Jan Peters
Papers
14
Total Citations
304
H-Index
6
About
Jan Peters is a prominent robotics researcher whose work spans robot learning, motion optimization, manipulation, and the integration of machine learning with robotic control. His contributions have significantly advanced how robots acquire and generalize complex skills across diverse environments. Among his most impactful work is his participation in the **Open X-Embodiment** collaboration (2024, 119 citations), which demonstrated that large-capacity models trained on diverse robotic datasets could dramatically accelerate downstream task performance — a landmark step toward generalizable robot intelligence. His research on **SE(3)-DiffusionFields** (2023, 89 citations) introduced elegant diffusion-based cost functions for jointly optimizing grasping and motion planning, tackling some of the field's most persistent multi-objective challenges. Peters has also made notable contributions to reactive motion generation through signed distance fields, implicit motion priors, and neuro-symbolic imitation learning — collectively pushing robots toward safer, more adaptable real-world operation. His earlier work on reinforcement learning for motor primitives reflects a sustained career-long commitment to bridging learning theory and physical robot control. With hundreds of citations across foundational and applied research, Peters stands as an influential voice shaping the future of intelligent, autonomous robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Regularized Deep Signed Distance Fields for Reactive Motion Generation32 citations · 2022
- 4Learning Implicit Priors for Motion Optimization21 citations · 2022
- 5
- 6
- 7Active Exploration for Robotic Manipulation5 citations · 2022
- 8Extended Tree Search for Robot Task and Motion Planning4 citations · 2024
- 9Controlling the Cascade: Kinematic Planning for N-ball Toss Juggling3 citations · 2022
- 10