An Analysis of the International Level of Machine Learning Application Research in Korean Manufacturing Quality Management

This study evaluates South Korea’s global competitiveness in machine learning applications for manufacturing quality management in the context of the transition to ‘Quality 4.0’ using a bibliometric analysis. A total of 1,624 articles, including 99 Korean studies, published between 2010 to 2025 were analyzed using Scopus data and a two-stage AI-assisted screening process. This study introduces a novel three-stage Machine Learning Maturity Framework (Level 1~3) and an time constraints analysis to enable an in-depth assessment of qualitative maturity levels. The results show that Korea ranks fifth globally in research volume and demonstrates world-class competitiveness in Level 1 (Defect Detection) and Level 2 (Virtual Metrology), particularly in the semiconductor and electronics industries, which primarily drive this competitiveness. However, whereas global research is increasingly shifting toward process optimization across diverse domains, Korea exhibits a relatively low proportion of Level 3 (Control) studies. Furthermore, although online monitoring of field data is prevalent (45.5%), real-time closed-loop control remains limited (20.2%), indicating a technological gap between sensing and control. Consequently, this study leverages Korea's strengths in data infrastructure and inspection technologies to propose an industry-specific roadmap for transitioning from monitoring-centric toward explainable autonomous control and PMQ (Production-Maintenance-Quality) integrated systems.