Enhancing product predictive quality control using Machine Learning and Explainable AI

Abstract

The integration of predictive quality and eXplainable Artificial Intelligence (XAI) in product quality classification marks a significant advancement in quality control processes. This study examines the application of Machine Learning (ML) models and XAI techniques in managing product quality, using a case study in the agri-food industry quality as an example. Predictive quality models leverage historical and real-time data to anticipate potential quality issues, thereby improving detection accuracy and efficiency. XAI ensures transparency and interpretability, facilitating trust in the model’s decisions. This combination enhances quality management, supports informed decision-making, and ensures regulatory compliance. The case study demonstrates how ML models, particularly Artificial Neural Network (ANN), can accurately predict product quality, with XAI providing clarity on the reasoning behind these predictions. The study suggests future research directions, such as expanding datasets, exploring advanced ML techniques, implementing real-time monitoring, and integrating sensory analysis, to further improve the accuracy and transparency of quality control in various industries.