Wheel defects are a significant concern in rail transport, impacting safety and increasing infrastructure wear. This study presents an automated framework for early detection and severity classification of wheel flats using data from a single wayside-mounted accelerometer. Acceleration signals are processed to extract time-domain features, which are then classified using machine learning (ML) algorithms, including K-Nearest Neighbors (kNN), Multi-Class Support Vector Machine (MSVM), Decision Tree (DT), Ensemble Tree (ET), Gaussian Mixture Model (GMM), Naive Bayes (NB), Random Forest (RF), and Discriminant Analysis (DA). The results indicate that the proposed approach consistently achieves over 90% accuracy in detecting and classifying wheel flats, irrespective of the sensor’s placement—whether on the rail or between sleepers—and with the use of only a single sensor. This cost-effective and scalable approach minimizes sensor requirements, making it practical for widespread implementation across large-scale railway networks while ensuring high accuracy in defect detection.
NielsenJCOJohanssonA. Out-of-round railway wheels-a literature survey. Proc Inst Mech Eng F J Rail Rapid Transit2000; 214(2): 79–91.
2.
JohanssonAAnderssonC. Out-of-round railway Wheels—a study of wheel polygonalization through simulation of three-dimensional wheel–rail interaction and wear. Veh Syst Dyn2005; 43(8): 539–559.
3.
LanQDhanasekarMHandokoY. Wear damage of out-of-round wheels in rail wagons under braking. Eng Fail Anal2019; 102: 170–186.
4.
BarkeDWChiuWK. A review of the effects of out-of-round wheels on track and vehicle components. Proc Inst Mech Eng F J Rail Rapid Transit2005; 219(3): 151–175.
5.
ZhangZEntezamiMStewartE, et al.Enhanced fault diagnosis of roller bearing elements using a combination of empirical mode decomposition and minimum entropy deconvolution. Proc IME C J Mech Eng Sci2015; 231(4): 655–671.
6.
TraquinhoNValeCRibeiroD, et al.Damage identification for railway tracks using onboard monitoring systems in In-Service vehicles and data science. Machines2023; 11(10): 981.
7.
MoslehACostaPCalçadaR. A new strategy to estimate static loads for the dynamic weighing in motion of railway vehicles. Proc Inst Mech Eng F J Rail Rapid Transit2020; 234(2): 183–200.
8.
PintãoBMoslehAValeC, et al.Development and validation of a weigh-in-motion methodology for railway tracks. Sensors2022; 22(5): 1976.
9.
MoslehAMontenegroPAAlves CostaP, et al.An approach for wheel flat detection of railway train wheels using envelope spectrum analysis. Structure and Infrastructure Engineering2021; 17(12): 1710–1729.
10.
MoslehAMontenegroPCostaP, et al.Railway vehicle wheel flat detection with multiple records using spectral kurtosis analysis. Applied Sciences2021; 11(9): 4002.
11.
PimentelRRibeiroDMatosL, et al.Bridge weigh-in-motion system for the identification of train loads using fiber-optic technology. Structures2021; 30: 1056–1070.
12.
ZhaoATianGYZhangJ. IQ signal based RFID sensors for defect detection and characterisation. Sensor Actuator Phys2018; 269(1): 14–21.
13.
AlemiACormanFPangY, et al.Evaluation of the influential parameters contributing to the reconstruction of railway wheel defect signals. Proc Inst Mech Eng F J Rail Rapid Transit2019; 234(9): 1005–1016.
14.
WeiCXinQChungWH, et al.Real-Time train wheel condition monitoring by fiber bragg grating sensors. Int J Distrib Sens Netw2011; 8(1): 409048.
15.
SunQChenCKempAH, et al.An on-board detection framework for polygon wear of railway wheel based on vibration acceleration of axle-box. Mech Syst Signal Process2021; 153: 107540.
16.
NielsenJCOIgelandA. Vertical dynamic interaction between train and track influence of wheel and track imperfections. J Sound Vib1995; 187(5): 825–839.
17.
SladkowskiASitarzM. Analysis of wheel–rail interaction using FE software. Wear2005; 258(7–8): 1217–1223.
18.
ZengYSongDZhangW, et al.A new physics-based data-driven guideline for wear modelling and prediction of train wheels. Wear2020; 456-457: 203355.
19.
DavariNV, BVelosoBCostaGA, et al.A survey on data-driven predictive maintenance for the railway industry. Sensors2021; 21: 5739.
20.
WangHLiHLiY, et al.Railway wagon wheelset fault diagnosis method based on DBN (PHM-Shanghai). Global Reliability and Prognostics and Health Management2020: 1–6.
21.
LiZHeQ. Prediction of railcar remaining useful life by multiple data source fusion. IEEE trans Intell Transp Syst2015; 16(4): 2226–2235.
22.
LiYZuoMLinJ, et al.Fault detection method for railway wheel flat using an adaptive multiscale morphological filter. Mech Syst Signal Process2017; 84: 642–658.
23.
NiYZhangQ. A Bayesian machine learning approach for online detection of railway wheel defects using track-side monitoring. Struct Health Monit2020; 20(4): 1536–1550.
24.
MoslehAMeixedoARibeiroD, et al.Early wheel flat detection: an automatic data-driven wavelet-based approach for railways. Veh Syst Dyn2023; 61(6): 1644–1673.
25.
