MIT Robotics Seminar - Greg Stein

"Learning, Introspection, and Anticipation: Making Robot Planners Comfortable with Missing Knowledge"

The next generation of service and assistive robots will need to operate under uncertainty, expected to complete tasks and perform well despite missing information about the state of the world or over their future. Many emerging approaches turn to learning to overcome the challenges of planning under uncertainty, yet can be brittle and myopic, limiting their effectiveness. Our work introduces a family of model-based approaches to long-horizon planning under uncertainty that augments planning with estimates from learning, allowing both high-performance and reliability. I will present several ongoing projects aimed at improving long-horizon planning in uncertain environments for both single-robots and multi-robot teams. Alongside these capabilities, we have also developed "anticipatory planning," in which a robot estimates how its actions now may impact tasks it may later be assigned. With such anticipation, preparatory and organization-focused behaviors emerge, improving performance over lengthy deployments. Finally, our learning-augmented planning abstractions afford introspection via counterfactual reasoning, allowing fast and reliable deployment-time policy selection and thus improved performance even in unfamiliar environments dissimilar from any seen during training.

Greg is an Assistant Professor of Computer Science at George Mason University, where he runs the Robotic Anticipatory Intelligence & Learning (RAIL) Group and co-directs the GMU Autonomous Robotics Lab. His research, at the intersection of robotics, planning, and machine learning, is centered around developing representations for planning and learning that allow robots to better understand the impact of their actions, so that they may plan quickly, intelligently, and reliably in a dynamic and uncertain world. Before joining Mason, he received his PhD in 2020 from the MIT Department of Electrical Engineering and Computer Science and previously graduated summa cum laude from Cornell University with a B.S. in Applied and Engineering Physics. His work was a finalist for Best Paper at the 2018 Conference on Robot Learning, at which he was additionally awarded Best Oral Presentation. His work has been supported by the US Army Research Lab, National Science Foundation, and US Air Force. 

The MIT Robotics Seminars have returned this Spring! Juraj Kabzan from Skydio kicked us off a few weeks ago. Additionally, we have an outstanding lineup of speakers planned for the rest of the semester, and we invite all of you to attend the seminars and meet the speakers and colleagues:

• Friday, March 6, 2026 — Greg Stein (George Mason University)
• Friday, March 13, 2026 — Oussama Khatib (Stanford University)
• Friday, April 3, 2026 — Mac Schwager (Stanford University)
• Friday, April 10, 2026 — Melanie Zeilinger (ETH Zurich)
• Friday, April 24, 2026 — Seth Hutchinson (Northeastern University)
• Friday, May 1, 2026 — Giuseppe Loianno (UC Berkeley)
• Friday, May 8, 2026 — Sanjiban Choudhury (Cornell University)

When & Where:
All seminars will be held on Fridays, 3-4 pm, in 45-230. A social event will follow the seminars at 4 pm!

Format:
All seminars will be in person. We will not live-stream the events, but we will record the talks as usual and release them on the MIT Robotics YouTube Channel: https://www.youtube.com/c/MITRoboticsSeminar/videos
(please like/subscribe/share!)

GWiRC: 
Some of the Friday lunches with the speakers will be held in conjunction with the Graduate Women in Robotics Community (GWiRC), an awesome association that organizes a variety of events at MIT and beyond.
You can learn more and subscribe to their mailing list here: https://robotics.mit.edu/graduate-women-robotics-community

Sponsors: 
A big shoutout to our sponsors, who are making this semester’s robotics seminars possible! We are pleased to have three amazing sponsors supporting the MIT Robotics Seminars this semester: Skydio, Symbotic, and Amazon. Skydio (https://www.skydio.com/) is a leader in vision-based autonomous navigation (and more!) for drones. Symbotic (https://www.symbotic.com/) is redesigning the future of warehouse automation with mobile robots (just a few miles from MIT!). Amazon (https://www.amazon.science/research-areas/robotics) is building new kinds of GenAI for robotics as they grow towards 1M deployed warehouse robots.

Admins: 
A big THANK YOU to our amazing Robotics Seminars admin, Vendi Pavic, who is handling the logistics of the seminars and making sure they run smoothly.