| |
 Pretraining Foundation Models for Embodied AI Speaker: 
Tasbolat TAUNYAZOV
Team Member
Light Robotics
| Date: Time: Venue: | | 13 November 2026, Friday 11:00 am - 12:00 pm School of Computing & Information Systems 1 (SCIS 1)
Level 4, Meeting Room 4-4
80 Stamford Road
Singapore Management University
S(178902) 
|
|
About the Talk
How can robots learn from large and diverse datasets to perform a wide range of tasks in the physical world? This talk explores pretraining foundation models for embodied AI, with a focus on vision-language-action (VLA) models that connect visual observations and language instructions to robot actions. It will introduce the role of pretraining and discuss key considerations in data collection, data quality, model design, and training objectives. Drawing on practical experience in developing robotic learning systems, the talk will examine the challenges of transferring learned capabilities to real robots and generalizing across tasks and environments. It will conclude with open questions and future directions for building more capable and adaptable embodied agents.
About the Speaker
Dr. Tasbolat TAUNYAZOV is a robotics researcher specializing in robotic foundation models, tactile perception, and Vision-Language-Action models. He earned his Ph.D. from the National University of Singapore, where his doctoral research received the NUS SoC Computer Science Thesis Award and IROS Best Paper Award. He currently builds foundational models for robotics at Light Robotics, where he is leading the technical team.
|
