelligence. His work has been recognized through awards and support from leading organizations, including the NVIDIA Academic Grant Program, Google Research, and the Swiss AI Initiative.
Marcel Torne(remote)
Stanford University & Physical Intelligence, USA
11:30 AM
TBD
Biography
Marcel Torne is a researcher at Stanford and Physical Intelligence. His research focuses on learning-based assistive robots and methods for in-context adaptation of policies to unseen scenarios with an emphasis on human-centric approaches.
Tapomayukh Bhattacharjee
Cornell University, USA
12:00 PM
TBD
Biography
Tapomayukh "Tapo" Bhattacharjee is an Assistant Professor in the Department of Computer Science at Cornell University, where he directs the EmPRISE Lab. His research aims to enable robots to assist people with mobility limitations with activities of daily living, spanning human-robot interaction, haptic perception, and robot manipulation. He received his PhD in Robotics from Georgia Institute of Technology and was an NIH Ruth L. Kirschstein NRSA postdoctoral research associate at the University of Washington.