Human expertise such as existing know-how plays an absolutely critical role in the industrial AI race that is transforming manufacturing through artificial intelligence (AI), experts said. With its vast talent pool built on a long manufacturing heritage, South Korea can maximize this strength to lead the industrial AI era as well, they added.
"Without human expertise, AI will not be that useful," Ravi Kunju, senior vice president of product strategy and marketing for the SIM division at Siemens Digital Industries Software, said in a Session 3 lecture at "Seoul Forum 2026," held at the Shilla Hotel in Jung-gu, Seoul, on the 28th. "'Engineering intelligence' — the capability of experts to secure reliable data and grasp what that data means — is the core pillar supporting industrial AI."
Kunju previously served as chief product and strategy officer (CPSO) at industrial simulation firm Altair Engineering and joined Siemens after it acquired the company, where he is leading the AI transition in the global manufacturing industry. As he noted, just as AI robots can only learn when there is "behavioral data" of welders welding at an actual shipyard, securing high-quality data from industrial sites requires a variety of on-site experts.
Kunju assessed that South Korea, a manufacturing powerhouse that has long cultivated such talent, can gain an edge in the industrial AI race. "Korea is a country with a great deal of expertise and is well positioned to build industrial foundation models," he said. "Its engineering talent pool, in particular, is enormous."
Lee Tae-jin, chief data officer (CDO) of HD Hyundai, who joined the panel discussion, agreed: "For AI to make optimal decisions, a common language that both AI and humans can understand in virtual space — that is, semantic data — is needed. Manufacturers must make efforts in this area."
The discussion also extended to digital twins, which will be realized through industrial AI. "The competitiveness of digital twins lies in how well they reflect reality," said Cha Jae-won, project manager at Naver Labs. "For example, if a digital twin predicting urban flooding has errors, the service loses its credibility." Kim Hyun-bae, CEO of Deepfine, predicted, "The era of digital twins will open as AI glasses (eyewear-type devices) become widespread, and it will become reality within two to three years."
Kim Jae-hyun, a professor at Sungkyunkwan University who chaired the session, added, "There are discussions ranging from creating digital twins on a small scale, such as factories, to all of Seoul, and Nvidia CEO Jensen Huang has said he will recreate Earth in a virtual world." In response, Hwang Young-ho, director general for technology innovation policy at the Ministry of SMEs and Startups, pledged, "We will support digital twin development not only for large companies but also for small and medium-sized enterprises, and work to ease data-related regulations."
