Equipment Condition Identification Based on Telemetry Signal Clustering

This paper deals with the problem of pattern detection in telemetry data, in particular, the approach of automatic machine state detection based on the vibration signal proposed. The approach based on the analysis of the signal via clustering. The paper provides basic information about telemetry data analysis, vibration data analysis, and machine condition monitoring. Also, an overview of cluster analysis methods provided. The proposed approach based on clustering of objects represented with feature set extracted from vibration signals. Given the explanation of the technique and illustrative example of the application of the proposed approach applied to vibration data provided by SmartEdge Agile device for industrial electric motor considered.