Papers
6
Total Citations
65
H-Index
4
About
Gijeong Kim is a robotics researcher whose work is reshaping how legged robots move, think, and adapt. His primary research areas span quadruped robot design, contact-implicit model predictive control (MPC), and reinforcement learning for diverse locomotion. Kim’s most impactful contribution is the development of the KAIST HOUND quadruped platform, which uses a novel mixed-integer nonlinear optimization of gear trains to achieve fast, efficient locomotion at target speeds of 3 m/s—a design that has garnered 30 citations. He is also a pioneer in contact-implicit MPC, a framework that enables robots to discover multi-contact motions in real time without pre-planned contact modes or foothold positions, as detailed in his highly cited 2024 paper (22 citations). This approach, merging differential dynamic programming with contact dynamics, allows for unprecedented agility and adaptability. Kim further addresses real-world challenges with his online friction coefficient identification method for slippery terrain (2025, 5 citations) and a barrier-based style reward learning framework for diverse gaits. His work, recognized with over 65 total citations, is advancing the frontier of autonomous, versatile legged robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6