Computer algorithm to detect and predict machine faults using cloud-based vibration data

In this research a machine fault detection and diagnostic algorithm is presented. The algorithm uses time wave-form acceleration data stored in a server for cloud computing to calculate RMS and Peak values from it and give information to the user for maintenance schedule. Detection algorithm analyses the change in time of the acceleration signals and establish urgency and severity of the studied machines. Furthermore, the diagnosis sub-system studies also the change in time of the signals in frequency domain to give a forecast of the possible existing fault by discarding faults throughout a predetermined decision table. Simulated and real cases are performed to show the efficiency and results of using the algorithm as well.