Hardware

Boston Dynamics Trains Atlas Inside Hyundai Factory

Boston Dynamics has opened a dedicated training facility inside Hyundai's Georgia EV plant to prepare its electric Atlas humanoid robots for a massive global deployment of 25,000 units.

AlphaSignal1 day agoHardware
Image: AlphaSignal

Boston Dynamics has launched the Robotics Metaplant Application Center (RMAC) inside the Hyundai Motor Group Metaplant America in Savannah, Georgia. This facility serves as a real-world training ground for the all-electric Atlas humanoid robot. Hyundai has committed to deploying more than 25,000 Atlas robots across Hyundai and Kia factories worldwide, supported by a planned U.S. production capacity of 30,000 robots annually. The initial phase of RMAC is operational now, preparing the humanoids for a scheduled 2028 deployment at the plant.

At the Georgia facility, Atlas will initially perform parts sequencing for the IONIQ 5 and IONIQ 9 electric vehicles, retrieving components from bins and arranging them for the assembly line. By 2030, the robots will transition to component assembly and heavier repetitive tasks. The electric Atlas features a four-hour battery life, supports autonomous battery swaps in under three minutes, and uses an onboard safety system to pause when humans enter its safety radius.

The development pipeline is highly integrated: Boston Dynamics designs the Atlas platform and software, Hyundai Mobis manufactures the joint actuators, and Hyundai provides the factory infrastructure. All of Boston Dynamics' 2026 Atlas production capacity is already allocated to RMAC and Google DeepMind, with new commercial customers slated for 2027. Next year, RMAC plans to expand into a new building roughly 10 times its current size to accommodate non-automotive clients.

For robotics practitioners, this initiative represents a shift from isolated lab testing to a continuous "data factory" loop. By integrating Atlas with Hyundai's "Software-Defined Factory" systems, engineers can capture physical run data—including sensor inputs and error recoveries—to retrain control policies in simulation before deploying updates. This structured pipeline addresses the scaling challenges faced by competitors like Tesla's Optimus and Figure's deployments at BMW, though Boston Dynamics has yet to release specific metrics on cycle times, unit costs, or error rates.

This is our own summary of reporting by AlphaSignal

More in Hardware