LG and Nvidia accelerate 'physical AI' collaboration, aiming for 12 years of data collection within the year

LG Electronics and Nvidia are intensifying their collaboration in the field of "physical AI," with a particular focus on advancing robotics technology. This partnership follows a recent Memorandum of Understanding signed between LG Group Chairman Koo Kwang-mo and Nvidia CEO Jensen Huang.

The collaboration centers on enhancing robot learning capabilities. LG's 'Data Factory,' a facility designed to train robots in simulated environments mimicking homes and factories, will be integrated with Nvidia's robotics technology. This synergy aims to accelerate the acquisition and expansion of data crucial for robot learning. LG is currently using its own robot, 'Cloid,' to collect and learn from operational data within the Data Factory.

Nvidia's humanoid inference models will be utilized to process and augment this collected data, transforming it into high-quality training material. This refined data will then be fed back into the learning process, creating a continuous improvement loop. LG Electronics announced its goal to collect a total of 100,000 hours of robot training data by the end of this year, a volume equivalent to 12 years of learning.

The Data Factory itself is a significant investment, spanning four floors with a total area of 10,000 square meters. LG plans to deploy hundreds of robots within this facility by year-end. Ryu Jae-cheol, CEO of LG Electronics, expressed confidence in the "One LG" approach, leveraging group-wide capabilities and strategic partnerships to secure leadership in physical AI and become a total robotics solution provider.