You will build the software systems that enable robots to learn dexterous manipulation skills from human data. You will work with dexterous hands and grippers across control, perception, teleoperation, simulation, and learned policy deployment, turning research ideas into working systems on physical robots. We are looking for robotics software engineers who like working on real hardware and want to apply learning-based approaches to it. You don't need prior experience with robot foundation models, but you should be excited about applying modern machine learning to robotics.
What you'll do
- Develop and maintain software for robot arms, dexterous hands, and grippers, including control, sensor integration, and real-time execution.
- Build teleoperation, motion-retargeting, and data-collection systems that map human hand motion, force, and muscle-activity data onto dexterous hands and grippers with different kinematics.
- Build perception for our robots and capture setups, including multi-camera calibration and synchronization, hand and object tracking, and 3D state estimation for teleoperation and policy deployment.
- Integrate cameras, tactile sensors, force/torque sensors, and other sensing modalities into robotic systems.
- Deploy and evaluate learned manipulation policies on physical robots, diagnosing failures across perception, planning, control, and hardware.
- Develop simulation environments and evaluation tools to support robot learning experiments.
- Work with machine learning and biomechanics researchers on imitation learning, reinforcement learning, and other approaches to dexterous manipulation.
- Build the robot runtime, data, and evaluation tooling the rest of the team relies on to run experiments quickly.
What we're looking for
- Strong software engineering skills in Python and/or C++.
- Experience developing software for physical robots, simulation, or closely related systems.
- Solid understanding of robotics fundamentals such as kinematics, dynamics, motion planning, control, or state estimation.
- Ability to independently debug complex systems spanning hardware and software.
- Experience with dexterous hands, contact-rich manipulation, or force and impedance control is a plus.
- Experience with computer vision, 3D perception, camera calibration, or pose estimation is a plus.
- Experience with ROS/ROS2, robot simulators, or robotics development frameworks is a plus.
- Experience with robot learning, imitation learning, reinforcement learning, or PyTorch is a plus.