DrTrack, Detection and Diagnosis of Railway Track Short Wave Anomalies

January 11, 2016 in

PhD Research by Siamand Rahimi

Rolf Dollevoet, Zili Li

09/01/2012 - ongoing


condition monitoring, short wave anomalies,

Asset Rail, Arcadis, DeltaRail, Eurailscout Inspection & Analysis B.V., Müller-BBM VAS GmbH, Tata Steel, Polytec GmbH, ProRail S&C, Track and RCF Groups, RailData


Link or Download
Not available


Track short waves anomalies such as damaged frogs and blades of S&C, damaged insulated joints and fastening, damaged and hanging sleepers and ballast void cause large dynamic train-track interaction forces. Defects causes by these forces will degrade rapidly and become one of main failure sources.

In order to increase the track safety and availability and reduce the maintenance and renewal cost , defects should be detected at their early stage. This can be achieved by 24/7 track monitoring utilizing Dr. Track measurement system.

Track short wave anomalies cannot be detected easily using existing methods and there is no reliable and automated detection system which can track them. Therefor very often they are detected on the appearance during manual inspection at the very late and severe stage. This is very labour intensive and expensive and the hit rate is normally very low. If these defects are identified at their early stage when the degradation is local and light the maintenance cost can be reduced and the track safety and availability can be enhanced.

DrTrack is devoted to train-borne measurement system which is able to detect and locate the short wave anomalies at their early stage, based on the acquired measurement data and by using identification algorithm. DrTrack is expected to make 24/7 track health monitoring of short wave anomalies possible and might result in optimal design of component and structure of tracks, switches and crossings to reduce the forces and stresses.

Frogs of switch and crossings are chosen for case study due to their criticality in the track system, their complex geometry and the resulting high dynamic train-track interaction forces. Based on laboratory and field testing as well as the identified signature tunes their condition can be assessed and a maintenance method can be selected.

DrTrack will make use of artificial intelligence, data mining and neural network information management to extract and optimize signature tunes and to identify related parameters and thresholds for anomaly detection.

DrTrack and its detection algorithm can be integrated into the information management and decision support systems of the railway infra manager to quickly localize and handle disturbances and to plan maintenance and renewal.