Karol Hausman
Karol Hausman | |
|---|---|
| Occupations | Computer scientist, robotics researcher |
| Employer | Stanford University |
| Organization | Physical Intelligence Inc. |
| Known for | Robot learning, simulation-to-real transfer, embodied AI |
| Academic background | |
| Education | University of Southern California (PhD) Technical University of Munich (MS) Warsaw University of Technology (MS, BS) |
| Academic work | |
| Discipline | Computer science |
| Sub-discipline | Robotics, Artificial intelligence, Machine learning |
| Institutions | Stanford University Google Brain Google DeepMind |
Karol Hausman is a professor, researcher and technology executive known for his work on robot learning and artificial intelligence.[1][2]
Early life
Hausman received a Ph.D. in Computer Science from the University of Southern California, a Master of Science in Robotics from the Technical University of Munich, a Master of Science and a Bachelor of Science in Mechatronics from the Warsaw University of Technology.[3]
Research career
Hausman's work is focused on creating an AI that can power any robot.[4] He has been serving as an adjunct professor at Stanford University since 2021. He researched perception–action loops and interactive perception, showing how robots can actively move to reduce uncertainty in object pose estimation by coupling probabilistic state estimation with control.[5]
In Scaling Simulation-to-Real Transfer by Learning a Latent Space of Robot Skills, published in The International Journal of Robotics Research (2020), Hausman and collaborators proposed learning a low-dimensional latent space of skills in simulation and adapting them efficiently to physical robots. The method improved transfer efficiency and reduced the amount of real-world data required for locomotion and manipulation tasks.[6]
At Google Brain and Google DeepMind, he worked as a senior researcher on large-scale robotic learning systems that trained generalizable grasping and manipulation policies from extensive robot interaction data.[7][8] His work emphasizes shared representations and scalable learning architectures for general-purpose robotic intelligence.[2][9]
He served as an chair/associate editor of the International Conference on Robotics and Automation,[10] International Conference on Learning Representations, Neural Information Processing Systems, International Conference on Machine Learning.[11][12]
Physical Intelligence
Hausman co-founded Physical Intelligence, a technology company that aims to develop foundation models and learning algorithms for physically embodied AI systems.[13]
References
- ^ Busoniu, Lucian; Tamás, Levente, eds. (2015). Handling Uncertainty and Networked Structure in Robot Control. Studies in Systems, Decision and Control. Vol. 42. Cham: Springer International Publishing. doi:10.1007/978-3-319-26327-4. ISBN 978-3-319-26325-0.
- ^ a b "Karol Hausman". scholar.google.com. Retrieved 2026-02-28.
- ^ "Trojan Trio Make Waves at Google DeepMind". USC Viterbi | School of Engineering. Retrieved 2026-03-01.
- ^ Mims, Christopher (2024-11-16). "A Powerful AI Breakthrough Is About to Transform the World". The Wall Street Journal. Retrieved 2026-03-05.
- ^ Julian, Ryan C; Heiden, Eric; He, Zhanpeng; Zhang, Hejia; Schaal, Stefan; Lim, Joseph J; Sukhatme, Gaurav S; Hausman, Karol (2020-09-09). "Scaling simulation-to-real transfer by learning a latent space of robot skills". The International Journal of Robotics Research. 39 (10–11): 1259–1278. doi:10.1177/0278364920944474. ISSN 0278-3649.
- ^ Julian, Ryan C; Heiden, Eric; He, Zhanpeng; Zhang, Hejia; Schaal, Stefan; Lim, Joseph J; Sukhatme, Gaurav S; Hausman, Karol (2020). "Scaling simulation-to-real transfer by learning a latent space of robot skills". The International Journal of Robotics Research. 39 (10–11): 1259–1278. doi:10.1177/0278364920944474. ISSN 0278-3649.
- ^ "Towards Helpful Robots: Grounding Language in Robotic Affordances". Research Google Brain. Retrieved 2026-02-28.
- ^ Roose, Kevin (2023-07-28). "Aided by A.I. Language Models, Google's Robots Are Getting Smart". The New York Times. ISSN 0362-4331. Retrieved 2026-03-03.
- ^ Preiss, James A; Hausman, Karol; Sukhatme, Gaurav S; Weiss, Stephan (2018-12-09). "Simultaneous self-calibration and navigation using trajectory optimization". The International Journal of Robotics Research. 37 (13–14): 1573–1594. doi:10.1177/0278364918781734. ISSN 0278-3649.
- ^ Hausman, Karol; Preiss, James; Sukhatme, Gaurav S.; Weiss, Stephan (2017-07-07). "Observability-Aware Trajectory Optimization for Self-Calibration With Application to UAVs". IEEE Robotics and Automation Letters. 2 (3): 1770–1777. arXiv:1604.07905. doi:10.1109/LRA.2017.2647799. ISSN 2377-3766.
- ^ "ICML Poster Actionable Models: Unsupervised Offline Reinforcement Learning of Robotic Skills". icml.cc. Retrieved 2026-03-05.
- ^ https://karolhausman.github.io/cv.pdf
- ^ "Training General Robots for Any Task: Physical Intelligence's Karol Hausman and Tobi Springenberg". Sequoia Capital. Retrieved 2026-03-01.