CoRL 2026 will put robot learning's founder pipeline in Austin this November
Yuke Zhu and Peter Stone are chairing the Nov. 9-12 conference as robotics labs race to turn foundation models into machines that work.
By Ryan Merket · Published
Why it matters
CoRL is where robot learning research turns into hiring pipelines, startup theses and technical credibility for physical AI companies.
Yuke Zhu and Peter Stone will bring CoRL 2026 to Austin from November 9 to 12, giving robot learning researchers, robotics founders and industrial AI labs a four-day checkpoint for a field that has moved from demos to company formation.
The Conference on Robot Learning says its 2026 edition will run workshops on November 9 and the main conference November 10 to 12 at JW Marriott Austin, 110 E 2nd St. The schedule is still light on accepted papers, pricing and capacity, but the program's center of gravity is already clear: robot foundation models, real-robot reinforcement learning, imitation learning, world models, safety, benchmarks and the messy transfer from simulation to deployed hardware.
Zhu and Stone are a telling pair for the Austin edition. Zhu is an associate professor of computer science at the University of Texas at Austin, director of the Robot Perception and Learning Lab, and a Director and Distinguished Research Scientist at NVIDIA Research, where he co-leads the Generalist Embodied Agent Research group. On his homepage, Zhu describes his research goal as building algorithms and systems for autonomous robots and embodied agents that reason about and interact with the real world. Stone is chair of UT Austin's computer science department, founding director of Texas Robotics and chief scientist of Sony AI. His own research statement puts adaptation, interaction and embodiment at the center of complete intelligent agents.
That makes CoRL 2026 less a calendar listing than a map of where robotics talent is clustering. The conference is governed by the Robot Learning Foundation, whose board lists Ken Goldberg of UC Berkeley as president. CoRL's first edition was held in Mountain View in November 2017, with Ken Goldberg, Sergey Levine and Vincent Vanhoucke as general co-chairs. Since then, CoRL has rotated across Europe, Asia and North America before landing in Austin for 2026.
The program is being written around embodied AI
The call for papers frames CoRL as a venue for original research at the intersection of robotics and machine learning. The topic list reads like the technical backlog for physical AI startups: learning representations for perception and control, robot foundation models, imitation learning, reinforcement learning for control of physical robots, automatic robotic data generation, robot task and motion planning, human-robot interaction, robot safety and alignment, video models, latent world models, benchmarks and datasets.
CoRL's CFP and author instructions emphasize a robotics focus and sim-to-real relevance. Authors are encouraged to report real-robot experiments or provide convincing evidence that simulation results transfer to physical systems.
Accepted papers will appear in the Proceedings of Machine Learning Research. The conference site notes that the author instructions were updated this spring with a generative AI policy.
For robotics founders, those details matter because CoRL is one of the places where hiring, partnerships and technical narratives get stress-tested before they turn into sales decks. A model that cannot leave the simulator, a benchmark that hides failure modes, or a demo without reproducibility all face a harsher audience here than in a product launch video.
Austin gives CoRL a local research base
The organizing committee lists Zhu and Stone as general chairs, with Ani Majumdar of Princeton, Dieter Fox of the University of Washington and AI2, and Georgia Chalvatzaki of TU Darmstadt as program chairs.
Austin gives CoRL a host city with a robotics research institution already tied into the general chairs. Stone founded Texas Robotics, while Zhu runs the Robot Perception and Learning Lab at UT Austin and also holds a senior research role at NVIDIA. The venue is a large downtown convention hotel suited to the poster-heavy, demo-heavy meeting that robot learning has become.
The keynote lineup posted so far lists Russ Tedrake (MIT), Fei-Fei Li (Stanford; World Labs) and Wolfram Burgard (University of Technology Nuremberg).
What is still missing
CoRL 2026 has posted the venue, workshop and conference dates, call for papers, workshop call, sponsor call, keynote page and organizing committee. It has not posted the 2026 registration fees, attendee capacity, accepted-paper count, acceptance rate or named sponsors.
Those omissions are normal four months before a November research conference, but they limit what can be inferred. The conference can already tell founders where the field wants stronger evidence: real-world transfer, reproducibility, safety and limitations. It cannot yet tell them which specific methods or labs will dominate the 2026 program.
The sponsor list will also be worth watching when it appears. CoRL sits in the channel between university labs and commercial robotics companies. The organizations willing to underwrite it often reveal which industrial players are recruiting, which labs want visibility with graduate students, and which parts of robot learning are becoming a budget line rather than a research bet.
For now, the date is fixed. CoRL 2026 begins on November 9 in Austin, with Zhu and Stone chairing a conference that will measure how much of the physical AI boom can survive contact with robots, reviewers and the real world.