Isabel Leal
Papers
9
Total Citations
1,087
H-Index
7
About
Isabel Leal is a pioneering researcher at the intersection of robotics, machine learning, and foundation models, whose work has fundamentally advanced how robots learn to interact with the physical world. She is best known for her central contributions to Google's Robotics Transformer series, including RT-1 (550+ citations across versions) and RT-2, which demonstrated that large-scale, diverse datasets and vision-language models trained on internet-scale data could be directly applied to real-world robotic control, enabling zero-shot generalization and emergent semantic reasoning. Her involvement in the Open X-Embodiment initiative further pushed boundaries by consolidating robotic learning datasets across multiple platforms into unified, generalizable models. Leal has also explored prompting strategies for robot manipulation through code-based policies and contributed to AutoRT, a framework for large-scale orchestration of robotic agents using embodied foundation models. Her most recent work on Gemini Robotics represents a bold step toward truly generalist physical AI. With hundreds of citations accumulated in just a few years, Leal's research is shaping the trajectory of embodied AI, making her one of the most impactful emerging voices in modern robotics.
Research Focus
Key Achievements
Top Papers
- 1RT-1: Robotics Transformer for Real-World Control at Scale512 citations · 2023
- 2RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control267 citations · 2023
- 3
- 4Open X-Embodiment: Robotic Learning Datasets and RT-X Models101 citations · 2023
- 5RT-1: Robotics Transformer for Real-World Control at Scale38 citations · 2022
- 6
- 7
- 8
- 9Gemini Robotics: Bringing AI into the Physical World4 citations · 2025