Andrey Kurenkov
Stanford University, Georgia Institute of Technology, Stanford Health Care
Papers
9
Total Citations
449
H-Index
7
About
Andrey Kurenkov is a leading researcher in robot learning, with a focus on task-oriented manipulation, mechanical search, and simulation for embodied AI. His work bridges the gap between grasping and task reasoning, most notably through his highly cited 2019 paper "Learning task-oriented grasping for tool manipulation from simulated self-supervision" (185 citations), which pioneered methods for robots to reason about a tool's desired effect rather than just grasp stability. He also introduced the concept of "Mechanical Search" (108 citations), enabling robots to retrieve occluded objects from cluttered environments—a critical capability for real-world deployment in warehouses and homes. Kurenkov contributed to the development of iGibson 2.0 (62 citations), an object-centric simulation platform that has become a key resource for training household robot skills. His work on AC-Teach (17 citations) advanced sample-efficient deep reinforcement learning by leveraging ensembles of suboptimal teachers. With additional contributions in deformable object reconstruction and error-aware imitation learning, Kurenkov's research consistently addresses the practical challenges of deploying robots in unstructured, human-centric environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2Mechanical Search: Multi-Step Retrieval of a Target Object Occluded by Clutter108 citations · 2019
- 3
- 4
- 5
- 6
- 7
- 8
- 9