Staff Research Scientist, Dexterous Manipulation
Develop reinforcement and imitation learning algorithms for dexterous robot hands, then test them in simulation and on physical robots.
Work on robots, autonomy, and the hardware around them. Roles link directly to the employer.
Develop reinforcement and imitation learning algorithms for dexterous robot hands, then test them in simulation and on physical robots.
Build the systems that train and run learned robot policies, working with researchers on perception, manipulation, planning and reasoning.
Lead experiments in robot learning, simulation transfer and foundation models, and validate which approaches can become production capabilities.
Develop teleoperation and demonstration collection software, including motion capture, robot targeting and visual and haptic feedback.
Research reinforcement and imitation learning for robot manipulation. The posting allows flexibility in internship start date and duration.
Design robotic hands, actuators and precision mechanisms, taking complex electromechanical systems from prototype toward production.
Research learned manipulation for manufacturing and build the simulation, teleoperation and training infrastructure needed to deploy it.
Improve a browser-based robotics simulation engine and digital twin workflows for machine design. Hybrid role.
Develop real-time motion components, embedded Linux services and C++/Python APIs for the MachineMotion automation platform. Hybrid role.
Apply vision models, pose estimation and synthetic data to robotic picking. In-person, eight-month term from January to August 2027.
Build real-time machine vision features and test robot integration in the lab. In-person, eight-month term from January to August 2027.
Build cloud services and robot telemetry systems for vision-guided robotics. In-person, eight-month term from January to August 2027.
Develop the Forge simulation platform with React and Python. In-person, eight-month term from January to August 2027.
A curated snapshot of employer postings. Applications go directly to the company. We check the role and location against the linked listing.