Papers
54
Total Citations
673
H-Index
13
About
Majid Khadiv is a leading robotics researcher whose work spans legged locomotion, whole-body motion planning, and robot learning, with particular expertise in model predictive control (MPC) for humanoid and bipedal robots. His research addresses some of the field's most demanding challenges: enabling robots to walk robustly, recover from disturbances, and plan complex motions in real time. Among his most influential contributions is BiConMP, a nonlinear MPC framework that generates whole-body trajectories online for legged robots, garnering 88 citations since 2023. His earlier work on combining step location and timing adjustment for robust gait generation (73 citations) and variable horizon MPC for bipedal locomotion (57 citations) established him as a key voice in principled locomotion control. He has also advanced push recovery strategies using capture point theory and tackled practical challenges such as walking on slippery surfaces through gait optimization. Beyond model-based methods, Khadiv bridges the gap between trajectory optimization and reinforcement learning, developing approaches that transfer policies to real robots without additional training. His 2025 survey on learning-based legged locomotion reflects his broad perspective on where the field is headed. His open-source TriFinger platform further demonstrates his commitment to democratizing robotics research for the wider community.
Research Focus
Key Achievements
Top Papers
- 1
- 2Step timing adjustment: A step toward generating robust gaits73 citations · 2016
- 3Variable Horizon MPC With Swing Foot Dynamics for Bipedal Walking Control57 citations · 2021
- 4Learning-based legged locomotion: State of the art and future perspectives34 citations · 2025
- 5
- 6
- 7Robust Walking Based on MPC With Viability Guarantees29 citations · 2022
- 8
- 9TriFinger: An Open-Source Robot for Learning Dexterity24 citations · 2020
- 10