A wayside monitoring system for rail wheel defect detection using thermography and dynamic measurements
Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit · 2026-11 · 240 (10) · 1216-1226
DOI: 10.1177/09544097261469246 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, online erschienen am 28.07.2026. Verfasst von Enrico Lubbe und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 54 Literaturangaben. Gefördert von: Gibela Rail Transport Consortium RF (Pty) Ltd.
Abstract (Originalsprache)
This research develops and evaluates a wayside monitoring system for rail-wheel tread defect detection that integrates thermography with conventional rail-mounted accelerometer measurements. The study has two aims: (1) design and validate an in-situ thermographic wayside system for detection and geometric characterisation of wheel-tread defects under depot conditions; and (2) benchmark thermography against an accelerometer-based reference with respect to detection performance, localisation, speed sensitivity, environmental robustness, processing time and cost. An experimental campaign on two X’Trapolis Mega train sets within a depot produced the dataset used for evaluation. The thermographic system records thermograms on the wheel tread for approximately one and a half wheel revolutions on the near rail. The analysis included calibrating the measurement field to map various railway coordinates, upon which a captured wheel flat could be back-projected to estimate the length of the defect. The accelerometer system comprised five rail-foot uniaxial sensors and two-point lasers for axle counting. Time-domain statistical features (peak-to-peak, root-mean-square, kurtosis, crest factor) were fused into a joint metric for automated flagging. Results demonstrate that at depot speed (3 km/h), both methods achieved high detection accuracy. The accelerometer joint metric yielded F1-scores of 0.96 (3 km/h), 1.00 (10 km/h) and 0.96 (20 km/h). Thermography produced perfect detection on the tested set at 3 km/h and revealed surface defects missed by manual inspection. Thermographic length estimation of captured wheel defects achieved typical errors of ±1.5 mm for defects ≤100 mm at speeds up to 10 km/h when compared against the manually inspected defects. Performance degraded at 20 km/h (mean error = 8.7%) due to motion blur and reprojection limitations. The study concludes that thermography is a feasible technique for wayside monitoring.
Förderung
- Gibela Rail Transport Consortium RF (Pty) Ltd
Zitieren
APA
Lubbe, E., & Bekker, A. (2026). A wayside monitoring system for rail wheel defect detection using thermography and dynamic measurements. Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, 240(10), 1216-1226. https://doi.org/10.1177/09544097261469246
Vancouver
Lubbe E, Bekker A. A wayside monitoring system for rail wheel defect detection using thermography and dynamic measurements. Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit. 2026;240(10):1216-1226. doi:10.1177/09544097261469246
BibTeX
@article{Lubbe2026,
author = {Lubbe, Enrico and Bekker, Anriëtte},
title = {A wayside monitoring system for rail wheel defect detection using thermography and dynamic measurements},
journal = {Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit},
year = {2026},
volume = {240},
number = {10},
pages = {1216--1226},
doi = {10.1177/09544097261469246}
}
EUVIMED