AI & ML Breaks Assumption

Zero-shot sim-to-real transfer for complex robotic manipulation is achievable using only synthetic simulated data at scale.

arXiv · March 18, 2026 · 2603.16861

Abhay Deshpande, Maya Guru, Rose Hendrix, Snehal Jauhri, Ainaz Eftekhar, Rohun Tripathi, Max Argus, Jordi Salvador, Haoquan Fang, Matthew Wallingford, Wilbert Pumacay, Yejin Kim, Quinn Pfeifer, Ying-Chun Lee, Piper Wolters, Omar Rayyan, Mingtong Zhang, Jiafei Duan, Karen Farley, Winson Han, Eli Vanderbilt, Dieter Fox, Ali Farhadi, Georgia Chalvatzaki, Dhruv Shah, Ranjay Krishna

The Takeaway

It challenges the consensus that real-world data or fine-tuning is required for effective robotic deployment. By releasing a pipeline for 1.8M expert trajectories and demonstrating zero-shot success on diverse platforms, it provides a blueprint for scaling robotics through simulation alone.

From the abstract

A prevailing view in robot learning is that simulation alone is not enough; effective sim-to-real transfer is widely believed to require at least some real-world data collection or task-specific fine-tuning to bridge the gap between simulated and physical environments. We challenge that assumption. With sufficiently large-scale and diverse simulated synthetic training data, we show that zero-shot transfer to the real world is not only possible, but effective for both static and mobile manipulati