Papers
12
Total Citations
139
H-Index
7
About
Daegyu Lim is a robotics researcher whose work centers on three critical pillars of human-robot interaction: collision detection for collaborative robots, robust humanoid locomotion, and intuitive teleoperation systems. His most impactful contribution is the development of deep learning approaches to overcome model uncertainty in collision detection, notably through a transferable modularized neural network (40 citations) and a momentum observer enhanced with LSTM networks (33 citations). These methods enable robots to detect unexpected collisions quickly and safely, addressing a fundamental safety challenge as robots move into human environments. In humanoid robotics, Lim has advanced balancing and walking control through model predictive capture point frameworks, foot stepping algorithms with double support time adjustment, and compliant motion control—work that has garnered over 50 combined citations. He contributed to the ANA Avatar XPRIZE finals, developing an intuitive robotic avatar system for tele-existence that allows operators to control humanoids naturally for both manipulation and imitation tasks. Lim’s research, spanning from sensorless torque estimation to online walking pattern generation, consistently pushes toward practical, safe, and intuitive deployment of humanoid and collaborative robots in real-world settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8Advances in Humanoid Robot Walking Technologies: A Review6 citations · 2024
- 9
- 10