As South Korea seeks to combine its manufacturing capabilities with physical artificial intelligence (AI) robotics technology, experts gathered for a lively discussion on how the country can surpass the United States and China to become a leader in physical AI.
The '18th Good Growth, Good Jobs Global Forum (2026 GGGF)', hosted by Aju Economy, continued successfully on September 3 at The Plaza Hotel in Jung-gu, Seoul, attracting over 300 attendees who listened to insights from industry experts.
Among the participants were Ju Si-hyun, head of robotics intelligence development at Hyundai Motor and Kia, Lim Kyung-seop, director of the Ministry of Trade, Industry and Energy's AI Manufacturing Transition Cooperation Division, and Choi Jae-sik, CEO of InEasy and a distinguished professor at KAIST. They presented various agendas for the advancement of South Korea's robotics industry and engaged in passionate discussions.
Experts agreed that physical AI will play a crucial role in addressing the challenges of an aging population and labor shortages in South Korea. Ju emphasized, "For physical AI to grow beyond a single robotic product into an industry, services and infrastructure surrounding the robots must also evolve. Partnerships with various industry players, including parts suppliers, transportation, services, insurance, and telecommunications, are essential."
He added, "Focusing solely on the humanoid's appearance and demonstration videos will not accelerate industrialization. We need to solve industrial challenges such as automation in manufacturing, aging, labor shortages, and the efficiency of logistics and services through robotics, connecting these solutions to new markets."
The government plans to invest 115.8 billion won next year to establish a 'Manufacturing Data Library' aimed at securing quality data for the development of the physical AI industry. To address past limitations where companies did not submit critical data on defects and accidents under the premise of public disclosure, the approach will shift to a principle of confidentiality.
Data will only be learned by AI models within a physically separated 'clean room,' and original data will not be allowed to leave the facility. Lim explained, "We will collect data on a non-disclosure basis, allowing learning within the clean room, but the data cannot be taken out. We are designing it so that only the AI model can be trained."
There was also interest in benchmarking China's physical AI competitiveness, which surpasses South Korea's manufacturing capabilities. Kim Soo-yeon, head of the AI Center at EY Han Young, noted the rapid expansion of China's robotics ecosystem, stating, "Currently, 80-85% of the AI agent market in China consists of in-house projects, with budgets ranging from 300,000 to 3 million yuan accounting for 78% of the total market growth."
Kim urged the South Korean industry to focus on how to create learning environments rather than just how to purchase and use robots, emphasizing the need for a 'structural strategy' that systematizes independent on-site data and work processes to design self-learning environments.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.