Isac Arnekvist
Papers
2
Total Citations
12
H-Index
2
About
Isac Arnekvist is a researcher advancing the frontiers of robotic manipulation and reinforcement learning. His work centers on developing efficient algorithms for complex physical tasks, particularly in non-prehensile rearrangement planning—where robots must move objects among obstacles without grasping them. His most cited paper (2019, 9 citations) introduces a novel planning algorithm that minimizes assumptions about a robot’s manipulation abilities, enabling more flexible and practical solutions for real-world cluttered environments. Arnekvist also tackles the critical challenge of policy transfer in reinforcement learning. In his work on Variational Policy Embedding (VPE, 2018, 3 citations), he proposes a method to learn robust policies that adapt to varying deployment conditions, addressing the costly retraining problem in robotics. By combining theoretical rigor with practical design, Arnekvist’s contributions help bridge the gap between controlled lab settings and unpredictable real-world applications. His research is particularly valuable for students and engineers seeking to build more adaptable and efficient autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2VPE: Variational Policy Embedding for Transfer Reinforcement Learning3 citations · 2018