Explainable and Trustworthy AI for Steel Plates Fault Detection Using Shap and Lime
This paper investigates the potential of Explainable Artificial Intelligence (XAI) methods to improve the interpretation of AI-based models in smart manufacturing applications. The study focuses on SHAP and LIME techniques, analyzing their ability to explain model decisions, identify the most influential input features, and provide insights into both individual predictions and overall model behaviour. A practical case study based on the Steel Plates Faults dataset is presented using a Random Forest classification model to investigate the application of explainable AI techniques for steel plate fault detection. The results demonstrate that SHAP and LIME methods provide valuable insights into model decision-making by identifying the key features influencing fault classification. These explainability approaches enhance model transparency and trustworthiness, supporting the development of more reliable AI-based solutions for industrial quality inspection and smart manufacturing applications.