Intelligent Machine Fault Detection in Industries using Supervised Machine Learning Techniques

In the modern epoch of Industry 4.0, Artificial Intelligence (AI) has made huge industry performances to shape our daily lives, increase production efficiency, and reduce economic losses. Predictive maintenance identifies the equipment faults earlier and reduces unexpected failures and sudden machine shutdowns. IoT sensors are fitted with the machinery parts, and the data gathered from the IoT sensors detect the misbehavior or faults in the machine. In this paper, three machine learning algorithms are applied for feature selection and prediction of faults in the machine. In the first stage, features from the machine were selected using the logistic regression method. From the selected features, the faults are classified with Support Vector Machine (SVM), Naïve Bayes, and Decision Tree classifier. The performance of the algorithms was evaluated with precision, recall, and accuracy. The results show that the supervised machine learning methods work well in predicting faults in machine parts.