University students bring practical AI solutions to cash-strapped small factories across Korea

Undergraduate researchers from Chungbuk National University built an AI tool for a small manufacturer in Cheongju, North Chungcheong, that could not afford to build one itself.

The company, Chang Myung CST, which makes industrial electronic control gear and medical devices, employs fewer than 50 people. Organizing thousands of technical drawings so that workers could search them by component, dimension and part number had been a problem for years, and the company had neither the budget nor the specialists to take it on.

Chungbuk National University's Industrial Technology Research Park broke the deadlock. Kim Tae-kyung, an associate professor of management information systems at the university, took a team of undergraduate researchers onto the factory floor. The researchers reviewed the workers' struggles and requests and built a tool to organize and analyze the drawing database.

"At first, I thought they were only students and wondered how much they could really manage," Lee Chang-ho, a managing director at Chang Myung CST, said. "Then I watched them build an AI tool that ran on the floor straight away, and it took me by surprise."

Smaller Korean manufacturers now treat AI transformation as a matter of survival rather than a choice. The second baby boom generation, born between 1964 and 1974, held the country's factory floors and is now reaching retirement age in large numbers; the hollowing out of the factory is becoming real. Technology has to fill the gap left by missing hands, but most small companies are caught between a cash squeeze and a shortage of specialists.

Of 163,273 small- and mid-sized manufacturers that own a factory, 31,782 had installed a smart factory, or 19.5 percent, according to a survey by the Ministry of SMEs and Startups and the Korea Smart Manufacturing Office in April 2025. Of those, 75.5 percent are at a basic level, and 0.1 percent of manufacturers have put AI into production itself.

Experts point to regional flagship national universities as the realistic route, with students and faculty going into factories in their own areas to work on specific problems.

"For a regional company that has no dedicated AI staff at all, simply having university students on site is a way through," Kim said. "It could also be extended into an employment-linked program for promising smaller companies that are struggling to hire."

Undergraduates who work on real production problems become hires the company has already watched perform, which is the circle the model is meant to close.

"There is a limit to how far this spreads on the voluntary effort of individual universities and companies," Chang Tai-woo, head of the Intelligence and Manufacturing Research Center at Kyonggi University, said. "Government work force programs have to move toward a hands-on, problem-solving model that links universities and companies before a sustainable industry-academia system can settle in."

Companies can also spend more on AI and end up less productive. If a manufacturer installs a different system at each stage of production without a tight system integration plan, management costs rise, and productivity can fall below that of manual work.

"The government needs to build a single integrated AI system that runs through the entire factory, the way the Android operating system runs through a smartphone," Jang Young-jae, a professor of industrial and systems engineering at KAIST, said. "Turning that into a standard platform that smaller manufacturers can pick up and use easily is the urgent task."

BY KIM SU-MIN, KIM KYUNG-MI [[email protected]]