Abstract

The growing demand for high-quality production and fast delivery has pushed manufacturers to rethink the traditional quality control methods. Although Six Sigma and other quality control tools provide a systematic quality improvement framework, they often struggle with data generated in today's modern production environments because of its volume and complexity. This study responds to this challenge by integrating Artificial Intelligence (AI) with the Six Sigma DMAIC methodology. This study was conducted using production data of KAC Fashion Wear Ltd., a renowned garment manufacturer in Bangladesh. This work creates an adaptive model programmed to predict potential defects, track the root cause, and help in decision-making to improve quality. Prior work has generally considered AI and Six Sigma as distinct tools, one anchored on statistical control, the other on machine learning, thus leaving a distinct gap in predictive quality management. Applying production data from garments, the combined Six Sigma - AI model (CART) was able to ascertain the accuracy of prediction as 76.93%, holding promise towards quality issues forecasting before occurrence. The CART results also helped identify the key variables that contribute to defects, allowing more precise corrective actions to be taken. More broadly, the findings suggest that AI can extend the analytical reach of Six Sigma, turning it from a backward-looking framework into a practical, adaptive system for continuous improvement and sustainable quality control across all manufacturing environments.