Early Detection of Bearing Failures Based on Temperature Trends Deviations Analysis

In this paper, we consider a practical problem of detecting upcoming mechanical failures of bearings in industrial pumps and motors equipped with temperature sensors. Several data processing procedures and early failures detection analytics based on temperature trends analysis have been designed and tested in the field. The paper results in a Fourier Transform-based predictive diagnostic algorithm ready to be implemented for any kind of rotating equipment.