[Edaily Reporter JAEMIN SONG ] LGELECTRONICS and NVIDIA are establishing a “virtuous cycle of data” for robot training by combining LG Corp.’s manufacturing and logistics data with NVIDIA’s physical artificial intelligence (AI) technology. LGELECTRONICS plans to deploy hundreds of robots at its Yangjae Data Factory in Seoul by the end of the year to secure 100,000 hours of training data—equivalent to approximately 12 years’ worth of data.

LGELECTRONICS announced on the 18th that it met with NVIDIA officials at the LG Data Factory in Seocho-gu, Seoul, to review the direction of their robotics collaboration and discuss commercialization strategies. Attendees at the meeting included Madison Huang, Senior Director of Product Marketing for Omniverse and Robotics at NVIDIA; Jeong So-young, CEO of NVIDIA Korea; Ryu Jae-cheol, CEO of LGELECTRONICS; Hyun Shin-kyun, CEO of LG CNS; and Jeong Su-heon, CEO of LG Science Park.

This meeting took place just four days after LG Corp. and NVIDIA signed a Memorandum of Understanding (MOU) on the 13th (local time) at NVIDIA’s U.S. headquarters to collaborate on future strategic business initiatives. It marked the translation of the agreement reached at the top management level into concrete collaboration plans at the actual robot development site.

Senior Director Huang quietly entered the campus around 8:40 a.m. that day without any public schedule, and CEO Jeong arrived about 10 minutes later. The meeting lasted approximately 2 hours and 30 minutes, until 11:12 a.m., exceeding the initially expected 2 hours.

The two companies reviewed the operational status of the Data Factory, which is currently under construction with the goal of achieving full-scale operation by the end of the year. The Data Factory spans four floors—from the first basement level to the third floor above ground—with a total floor area of 10,000 square meters. It consists of spaces where robots learn various tasks and where collected data is verified and refined.

Currently, at the Data Factory, “LG CLOiD,” a humanoid robot developed in-house by LGELECTRONICS, is performing data generation, collection, and learning tasks. CLOiD cleans in a space designed to resemble a home and moves, loads, and assembles parts in a manufacturing environment replicating LGELECTRONICS’s washing machine factory in Tennessee, U.S. Relevant data is also being collected at LG CNS’s logistics automation solution facility and LG Innotek’s robotic hand training space.

LGELECTRONICS is combining data accumulated over decades in manufacturing and logistics settings, along with the know-how of skilled workers, with NVIDIA’s robotics technology. Using the NVIDIA Omniverse library, the Cosmos OpenWorld model, and the Isaac Open Robotics development platform, the company augments and synthesizes real-world data to expand it into a large-scale training dataset.

The total volume of data—combining data LGELECTRONICS will collect directly and virtual data generated through augmentation and synthesis—is expected to reach 100,000 hours by the end of the year, equivalent to approximately 12 years’ worth of data. The company plans to establish a virtuous cycle of high-quality data training and iterative robot performance improvements to further refine the Robot Foundation Model (RFM), which governs a robot’s perception, decision-making, and actions.

During his tour of the Data Factory, Senior Vice President Hwang personally left the message “AMAZING LG Corp. (LG Corp.)” on LG Cloid.

After the meeting, she stepped outside the building with LG Corp. CEO Ryu and other executives to exchange greetings before getting into a black Palisade. As the car passed in front of the press, she rolled down the window and, in response to a question asking, “How was the meeting?” replied with a beaming smile, “Incredible.” He then gave a thumbs-up and waved, while LG Corp. executives and employees, who remained outside the building, waved back at the car until it was out of sight.

CEO Ryu stated, “Through synergies based on ‘One LG Corp.,’ which brings together the group’s core competencies, and strategic collaboration with global partners, we will secure competitiveness in physical AI and transform ourselves into a total solutions provider for robotics.”