■ Interview Series ① – Advancing Manufacturing Innovation with Physical AI: “Korea Should Target Global Markets with Factory Packages”

Professor Youngjae Jang, Director of the Physical AI Research Center for Manufacturing at KAIST, emphasized the need to drive manufacturing innovation and develop a factory export model based on Physical AI. He stated that approaching robotics, automation, logistics, and manufacturing separately is no longer sufficient to compete globally, and that Korea should instead export integrated dark factory systems. Through the Jeonbuk demonstration project, the application of Physical AI to manufacturing processes including assembly, inspection, and labeling has led to improvements in productivity, quality, and process efficiency. The project, with a budget of KRW 22.9 billion, has established demonstration labs in Jeonbuk and at KAIST to validate its real-world applicability. Building on these results, the government is considering launching a full-scale program worth approximately KRW 1 trillion by 2030 while expanding private-sector participation. Professor Jang described Physical AI as an opportunity to integrate manufacturing, robotics, and IT capabilities, and emphasized the need for a turnkey export strategy that combines factory design and construction.

https://www.edaily.co.kr/News/Read?newsId=05182406645322312&mediaCodeNo=257&OutLnkChk=Y 

■ Interview Series ② – Physical AI Is Different from ChatGPT: A Once-in-a-Generation Opportunity for Korea's Manufacturing Industry

Professor Youngjae Jang, Director of the Physical AI Research Center for Manufacturing at KAIST, explained that manufacturing environments require a fundamentally different approach from generative AI due to the many variables and unique characteristics of production processes. He noted that even when the same process is replicated, outcomes can vary depending on factors such as temperature, humidity, and operator behavior, making manufacturing an environment without a single correct answer. As a result, he identified reinforcement learning and digital twins as the core technologies of Physical AI for manufacturing, explaining that they make it possible to identify optimal solutions through thousands or even tens of thousands of simulations in a virtual environment before applying them to real production processes. Because repeated trial and error is impractical in actual factories, he emphasized the importance of validating solutions in a digital environment without operational risk. The project currently underway aims to develop a Physical AI-based factory operating system (OS) that will evolve into a platform capable of orchestrating robots and manufacturing equipment within a unified system. Professor Jang stated that these technologies will provide the foundation for dark factories and future factory export models.

https://www.edaily.co.kr/News/Read?newsId=05188966645322312&mediaCodeNo=257&OutLnkChk=Y 

■ Interview Series ③ – Without Physical AI, Factories Struggling to Hire Workers May Not Survive

Professor Youngjae Jang, Director of the Physical AI Research Center for Manufacturing at KAIST, emphasized that Physical AI-based automation is essential to address labor shortages in regional manufacturing. He explained that many small and medium-sized factories outside major metropolitan areas continue to struggle with production because they cannot hire enough workers, making automation not a matter of choice but of survival. He added that adopting robots is not about replacing jobs, but about filling workforce gaps that already exist. Physical AI is applied by orchestrating diverse robots and manufacturing equipment within a unified system to improve the operational efficiency and adaptability of entire factories. Professor Jang also highlighted practical examples, including the simultaneous orchestration of more than 1,000 robots using reinforcement learning to eliminate congestion, and reducing factory design time from one month to half a day. He emphasized that Physical AI is key to lowering the barriers to manufacturing automation and building a sustainable production system.

https://www.edaily.co.kr/News/Read?newsId=05195526645322312&mediaCodeNo=257&OutLnkChk=Y