Hello!
GitHub · LinkedIn · Twitter · E-mail
I’m Louis Le Lay, a robotics engineer student turning advanced simulation and AI into real-world capabilities. I work on reinforcement learning, sim-to-real transfer, and multi-sensor fusion, building robots that can move, see, and adapt in both simulation and reality.
I’ve contributed to open-source projects like Isaac Lab, MuJoCo Menagerie, and RealSense-ROS, and built sim-to-real tools from Kinova Gen3 manipulation pipelines to Disney BD-X locomotion tasks.
My experience spans PAL Robotics, the University of Tokyo, Naova, and AIRBUS, with skills in Python, C++, ROS/ROS2, and simulation platforms like Isaac Lab and MuJoCo. I’ve worked with hardware from Kinova arms and Unitree quadrupeds to custom bipeds and full humanoids.
My current work at PAL Robotics focuses on developing a RL-based locomotion for PAL's humanoid robot Kangaroo, using Isaac Lab.
Outside of work, I enjoy mentoring students on Isaac Lab and building my own biped robots.
Simulation & Reinforcement Learning
My work centers on physics simulation: designing, training, and validating control policies before taking them to hardware. I craft custom RL environments, refine rewards for stable locomotion, and bridge simulation and robots for dependable sim-to-real transfer.
Open access
I keep most of my work public whenever possible.