Luminous Robotics, a MassRobotics resident in Boston, is deploying 4,000-pound industrial robotic systems at active solar farm construction sites to address the manual, labor-intensive installation of solar panels, which requires centimeter-level accuracy. Operating under a robots-as-a-service model, the LUMI robots use multi-camera perception systems to lift and place panels onto steel “torque tubes,” but placement previously relied on human operator confirmations. The Physical AI Fellowship program, an 8-week collaboration powered by Amazon Web Services, NVIDIA, and MassRobotics, worked with Luminous to develop a multi-purpose AI pipeline for autonomous placement. Previously, each LUMI robot required a dedicated operator who watched camera feeds and made real-time corrective adjustments, creating a 1:1 robot-to-operator ratio that limits throughput; under this paradigm, 3 robots require 3 operators to install thousands of panels within an 8-hour shift. Luminous aims to target 2 or more robots managed by a single shared operator who intervenes only when the AI requires human input. The solution uses operational data from tens of thousands of recorded corrective episodes, where operator interactions with the controller are synchronized with feeds from the robot’s six onboard RGBD cameras: two front-facing, two side-facing, and two arm-mounted, captured on active construction deployment sites.