Ford partnered with Humanoid on a proof-of-concept project completed within six weeks, demonstrating how humanoid industrial robots can undertake real industrial tasks in automotive manufacturing with speed, precision, and reliability. The project focused on two complex workflows: bin picking for assembly and two-arm manipulation of large metal body components.
During the proof of concept, Humanoid's wheeled Alpha HMND 01 industrial robot autonomously transported bins with an eight-kilogram payload for one hour, exceeding the target. In the fully autonomous pick-and-place task, it achieved 97% reliability and a pick rate of 83 units per hour, 60% higher than expected. Utilizing AI technology trained on large, diverse datasets, Humanoid delivered a high-performance autonomous model with just one hour of industrial robot data collection.
In the second workflow, the Alpha industrial robot demonstrated its ability to precisely handle thin sheet metal body components, successfully navigating to a rack, locating, and safely placing the parts. Despite the complexity of the process, the robot performed safely and consistently, validating tasks traditionally challenging for autonomous systems.
The proof of concept was conducted in two phases: initial development in Humanoid's internal lab, followed by an on-site demonstration at Ford's Cologne plant. Thanks to close collaboration between both teams, the on-site phase progressed smoothly. The project also leveraged the latest NVIDIA technologies, including an operational digital twin developed on Omniverse.
Based on these results, Humanoid and Ford Cologne are exploring opportunities to introduce humanoid industrial robots into actual production environments and plan further demonstrations at the Innovation Center. Artem Sokolov, founder of Humanoid, stated: "Innovation only matters when it works on the factory floor. Our joint proof-of-concept with Ford in Cologne proves that humanoid industrial robots are ready to take on real industrial tasks—not in a few years, but today. The team moved from discussion to on-site demonstration in six weeks, and the results exceeded all benchmarks, showing that rapid progress is possible when both sides align on scope and adhere to safety commitments."
