Moving Beyond Labor-Intensive Manufacturing with AI Robots
- 2025.09.15

Professor Youngjae Jang of KAIST emphasized that the recent detention of Korean engineers in the United States underscores the urgent need to reduce dependence on human labor in overseas factory operations. He pointed out that today's increasingly complex manufacturing environments are not sustainable in a global context and presented Physical AI and the Software-Defined Factory (SDF) as the solution. Physical AI transfers intelligence trained in a virtual environment to real manufacturing facilities, and has already demonstrated results in semiconductor production by reducing logistics time by 32% and construction time by 75%.
Professor Jang also announced plans to make an automated design system available free of charge by next year, enabling work that would normally require a month of expert analysis to be completed in just three hours. He further emphasized that the ultimate goal is to transition to a "dark factory," where the entire factory operates autonomously like a single operating system. In addition, he stressed that, rather than continuing policies focused primarily on data collection, Korea should build an execution-oriented Physical AI ecosystem to strengthen its global leadership in manufacturing.