Intelligent predictive maintenance model for rolling components of a machine based on speed and vibration

Machines have come a long way, from the industrial revolution to a modern-day industry 4.0. In this massive transition, one thing that has never changed within a machine is the moving part. Most industries use rotating machine with different load capacity and speed. These machines run at variable load and variable speed creating vibration bootstrap thus causing machine failure due to an increase in vibrations. Most of the researcher used vibration for fault detection in bearings but sometimes it caused by miss alignment in a shaft due to a fraction of overloading the machine. In this paper, we address it to solve those problems by using two parameters speed and vibration. To verify our approach, we use three different kinds of machine learning algorithms: Support Vector Machine (SVM), Naïve Bays, and Random Forest. By using these machine learning algorithms, we tried to find out the relationship between machine failure due to speed and vibration by predicting good and faulty bearings. After applying these models, we have seen that the SVM has 78% accuracy as compared to Naïve Bays, and Random Forest.