Andrey Kolobov
Andrey Kolobov is a Principal Researcher and Research Manager at Microsoft Research in Redmond, Washington, where he leads the Physically Embodied AI & Robotics (PEAR) group. He focuses on decision-making for physically embodied AI, with emphasis on general-purpose robotic manipulation and large model-based approaches. His research spans online and offline reinforcement learning, imitation learning, planning, scheduling, and model-predictive control, and has been applied to systems including robotic manipulators, drones, the Edge browser, and Bing's web crawler. He is known for leading Microsoft's Rho-alpha robotics model project and the company's research partnerships in the humanoid robotics space.
“Andrey Kolobov is the project lead for Rho. His leadership drives the strategic separation of embodiment and task adaptation, which is the core innovation of the paper.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.