LG is building a next-generation bipedal LG humanoid robot on NVIDIA’s stack. The company’s August 13 announcement targets a public reveal in the first quarter of 2027.
LG Corp chairman Kwang Mo Koo and NVIDIA CEO Jensen Huang signed the memorandum of understanding at NVIDIA’s Santa Clara headquarters. The deal covers robotics, AI factories, and an AI-defined vehicle platform. It is an execution pact, not a new product listing.
What the LG humanoid robot will use
LG says the biped will run NVIDIA Jetson Thor for onboard compute. It will also use Isaac GR00T, NVIDIA’s open humanoid foundation model, plus NVIDIA Halos for robotics safety. LG Electronics, LG Innotek, and LG Energy Solution will supply actuators, sensors, and batteries.
The company is not waiting until 2027 to put a robot on a factory floor. This year, LG will send its wheel-based CLOiD robot to the LG Electronics washing-machine line in Tennessee. After that trial, LG says it will refine the design and expand to other plants, homes, and commercial sites.

Meanwhile, LG CNS PhysicalWorks will power a robot data factory. That system would collect site data, make synthetic data, train models, and verify robots. LG also says it will keep building its own robot foundation model beside Isaac GR00T. This sits next to other humanoid work TMM has tracked, including Gemini Robotics 2 whole-body control.
AI factories and cars sit in the same deal
First, LG plans an AI factory reference site on NVIDIA Vera Rubin in the first half of 2027. Then it wants an 80-megawatt LG AI Factory in Cheonan, South Korea, by the first half of 2028. The company says it wants to sell a One LG turnkey package to other big tech buyers.
On the road, LG plans an HPC platform for AI-defined vehicles on NVIDIA DRIVE Hyperion. It wants to fold its in-vehicle infotainment work into autonomous driving software for carmakers. Other EV brands have mixed robot talk with vehicle bets, as Faraday Future’s robot-and-car lineup already showed.
Huang framed the work as machines that can understand a room, reason, and act next to people. Koo said the AI factory and physical AI tasks are now clear, and LG will chase industry-leading reference sites. A joint technical and business task force will run R&D, site tests, and later sales.
Still, this is a timeline, not a shipping humanoid. LG has not published a name, height, payload, or price. The Tennessee CLOiD trial is the near-term test. The Q1 2027 biped is the later showpiece.


































