Ultimately, 'Learning' Determines Robot Value... The 'Brain Tutor' Set to Disrupt the $81.5 Billion Market

Hyun Shinkyun, President of LG CNS, Interviewed at CES 2026
Full-Scale Deployment Expected in Two Years as Robot Manufacturers Establish Mass Production Systems
Accelerating Expansion into Defense and Military AI Markets and Broadening Global Business Reach
Focusing on Solid Growth to Boost Share Value While Maintaining a Policy of Increasing Dividends

As the global humanoid robot market is poised for explosive growth, the 'Robot Brain Tutor' market-which focuses on training general-purpose robots to become skilled workers in industrial settings-is emerging as a key battleground for the future artificial intelligence (AI) industry. Analysts suggest that once hardware performance reaches a standardized level, the decisive value-add for corporate productivity will be determined by how effectively robots have learned on-site tasks.

On January 7 (local time), Hyun Shinkyun, President of LG CNS, held a press conference in Las Vegas, United States, during 'CES 2026,' the world’s largest electronics and IT exhibition. He stated, "The role of experts who take robots with general-purpose brains and train them for specific field applications is becoming increasingly important. Just as a new university graduate is hired and mentored through on-the-job training (OJT) to become a skilled worker, LG CNS aims to serve as a 'Robot Brain Tutor,' injecting industrial intelligence into robots."

Hyun Shinkyun, President of LG CNS. LG CNS

The Robot Brain Tutor goes beyond simply programming robot movements; it refers to a sophisticated process of infusing general-purpose robot intelligence models with the language, rules, and know-how specific to industrial sites. This involves fine-tuning the Robot Foundation Model (RFM)-which combines 'visual intelligence' for recognizing objects in factories, 'language intelligence' for understanding work instructions, and 'action intelligence' for assembling items with precision-using on-site data. In particular, LG CNS has developed the capability to overcome variables in the physical world by using digital twin-based simulation technology to train robots tens of thousands of times in a virtual environment before actual deployment.

In fact, global market research firms are highly optimistic about the potential of the robot brain training and tuning market. Goldman Sachs, in its latest report, forecast that the global humanoid robot market will surge to approximately 38 billion dollars (about 51.5 trillion won) by 2035. The report noted that as robot manufacturing costs have plummeted by about 40%-from 250,000 dollars per unit in 2022 to around 150,000 dollars recently-the market’s focus is shifting from hardware to 'software and learning algorithms.' Research Nester projected an even more aggressive outlook, estimating the robot market will reach 81.5 billion dollars (about 110 trillion won) by 2035, with the education and training segment expected to post an overwhelming compound annual growth rate of 44.7%.

Amid these market trends, LG is mobilizing its group-wide capabilities to build a 'robot intelligence' ecosystem. At this year’s CES, LG CNS unveiled its 'Robot Brain Tutor' strategy to gain a foothold in the domestic and global physical AI markets. Notably, LG CNS formed a strategic partnership and equity investment with Skild AI, a U.S. company that has attracted large-scale investments from Nvidia and SoftBank and is now valued at nearly 14 billion dollars (about 20 trillion won). Based on Skild AI’s general-purpose RFM, LG CNS is optimizing robots by integrating the 'industrial intelligence' it has accumulated over years in manufacturing and logistics. Currently, LG CNS is conducting proof-of-concept (PoC) projects for high value-added tasks such as ship component assembly inspections at over 10 client sites, using devices from Unitree and other top three global robotics companies.

The Robot Brain Tutor market is evolving beyond technology provision into a 'Learning as a Service' (LaaS) subscription model. Instead of incurring massive upfront costs, companies subscribe to robot 'intelligence' and continuously enhance it by leveraging data generated on-site for further training. Nvidia’s 'Project GR00T' and Google’s 'RT-2' model provide such infrastructure, while Hyundai Motor Group and Samsung Group are also joining the race for physical AI leadership, leveraging their manufacturing capabilities and data factories. Industry insiders believe that the next two years, as robot hardware becomes mainstream, will mark a true inflection point for the market.