Papers
62
Total Citations
1,640
H-Index
20
About
Ioannis Havoutis is a leading robotics researcher whose work sits at the intersection of legged locomotion, motion planning, and machine learning for autonomous robotic systems. Based primarily at the Oxford Robotics Institute, he has made foundational contributions to enabling quadruped robots to navigate complex, uneven terrain with agility and robustness. His early work on the hydraulic quadruped HyQ established key frameworks for dynamic trot-walking and onboard perception-driven locomotion, while his influential 2016 paper on high-slope terrain locomotion (204 citations) demonstrated torque-controlled robots tackling previously formidable real-world challenges. Havoutis has consistently advanced the state of motion planning, developing coupled foothold-trajectory optimization methods and nonlinear MPC frameworks such as BiConMP that allow robots to plan whole-body motions online. His 2022 work on RLOC (122 citations) exemplifies his bridging of classical optimal control with modern reinforcement learning. Beyond locomotion, he has contributed to imitation learning on Riemannian manifolds and loco-manipulation for arm-equipped quadrupeds. With over 900 citations across his most recognized papers alone, Havoutis represents a pivotal voice in making legged robots genuinely deployable in demanding real-world environments.
Research Focus
Key Achievements
Top Papers
- 1High-slope terrain locomotion for torque-controlled quadruped robots204 citations · 2016
- 2
- 3
- 4An Approach for Imitation Learning on Riemannian Manifolds109 citations · 2017
- 5
- 6
- 7
- 8
- 9RoLoMa: robust loco-manipulation for quadruped robots with arms57 citations · 2023
- 10