I'm a robotics engineer and researcher passionate about robot manipulation, robot system identification, reinforcement learning, and simulation-to-real transfer. I build tools and frameworks that help bridge the gap between robot models and real hardware.
- Robot system identification frameworks for manipulators and humanoids
- Reinforcement learning for locomotion and manipulation in MuJoCo
- End-to-end manipulation pipelines for low-cost robot arms
- Long-horizon task and motion planning (TAMP) for manipulation
A system identification framework for robots — manipulators, mobile manipulators, and humanoids.
Comprehensive examples and tutorials for the FIGAROH-PLUS library.
The core algorithms of the Humanoid Path Planner framework.
Monorepo workspace for the SO-ARM100 project — integrating SDK, cloud RL training (mjlab), teleoperation, camera calibration, and LeRobot support.
RL training of the Walka biped humanoid in MuJoCo via mjlab.
Simple quadcopter simulator in Python.