Paper Title
ENHANCING PREDICTIVE MAINTENANCE WITH SHAP AND LIME: A FRAMEWORK FOR EXPLAINABLE AI

Abstract
One of the most important strategies for reducing machine downtime and increasing operational effectiveness in industrial settings is predictive maintenance. Traditional machine learning models may forecast equipment failures with great accuracy, but their interpretability is generally limited by their complexity, which presents problems in high-stakes situations when explainability is crucial. In order to improve predictive maintenance, this study integrates two cutting-edge explainable AI techniques: SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-Agnostic Explanations). The framework enables maintenance engineers to comprehend the critical elements influencing each prediction in addition to making highly accurate failure predictions by applying these techniques to actual industrial datasets. The results we obtain show how well this strategy works to improve decision-making, offer clear, useful insights, and build confidence in AI-driven maintenance solutions. The implications for industrial AI, model interpretability, and future possibilities for improving explainable predictive maintenance systems are discussed in the paper's conclusion. Keywords - SHAP(Shapley Additive exPlanations),LIME(Local Interpretable Model-Agnostic Explanations), Predictive maintenance (PdM), Explainable AI (XAI) , Black-box , Remaining Useful Life(RUL) , ROC(Receiver Operating Characteristic curve),AUC(Area Under the Curve), Logistic Regression model, DecisionTree Classifier, Random Forest Model.