Abstract
Machine learning a subdivision of artificial intelligence which is popular nowadays particularly where large amounts of data are to be evaluated unsupervised by applying algorithms thereby requiring less human effort to process the measurements. This paper presents an intelligent fault monitoring on the overhead wires using wireless sensor network (WSN). The current method used to monitor failures requires both foot patrols and vehicle measurements using cameras, however these methods are both labour and time intensive in preparing and analysing the data from the inspections. An intelligent method is proposed to reduce the amount of time spent on labour intensive inspections through data aggregation and machine learning. Ma-chine learning offers additional flexibility for identifying the type of the faults, finding otherwise hidden pat-terns and grouping instances of events accordingly based on similarities. WSN will convey the measured data to the cloud via the router for computation thereby providing notifications in real-time and also the data can be viewed anywhere by the operator. K-means clustering algorithm will then be applied later using sensors data via Matlab/Simulink.