Inbizzy, Gumi, South Korea — Intel and LG Innotek have announced a strategic collaboration to deploy an AI-powered smart factory at LG Innotek’s production facility in Gumi, South Korea. The initiative integrates Intel® Core™ and Intel® Xeon® processors, Intel® Arc™ GPUs, and the OpenVINO™ toolkit to enhance inspection accuracy and production efficiency in high-precision electronic component manufacturing.
AI-Driven Automated Inspection System
The project began in 2024 when LG Innotek started developing an AI-powered inspection system supported by Intel technologies. The system operates in several stages:
- Intel Core CPUs process defect data from production lines in real time using integrated graphics for cost-effective analysis.
- Intel Arc GPUs handle heavy workloads, including high-resolution image processing.
- Intel Xeon CPUs manage pre-training of AI models with large datasets.
- Future plans include integrating Intel® Gaudi® AI accelerators to further accelerate AI model training.
This approach significantly reduces the cost of building AI inspection systems while enabling large-scale manufacturing deployment.
Cost Efficiency and Production Expansion
The adoption of Intel Arc GPUs in LG Innotek’s main facility has delivered better cost efficiency compared to alternative solutions. The AI vision inspection system was first implemented on the company’s mobile camera module production line in 2024.
In 2025, LG Innotek plans to expand the system to its Gumi 4 facility, which produces flip-chip ball grid arrays (FC-BGA), as well as to overseas production sites in stages.
OpenVINO Simplifies AI Integration
OpenVINO, Intel’s open-source AI toolkit launched in 2018, has simplified the integration process by allowing developers to write AI models once and deploy them across multiple platforms without significant code modifications.
AI Model Retraining with Intel Xeon CPUs
Going forward, LG Innotek plans to leverage Intel Xeon CPUs with built-in AI accelerators, including Intel® Advanced Matrix Extensions, to retrain deep learning models as production processes or raw materials change.
By reducing reliance on third-party GPUs for large-scale AI workloads, the company expects to optimize costs and improve operational efficiency.
This collaboration is expected to accelerate digital transformation in the electronics manufacturing sector, delivering cost-effective and AI-driven smart factory solutions.
