"AI Solutions for Industrial Challenges"... LG Unveils Expert AI for Manufacturing, Finance, and Science

Solving On-Site Challenges with AI Specialized in Manufacturing, Science, and Finance
368 Papers Presented at AI Conferences, 1,080 Patents Filed
Development of Robot Foundation Model: Toward Autonomous Factories

LG AI Research Institute has unveiled specialized AI (artificial intelligence) solutions designed to address challenging issues in the manufacturing, finance, and science sectors.

On September 14, at the Convergence Hall of LG Science Park in Magok, Gangseo-gu, Seoul, LG AI Research Institute hosted the ‘LG AI Talk Concert 2026’, where it showcased AI technologies tailored to various industries and presented real-world application cases.

Woo Hyung Lim, Co-Director of LG AI Research Institute, explained during his keynote speech, “While creating high-quality AI models is important, LG AI Research Institute’s mission is to solve difficult problems that have long persisted in the industrial field.” He further stated, “There already exist countless general-purpose AI models in the world that generate excellent answers. However, to solve issues in the industrial landscape, we needed AI that could understand and judge countless variables and even the rare 1% of exceptions. Only by demonstrating tangible performance can AI prove its value at industrial sites.” This was his rationale for focusing on the development of expert AI.

During the event, LG AI Research Institute showcased specific achievements of its expert AI systems developed for key sectors such as manufacturing, science, and finance. In manufacturing, it introduced ‘EXAONE Tabular’, which understands and predicts process conditions and quality information, and ‘EXAONE Omni-Inspect’, which can inspect processes and products even when external images change, without requiring retraining.

Soonyoung Lee, Head of the Data Intelligence Lab at LG AI Research Institute, and Jungseon Yoo, Head of AI Transformation (AX) at LG Innotek, explained that by using industry-specific AI foundation models, rapid and accurate predictions can be made with only small amounts of data, and that these models flexibly respond to changing process conditions in manufacturing environments. LG AI Research Institute stated that its vision inspection agent autonomously performs data sampling, labeling, and model training, marking the evolution of manufacturing AI toward a phase where it solves problems independently.

LG AI Research Institute. LG Corporation.

In the science sector, the institute presented ‘EXAONE Discovery’, which designs materials and predicts synthesis results. LG AI Research Institute is planning to establish an autonomous experimental system that goes beyond material design and result prediction to conduct actual experiments. An example was unveiled where, in collaboration with LG Household & Health Care, the institute reviewed 420,000 candidate compounds and discovered the hair loss treatment ingredient ‘Rhamsydil’ in a single day. Another ongoing project with D&D Pharmatech involves developing next-generation oral peptide drugs, and ‘EXAONE 4.5’ has been integrated with the AI system for new drug screening at the Ministry of Food and Drug Safety. It is expected to help shorten future new drug approval periods.

In the financial sector, LG AI Research Institute highlighted the expanded business achievements of its explainable prediction solution, ‘EXAONE Business Intelligence (BI)’. Hwayoung Lee, Head of AI Business Development, explained, “After the official launch of EXAONE BI, a finance-specialized AI solution comprising multiple AI agents that collaborate to generate everything from data analysis and inference to prediction and explanation, we have expanded our business globally with the London Stock Exchange Group (LSEG), and in the domestic market with Koscom.”

‘EXAONE Forecast’, an AI foundation model capable of understanding information as it changes over time, has been integrated into EXAONE BI. LG AI Research Institute plans to expand its business in South Korea, North America, Europe, and the Middle East, and to broaden the use of explainable and reliable predictive AI applications to a variety of asset classes, including bonds and commodities.

Since its establishment in December 2020, LG AI Research Institute has addressed more than 100 industry challenges, including battery life and capacity prediction, defective product screening, optimization of production and material management plans, and discovery of new drug candidates. Lim emphasized, “We were able to tackle these difficult challenges because of our top-tier AI research capabilities and advanced technologies, demonstrated through 368 presentations at the world’s leading AI conferences and 1,080 patent applications filed domestically and overseas.”

LG AI Research Institute is also preparing for the next phase in which AI acts directly on site. Lim introduced the ongoing development of a robot foundation model to serve as the ‘brain’ of robots, and explained that the team is researching source algorithms that prioritize safety to prevent accidents caused by robot operation in advance.

He stated, “Over the past six years, LG has enabled AI to understand industrial problems, and now we are preparing for a stage where AI can make decisions and take action autonomously. Based on the robot foundation model, we are moving beyond the automation of individual robots towards a future where entire factories operate as one intelligent, autonomous system.”

Meanwhile, Honglak Lee, Co-Director and Chief Scientist for AI (CSAI) at LG AI Research Institute, held a special invitation session and panel discussion on the topic of ‘Global AI Trends and Insights on AI in Korea’. As AI evolves from simply answering questions to acting as an agent capable of performing actual work, the discussion also covered the competencies required for AI and the future direction of the domestic AI industry.

This content was produced with the assistance of AI translation services.