Papers
10
Total Citations
214
H-Index
6
About
Jung-Su Kim is a leading researcher in robotics and artificial intelligence, specializing in motion planning and control for complex robotic systems. His primary contributions lie at the intersection of deep reinforcement learning and robot manipulation, where he has developed novel algorithms for path planning in dynamic environments. Notably, his 2020 paper on using Twin Delayed DDPG with Hindsight Experience Replay for smoother robot manipulator motion has garnered 99 citations, establishing a foundation for automated motion planning in manufacturing. Kim has advanced multi-arm manipulator coordination, introducing methods that handle both static and periodically moving obstacles using Soft Actor-Critic algorithms combined with LSTM-based position prediction. His work on adaptive discount factors for reinforcement learning in uncertain environments (24 citations) and automated hyperparameter tuning for quadrupedal robot locomotion demonstrates his commitment to making RL more practical and robust. Beyond manipulation, Kim has contributed to nonlinear synchronization theory and control for omnidirectional mobile robots. His research continues to push boundaries in perception-based control and traversability prediction for legged robots, addressing real-world challenges in slippery and deformable terrains.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6A Nonlinear Synchronization Scheme for Hindmarsh-Rose Models10 citations · 2010
- 7Learning Robust Perception-Based Controller for Quadruped Robot4 citations · 2023
- 8
- 9
- 10A nonlinear synchronization scheme for polynomial systems2 citations · 2007