World Congress 2026 North America • Sep 25, 2026 • Session details

From Simulation to Reality: Overcoming the Data Challenge in Physical AI

Mitesh Patel

How do we train generalist robots without millions of slow, expensive real-world hours? Discover how massive-scale synthetic data and foundation models bridge the sim-to-real gap.

From Simulation to Reality: Overcoming the Data Challenge in Physical AI thumbnail

Checking access…

Playback and chapters load privately for Free videos.

Matching moments

2:22 min

Introduction to physical AI and the physical world

Moe Sani Moe Sani · World Congress 2026 Europe

2:11 min

Data and simulation requirements for AI robotics

Teresa Conceicao · World Congress 2022

1:54 min

Overcoming physical AI data scarcity through synthetic generation

Sergio Perez Sergio Perez · World Congress 2026 Europe

1:20 min

Shifting focus from large language models to physical AI

Sergio Perez Sergio Perez · World Congress 2026 Europe

2:58 min

Overcoming the lack of training data in physical environments

Ashutosh Saxena Ashutosh Saxena · World Congress 2026 North America

2:23 min

Development roadmap for world models and physical robotics

Joerg Krall Joerg Krall · World Congress 2026 Europe