Papers
9
Total Citations
256
H-Index
5
About
Abhishek Padalkar is a leading researcher in robotic learning and manipulation, whose work bridges the gap between data-driven AI and real-world physical interaction. His primary research areas include reinforcement learning for compliant manipulation, robotic assembly, and large-scale cross-embodiment learning. Padalkar’s most significant contribution is his pivotal role in the Open X-Embodiment collaboration, which produced two highly cited papers (119 and 101 citations) that introduced massive, diverse robotic datasets and the RT-X models. This work demonstrated that training high-capacity models on heterogeneous robot data can dramatically improve generalization and efficiency, marking a paradigm shift toward foundation models for robotics. He has also made notable advances in combining task frame formalism with reinforcement learning for dexterous skills like vegetable cutting, and in using Shared Control Templates to safely guide RL in contact-rich tasks. His research on flexible robotic assembly, grounded in ontological task representation, further showcases his ability to tackle complex industrial challenges. With over 250 total citations and a growing portfolio of impactful publications, Padalkar is establishing himself as a key figure in the movement to make robots more adaptable, data-efficient, and capable of operating in unstructured human environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2Open X-Embodiment: Robotic Learning Datasets and RT-X Models101 citations · 2023
- 3
- 4
- 5
- 6
- 7
- 8Guiding Reinforcement Learning with Shared Control Templates2 citations · 2023
- 9