Fault behaviour pattern analysis and recognition

The dynamics of a faulty gearbox that includes the rolling bearings element allows the engineer and the analyst to study the behavior of damaged parts and collect the data so as to analyze the defect. This helps to recognize it without bearing the cost of actual failures in real time. In this case study, a tiny hardware setup has been used for emulating the similar behavior of a gearbox and the same has been used for analysis with k-Means unsupervised machine learning model for in depth diagnostics of the device.