Papers
19
Total Citations
610
H-Index
10
About
Eugen Solowjow is a robotics researcher whose work spans reinforcement learning for robotic manipulation, multi-fingered grasping, and autonomous underwater systems. He is best known for pioneering applications of deep reinforcement learning to high-precision industrial assembly tasks, most notably demonstrating how force/torque feedback can be integrated into variable impedance controllers to enable robots to autonomously acquire delicate manipulation skills — work that has garnered over 177 citations. His 2020 paper introducing UniGrasp, a unified model enabling multifingered robotic hands to generalize across novel object geometries, further cemented his reputation as an innovator in dexterous manipulation, attracting over 110 citations. Solowjow has also made significant contributions to residual reinforcement learning, where learned policies complement conventional feedback controllers to handle contact-rich, friction-laden environments. Beyond manipulation, his early work developing the HippoCampus micro underwater vehicle laid important groundwork for swarm robotics research in aquatic environments. More recently, he has explored self-rewarding offline-to-online policy finetuning for novel connector insertion from vision. Across more than a decade of research, Solowjow's contributions consistently bridge the gap between theoretical machine learning and practical industrial robotics deployment.
Research Focus
Key Achievements
Top Papers
- 1
- 2UniGrasp: Learning a Unified Model to Grasp With Multifingered Robotic Hands110 citations · 2020
- 3
- 4Residual Reinforcement Learning for Robot Control45 citations · 2019
- 5HippoCampus: A micro underwater vehicle for swarm applications44 citations · 2015
- 6Domain Randomization for Active Pose Estimation33 citations · 2019
- 7
- 8
- 9
- 10