Papers
10
Total Citations
285
H-Index
7
About
Byeongho Yu is a leading robotics researcher whose work spans quadrupedal locomotion, aerial manipulation, and autonomous navigation in unstructured environments. His most influential contribution is **DreamWaQ** (122 citations), a deep reinforcement learning framework that enables quadrupedal robots to robustly traverse complex terrains by implicitly imagining terrain properties—eliminating the need for explicit mapping. Yu also advanced state estimation for legged robots with **STEP** (41 citations), introducing a preintegrated foot velocity factor that relaxes the non-slip assumption, and its improved version **BIG-STEP** (3 citations), which incorporates fast ground segmentation. In aerial robotics, he developed **CAROS-Q** (29 citations), a climbing aerial robot system with a quasi-decoupling controller for vertical surface inspection, and **Retro-RL** (13 citations), which combines nominal controllers with deep reinforcement learning for tilting-rotor drones. His survey on robotics for civil infrastructure inspection (53 citations) highlights the growing demand for autonomous inspection systems. Yu’s recent work includes **TRG-Planner** (10 citations), a traversal risk graph-based path planner for safe navigation in challenging environments. With over 280 total citations, Yu’s research is defining new capabilities for robots operating in the real world.
Research Focus
Key Achievements
Top Papers
- 1
- 2Survey of robotics technologies for civil infrastructure inspection53 citations · 2022
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10