Egor Davydenko
Papers
2
Total Citations
3
H-Index
1
About
Egor Davydenko is a roboticist specializing in humanoid locomotion, with a particular focus on reinforcement learning (RL) for bipedal walking. His research addresses the critical challenge of enabling humanoid robots to achieve stable, adaptive gaits on complex and uneven terrain. Davydenko’s major contributions include a systematic investigation into the role of reference trajectories in RL-based gait optimization. His 2023 paper, “Benchmarking the Full-Order Model Optimization Based Imitation in the Humanoid Robot Reinforcement Learning Walk,” explores the trade-offs between using and omitting reference trajectories, offering valuable guidance for control developers. Building on this, his 2025 work, “Reinforcement Learning-Based Footstep Control for Humanoid Robots on Complex Terrain,” proposes a novel RL framework for footstep planning that allows a robot to execute commands with varying positions, orientations, and heights, enabling stable locomotion over challenging surfaces. While his citation counts are still growing, Davydenko’s work is at the forefront of practical, learning-based control for humanoid robots, directly addressing the real-world need for robots that can navigate stairs, rubble, and other obstacles. His research is a key resource for students and engineers aiming to develop more capable and resilient walking robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2