From learning to physical systems
My work spans humanoid, quadruped, and wheeled robots, connecting reinforcement learning and simulation with the challenges of real-world autonomy.
Current research · Humanoid
Whole-body control and manipulation
A compact humanoid with articulated arms, legs, and hands for studying coordinated movement and interaction with objects. My current work combines reinforcement learning and demonstrations for locomotion and manipulation.
Platform details
Quadruped
Legged navigation and perception
A four-legged robot that can move through cluttered spaces, climb stairs, and traverse uneven ground. Its payload interfaces support additional sensing and onboard computing for mapping and navigation research.
Platform details
Quadruped
Compact, agile legged mobility
A lightweight quadruped with onboard perception and developer interfaces for locomotion and autonomy research. The EDU platform supports experiments with robot control, mapping, and navigation.
Platform details
Quadruped
Mobility in challenging terrain
A rugged quadruped built for outdoor operation and uneven terrain. Its modular payload mounts accommodate sensors and computing hardware for perception and autonomous navigation.
Platform details
Wheeled
Outdoor autonomy and field robotics
A four-wheeled research platform for outdoor robotics. Its larger chassis carries sensing and computing payloads, making it useful for mapping, terrain perception, and navigation experiments.
Platform details
Wheeled
Compact mobile-robot navigation
A compact four-wheeled robot for indoor and outdoor research. Its small footprint and integrated computing make it a practical platform for testing perception, planning, and navigation in constrained spaces.
Platform details
Wheeled
ROS prototyping and indoor autonomy
A small, modular ROS-based mobile robot for research and prototyping. Its accessible hardware and software ecosystem support experiments in mapping, localization, and indoor navigation.
Platform details
Wheeled
Early PhD work · 2021
A 1/18-scale autonomous racing car for hands-on reinforcement learning, with driving policies trained in simulation and deployed on the vehicle. I used DeepRacer in 2021, during the early stages of my PhD.
Platform details