[E-Daily Marketin Soyoung Park Reporter] "In the energy industry as well, the use of artificial intelligence (AI) to improve operational efficiency and safety management could open up new business opportunities."

Lee Hwa-young of LG AI Research offered this outlook on the 5th at the “Energy Leaders Summit (ELS)” held at the Kim Dae-jung Convention Center in Gwangju, stating, “AI technologies used in industrial settings can also be applied to the energy sector.” ELS was held as a major conference within BIXPO 2026.

In manufacturing, AI is emerging as a core technology for boosting productivity and efficiency. This evolution has progressed beyond analyzing vast amounts of industrial data to predict quality and demand to a stage where robots directly inspect production facilities and detect equipment malfunctions.

The researcher pointed out the problem of rapidly rising data and model management costs associated with AI adoption in global manufacturing. LG operates various businesses in more than 160 countries. Since the equipment and production environments differ from factory to factory, AI models for quality inspection must be built individually for each facility. This means that approximately 9,600 AI models are required for a single inspection task. When scaled across global operations, the number of models could rise to over 100,000.

To alleviate this burden, LG AI Research developed “ExaOne Tabular,” an AI model that analyzes tabular data. It is designed to operate quickly even with limited computational resources, allowing it to be deployed on-premises even on systems with low GPU specifications. In addition, the institute is developing various industrial AI technologies, such as AI that predicts the properties of new materials using chemical knowledge and time-series AI that forecasts product-specific demand.

Next, Jason Yoon, Head of Asia at Boston Dynamics, introduced the potential of “Physical AI,” where AI extends into the physical environment of the real world. Boston Dynamics is a robotics company wholly owned by Hyundai Motor Group. It is developing the quadruped robot “Spot,” the humanoid “Atlas,” and the logistics robot “Stretch,” among others.

Yoon explained that the role of robots in industrial settings is expanding beyond simple, repetitive tasks to include equipment inspections and safety management. For example, by equipping Spot with thermal imaging cameras and vibration and sound sensors, it can detect signs of equipment malfunctions that are difficult for workers to identify directly and alert operators in real time. It can also be used to build digital twins through 3D scanning, as well as for patrolling production facilities and security management.

He predicted, in particular, that the use of robots would increase in the maintenance and management of aging manufacturing facilities and energy infrastructure. If robots take over the inspections of power and water treatment facilities—tasks previously performed regularly by humans—work time can be reduced. At the same time, he explained, this can lower the risk of safety accidents. He also cited the ability to integrate robotic solutions without completely replacing existing industrial systems as a factor that lowers the barriers to adoption.

“AI and robotics technologies should be viewed not as replacements for human workers, but as solutions that assist on-site operators and workers,” said Manager Yoon. “We expect to achieve multiple benefits simultaneously, including high-quality inspections, improved productivity, and enhanced safety.”