Science Café – Artificial Intelligence in the Physical World
Wednesday, October 21, 2026 from 6:30 pm to 8:00 pm
- In-person event
- 1049 Bank St , Ottawa , ON , K1S 3W9
- Contact
- odscomms@carleton.ca
While Artificial Intelligence (AI) and Machine Learning have revolutionized our digital world, their reliable deployment in the physical world remains an open challenge. Unlike digital tasks, most physical world tasks involve sequential decision-making across many interacting agents, where collecting clear labels of correct behaviour is difficult and mistakes are costly. Reinforcement Learning is an active area of research that enables autonomous agents to learn directly from experience, and Multi-agent Reinforcement Learning extends this to settings with more than one collaborating or competing agents. However, these approaches are notoriously data-hungry and expensive to train, posing significant barriers to real-world deployment. This talk will describe these challenges and present our research on enabling AI agents to learn efficiently from external knowledge sources, accelerating training, and improving performance, with applications in autonomous driving and robotics.
About the Speaker
Sriram Ganapathi Subramanian is an Assistant Professor at the School of Computer Science at Carleton University, where he holds a Canada Research Chair (Tier II) in Artificial Intelligence. He is also a Faculty Affiliate at the Vector Institute for Artificial Intelligence, Toronto, and the Schwartz Reisman Institute for Technology and Society, Toronto. He received his Ph.D. in Electrical and Computer Engineering from the University of Waterloo and was a Distinguished Postdoctoral Fellow at the Vector Institute for Artificial Intelligence, before joining Carleton University in July 2025. His doctoral dissertation received the Best Doctoral Dissertation Award from the Canadian AI Association. Sriram’s research focuses on developing principled algorithms at the intersection of Multi-agent Reinforcement Learning, Game Theory, and Deep Learning, with applications in autonomous systems, robotics, and wildland fire management. His work spans both theoretical foundations and empirical validation, with the goal of designing new AI algorithms that can operate reliably and safely in the physical world, accelerating their deployment in many real-world applications.
Personal Website: https://sriramsubramanian.com/
Carleton Profile: https://carleton.ca/scs/people/sriram-subramanian/
Register for Science Café: Artificial Intelligence in the Physical World October 21
Personal information collected through this form will be used and disclosed by Carleton University under the authority of the Carleton University Act, and in accordance with sections 39, 41 and 42 of Ontario’s Freedom of Information and Protection of Privacy Act. The purpose of this processing is for advancement activities and alumni development purposes. If you have any questions about the processing of personal information by Carleton University, please contact the Manager, Privacy & Access to Information, by phone at 613-520-2600 ext. 2047 or by e-mail via University_Privacy_Office@carleton.ca. You may also visit carleton.ca/privacy for additional information and notices.