Predictive Maintenance on Milling Machines Using Long Short-Term Memory and Support Vector Machines

In the era of Industry 4.0, predictive maintenance has become a key strategy to improve the reliability and efficiency of industrial equipment by anticipating potential failures. This study develops and implements machine learning models to predict the Remaining Useful Life (RUL) of machinery and classify machine conditions for predictive maintenance. Using multi-sensor data, including air temperature, process temperature, rotational speed, torque, and tool wear-this research integrates Long Short-Term Memory (LSTM) and Support Vector Machine (SVM) models. The LSTM model handles time-series data to predict RUL, while the SVM model classifies machinery conditions based on features derived from LSTM outputs. An experimental approach with quantitative analysis divides the data into training and testing sets to validate model performance. Metrics such as accuracy, precision, recall, F1-score, and AUC (Area Under Curve) evaluate the models' effectiveness in predicting failures and classifying conditions. Results show that the combination of LSTM and SVM improves predictive maintenance accuracy, facilitating better repair scheduling and reducing unplanned downtime. This study provides valuable insights into the practical use of machine learning in industrial settings, advancing smart manufacturing and data-driven maintenance strategies.