LG Electronics is accelerating its efforts to secure a leading position in the robotics sector by collaborating with NVIDIA to implement its future core strategy of Physical AI. This initiative comes just four days after Koo Kwang-mo, chairman of LG Group, formed an AI alliance with NVIDIA CEO Jensen Huang at the company's headquarters in the United States.
On August 18, LG Electronics invited key figures, including Madison Huang, NVIDIA's Senior Vice President, to its Yangjae Data Factory in Seoul, a facility under construction with a total area of 10,000 square meters (approximately 2,625 acres), to discuss robotics collaboration and synergy strategies. Madison Huang, the daughter of Jensen Huang, is a key executive leading NVIDIA's future business and is on a succession path. The event was attended by Ryu Jae-cheol, president of LG Electronics, along with other top executives from LG Group's major affiliates, including Hyun Shin-kyun of LG CNS and Jeong Soo-hun of LG Science Park.
Previously, on August 13, Koo met with Jensen Huang at NVIDIA's headquarters in California to sign a strategic memorandum of understanding (MOU) covering three key future areas: robotics, AI factories, and mobility. Koo stated, "We will lead structural changes in future businesses through AI," emphasizing the need to quickly secure competitiveness in Physical AI, which involves robots with actual physical bodies beyond software.
The executives from both companies toured the Yangjae Data Factory for about two and a half hours, which is set to be fully operational by the end of the year. The facility serves as a base for robotic physical learning, where robots perform real tasks in environments that simulate home settings or manufacturing logistics, such as a washing machine factory in Tennessee, accumulating and validating high-quality data.
They also reviewed the operational status of the data collection and learning system using LG's proprietary robot, LG Cloi. LG Cloi performs various tasks, including cleaning, transporting parts, loading, and assembly, generating and collecting high-quality learning data in real-time to enhance robot capabilities. The collected data will be integrated with NVIDIA's advanced AI solutions to significantly improve the robots' intelligence.
LG Electronics plans to leverage decades of expertise from skilled workers in global appliance manufacturing and unique data, while NVIDIA will provide AI solutions to complete the robotic evolution system that connects data collection, learning, and performance improvement. Key NVIDIA technologies, including the Omniverse library, Cosmos open-world model, and Isaac open robotics development platform, will be fully implemented.
By the end of this year, LG Electronics aims to combine data collected directly from the Yangjae Data Factory with virtual data to create a database totaling 100,000 hours, equivalent to 12 years of data. The foundational model for robots, which influences their intelligence and operational performance, will also be further improved.
Under Koo's strong push for new business initiatives, the 'One LG' strategy, which consolidates the capabilities of LG Group affiliates, is gaining momentum. LG Group plans to combine NVIDIA's powerful AI chips and robotics platform with LG Electronics' hardware manufacturing technology, LG Innotek's sensors, and LG Energy Solution's battery capabilities to unveil next-generation humanoids in the first half of next year.
In June, LG Electronics declared the year as the starting point for its robotics business and established a 'Robotics Business Center' directly under the CEO. This move aims to consolidate all robotic capabilities within the organization, including business development, sales, and operations.
Ryu Jae-cheol, president of LG Electronics, emphasized, "We will secure Physical AI competitiveness early through synergies based on 'One LG' and strategic collaboration with global partners, aiming to become a leading total solutions provider in robotics on the global stage."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.