LG AI Research has introduced two AI foundation models designed specifically for manufacturing: EXAONE Tabular for analysing structured production and quality data, and EXAONE Omni-Inspect for automated visual quality inspection. The models are designed to adapt to changing production conditions while reducing the need for repeated model training.
EXAONE Tabular can analyse manufacturing and product-quality data to predict production conditions, while Omni-Inspect uses camera images to identify defects in components and products. LG says the approach can help factories adapt AI systems when products or production processes change without rebuilding the entire model.
LG is also developing a vision inspection agent that can automate parts of data sampling, labelling and model training. Its wider manufacturing AI work includes predictive maintenance, defect detection, production planning and robotics, pointing toward increasingly autonomous factory environments.
