Generalizing Embodied AI Beyond the Lab: Commonsense, Affordances, Steering and Safety
What does it take for embodied AI to work reliably beyond controlled environments?
Dr. Cheston Tan will explore how embodied AI systems can generalize beyond the lab, where robots must contend with unfamiliar situations, physical uncertainty, and failures that are difficult to anticipate during development.
Drawing from published and ongoing work at A*STAR's Institute of Advanced Intelligence & Computing (A*STAR IAIC), he'll explore four complementary capabilities for more robust generalization:
Physical commonsense and intuitive physics
Affordance understanding
Deployment-time steering
Safety and failure recovery
He'll discuss how these capabilities can help embodied AI systems better understand and adapt to the physical world, respond when things do not go as expected, and move towards more reliable operation in open and unpredictable environments.
More About the Speaker
Dr. Cheston Tan is a Senior Principal Scientist and Head of Embodied AI & Robotics at A*STAR's Institute of Advanced Intelligence & Computing (A*STAR IAIC).
He obtained his PhD from the Department of Brain and Cognitive Sciences at MIT and his BSc with Highest Honours from the Department of Electrical Engineering and Computer Science at UC Berkeley. He serves as Senior Area Chair for NeurIPS and ICML and Area Chair for ICLR, and was previously an Associate Editor for IEEE Transactions on Cognitive and Developmental Systems.
His awards and honours include the Ubiquitous Robotics conference Best Application Paper, CVPR Outstanding Reviewer, STAR Mentor Award and NeurIPS Travel Award. At MIT, he received the Mark Gorenberg Graduate Student Fellowship and Walle Nauta Award for Excellence in Graduate Teaching, and at UC Berkeley, the Chancellor Tien Award for Engineering and Computer Science.
More About the Series
Singapore Embodied AI Colloquium (SEAIC) is a focused technical research series built around researcher-driven discussions and sharing on frontier topics in embodied AI, robot learning, and physical AI.
Get involved: Learn more about Lorong AI | Speaker Sign-up | WhatsApp Community | LinkedIn | X