
A researcher demonstrates robotic technology at Jeonbuk National University's Physical AI Test Lab, May 19. Korea Times photo by Choi Won-suk
JEONJU, North Jeolla Province — Inside a manufacturing test lab at Jeonbuk National University, robots move with synchronized precision across a factory floor designed to mimic a real industrial site.
Autonomous mobile robots (AMRs) shuttle components between workstations, weaving through production lines without human guidance. Nearby, robotic arms pick up, assemble and inspect parts in a continuous sequence, while cameras and sensors track every movement in real time. The machines communicate with one another, make decisions based on real-time data and adjust on the fly.
It is not a commercial factory — not yet. It is an 846-square-meter manufacturing technology test lab designed to replicate real industrial conditions and determine whether physical artificial intelligence (AI) technologies are ready for deployment on actual production floors.
The facility sits at the center of Jeonbuk National University's push into physical AI, a field gaining momentum nationally as President Lee Jae Myung works to position Korea as a global leader in the field.
Through its test lab and industry-linked research infrastructure, the university is working to bridge the gap between laboratory breakthroughs and real-world industrial deployment, while training the next generation of specialists to operate in that space.

This photo shows the Physical AI Test Lab at Jeonbuk National University, May 19. Korea Times photo by Choi Won-suk
From lab to factory floor
The test lab is the first phase of the university's broader AI transformation strategy, launched as a Physical AI Proof of Concept project with 21.9 billion won ($14.4 million) in government funding. The initiative was carried out in collaboration with the Korea Advanced Institute of Science and Technology, Sungkyunkwan University, Hyundai Motor Group, Jeonbuk Technopark and three other regional manufacturers.
By replicating real factory conditions, the facility allows researchers and industry partners to evaluate whether AI-powered manufacturing systems can hold up outside the lab — testing both performance and reliability under production-like scenarios.
"The project has shown that physical AI is no longer confined to laboratories and can be successfully applied in manufacturing settings," said Jeonbuk National University President Yang O-bong.
The pilot project produced measurable results. According to the university, three automotive parts manufacturers participating in the project applied physical AI technologies to key production processes, resulting in gains in both productivity and quality. Production output increased by between 5.1 percent and 11.4 percent, while processing times were reduced by around 10 percent.
Yang highlighted that the results demonstrate the potential of physical AI to help manufacturers address growing challenges, including labor shortages, rising production costs and increasing demand for customized production.

A researcher inspects a robot at Jeonbuk National University's Physical AI Center, May 19. Korea Times photo by Choi Won-suk
Two zones, two purposes
The Manufacturing Technology Test Lab is divided into two sections: the Production Zone (P-Zone) and the Innovation Zone (I-Zone).
The P-Zone serves as a simulated factory floor where AMRs, industrial robots and machine vision systems work together in an integrated production process. From transporting materials and assembling components to inspection and shipment, the entire workflow is coordinated through AI-powered systems that monitor operations as they happen.
The facility was designed to test manufacturing scenarios involving high-mix, low-volume production, a model expected to become increasingly common as consumer demand grows more diverse and customized.
The I-Zone focuses on developing and testing next-generation physical AI technologies.
The space is equipped with humanoid robots, quadruped robots and other advanced robotic systems that allow researchers to collect training data and evaluate how AI models perform in real-world environments. The facility is also used to test emerging technologies, including AI-powered robotic control, multirobot collaboration and systems that enable robots to transfer skills learned in virtual simulations to physical settings.
Kim Soon-tae, a software engineering professor at Jeonbuk National University and head of the university's Physical AI Convergence Technology Project Group, said the initiative is aimed at developing homegrown technologies rather than relying entirely on foreign software platforms.
"The goal is not simply to import overseas software, but to develop and standardize the collaborative intelligence that allows robots, machines and production systems to work together seamlessly," Kim said.
"Ultimately, we want to create factories that operate as a single integrated system rather than a collection of separate machines."

A robot operates in response to commands displayed on a monitor at Jeonbuk National University's Physical AI Center, May 19. Korea Times photo by Choi Won-suk
Toward 'dark factory'
The university plans to continue working with industry partners to validate emerging technologies in real-world settings, with the goal of establishing North Jeolla Province as a hub for next-generation manufacturing innovation and a key player in the global physical AI ecosystem.
The effort is also being reinforced through international collaboration. In March, Jeonbuk National University launched the JBNU-Purdue Research Institute in partnership with Purdue University, creating a joint platform for research, education and technology development.
The partnership is expected to expand the university's global research network and accelerate collaboration, supporting its broader ambition to become a globally competitive innovation hub.
Beyond developing individual technologies, the university is pursuing a larger vision for the future of manufacturing.
Rather than simply exporting robots, it aims to help build and commercialize complete operating systems for AI-powered factories that integrate robotics, autonomous decision-making and smart manufacturing technologies.
Its long-term goal is the realization of the "dark factory," a fully automated manufacturing environment where AI-powered robots oversee production 24 hours a day with little human involvement.