A former OpenAI robotics researcher constructed a tabletop manipulation setup that challenges assumptions about research costs. The system uses an industrial-grade UFACT ORY x Arm Lite 6 with dual cameras and teleoperation hardware, all positioned conveniently beside their desk.

Between 2017 and 2020, the author worked on a tabletop setup at OpenAI that cost roughly ten times more and required approximately twenty people to operate effectively. The new configuration totals €4,569.80, representing a dramatic reduction in both financial and personnel barriers for robotics experimentation.

Design choices reflect practical constraints rather than technical limitations. The single-arm configuration sacrifices bimanual tasks like cloth folding but enables creative manipulation strategies using environmental constraints. Rather than building on established frameworks like LeRobot or ROS 2, the researcher developed a custom software stack to maintain flexibility for their specific research goals.

This project demonstrates how accessible hardware and foundation models have democratized robotics research. By documenting the process openly, the author aims to prove that meaningful work in robotic manipulation is now feasible for individuals rather than requiring institutional-scale teams.