An Explainable Artificial Intelligence Framework for Predictive Maintenance in Agricultural Facilities

Abstract

Predictive maintenance has become an essential approach for improving the reliability and operational efficiency of agricultural machinery and facilities. Traditional maintenance strategies such as reactive and preventive maintenance often result in unexpected equipment failures, increased downtime, and higher operational costs. This study proposes an Explainable Artificial Intelligence (XAI) based predictive maintenance system implemented using a Django web framework integrated with machine learning models. The system enables real-time monitoring, data preprocessing, model training, and failure prediction through an interactive web interface. The proposed model utilizes machine learning algorithms such as Logistic Regression, Support Vector Machine, and Random Forest to analyze equipment operational parameters including temperature, rotational speed, torque, and tool wear to predict potential machine failures. The dataset is preprocessed through normalization, encoding, and train–test splitting to improve model performance. The system automatically selects the best-performing model based on accuracy and provides visual performance comparisons using graphical analysis. Additionally, the predictive model allows users to input machine parameters and obtain failure predictions along with probability scores, enhancing interpretability and decision support. By incorporating explainable AI concepts, the system improves transparency in machine learning predictions, allowing agricultural operators and technicians to better understand the factors contributing to equipment failure. The proposed approach supports proactive maintenance planning, reduces downtime, and improves productivity in agricultural facilities. The implementation demonstrates how web-based machine learning systems can be effectively utilized for intelligent predictive maintenance in smart agriculture environments.