NickWAsameneKBullockG, et al.A study of machine learning techniques for detecting and classifying structural damage. Int J Mach Learn Comput2015; 5(4): 313–318.
26.
MoslehAMeixedoARibeiroD, et al.Automatic clustering-based approach for train wheels condition monitoring. Int J Real Ther2023; 11(5): 639–664.
27.
AddinOSapuanSMMahdiE, et al.A naïve-bayes classifier for damage detection in engineering materials. Mater Des2007; 28(8): 2379–2386.
28.
VitolaJPozoFTibaduizaD, et al.Distributed piezoelectric sensor system for damage identification in structures subjected to temperature changes. Sensors2017; 17(6): 1–26.
29.
ShafiqueRSiddiquiHRustamF, et al.A novel approach to railway track faults detection using acoustic analysis. Sensors2021; 21(18): 6221.
30.
KrummenacherGOngCSKollerS, et al.Wheel defect detection with machine learning. IEEE Trans Intell Transp Syst2018; 19(4): 1176–1187.
31.
ShuJZhangZGonzalezI, et al.The application of a damage detection method using artificial neural network and train-induced vibrations on a simplified railway bridge model. Eng Struct2013; 52: 408–421.
32.
ZhaiW. Vehicle-track coupled dynamics: theory and applications. Springer Nature, 2020.
33.
AktasMGunelEHYılmazerP, et al.Detection of wheel flatten defect on the moving train with acoustic emission sensor. 2018 IEEE 29th Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), Bologna, Italy, 20 December 2018. IEEE.
34.
BuckleyTGhoshBPakrashiV. A feature extraction & selection benchmark for structural health monitoring. Struct Health Monit2023; 22(3): 2082–2127.
35.
GhiasiRGordanMMoslehA, et al.M-CLUSTER: multistage clustering for unsupervised train wheel condition monitoring. Veh Syst Dyn2024: 1–26.
36.
GhiasiRMalekjafarianA. Monitoring of railway tracks maintenance needs using dynamic responses collected by an in-service train. Railw Eng Sci2025.
37.
GhiasiRand AMalekjafarianA. Feature subset selection in structural health monitoring data using an advanced binary slime mould algorithm. J Struct Integr Main2023; 8(4): 209–225.
38.
SonHKimCHwangN, et al.Classification of major construction materials in construction environments using ensemble classifiers. Adv Eng Inform2014; 28(1): 1–10.
39.
MartínezIL. Structural damage assessment under uncertainty. PhD Dissertation, University of California, 2010.
40.
WuXKumarVRoss QuinlanJ, et al.Top 10 algorithms in data mining. Knowl Inf Syst2008; 14(1): 1–37.
41.
GordanMSabbagh-YazdiSRIsmailZ, et al.Data mining-based structural damage identification of composite bridge using support vector machine. Journal of AI and Data Mining2021; 9(4): 415–423.
42.
SugumaranVSabareeshGRRamachandranKI. Fault diagnostics of roller bearing using kernel based neighborhood score multi-class support vector machine. Expert Syst Appl2008; 34(4): 3090–3098.
43.
AmbwaniT. Multi class support vector machine implementation to intrusion detection. Proceedings of the International Joint Conference on Neural Networks, 2003, Portland, OR, USA, 20–24 July 2003.
44.
DuanHDengZDengF, et al.Assessment of groundwater potential based on multicriteria decision making model and decision tree algorithms. Math Probl Eng2016; 2016: 1–11.
45.
ZhangLSuganthanPN. Oblique decision tree ensemble via multisurface proximal support vector machine. IEEE Trans Cybern2015; 45(10): 2165–2176.
46.
GordanMSabbagh-YazdiS-RIsmailZ, et al.State-of-the-art review on advancements of data mining in structural health monitoring. Measurement2022; 193: 110939.
47.
XuWFanSWangC, et al.Leakage identification in water pipes using explainable ensemble tree model of vibration signals. Measurement2022; 194: 110996.
48.
ShiPWangZ. An ensemble tree classifier for highly imbalanced data classification. J Syst Sci Complex2021; 34(6): 2250–2266.
49.
WengJMengQ. Ensemble tree approach to estimating work zone capacity. Transp Res Rec2012; 2286(1): 56–67.
GeleteGYaseenZM. Hybridization of deep learning, nonlinear system identification and ensemble tree intelligence algorithms for pan evaporation estimation. J Hydrol2024; 640: 131704.
52.
LiuJCaiDHeX. Gaussian mixture model with local consistency. Proc AAAI Conf Artif Intell2010; 24(1): 512–517.
53.
WanHWangHScotneyB, et al.A novel Gaussian mixture model for classification. 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC), Bari, Italy, 06-09 October 2019.
54.
LiJQBarronAR. Mixture density estimation. Advances in neural information processing systems, 2000, pp. 279–285.
55.
GordanMRazakHAIsmailZ, et al.Recent developments in damage identification of structures using data mining. Lat Am J Solid Struct2017; 14(13): 2373–2401.
56.
DawseyWJMinskerBSVanBlaricumVL. Bayesian belief networks to integrate monitoring evidence of water distribution system contamination. J Water Resour Plan Manag2006; 132(4): 234–241.
57.
MohammadzadehSHeydariHKarimiM, et al.Correlation analysis of railway track alignment and ballast stiffness: comparing frequency-based and machine learning algorithms. Algorithms2024; 17: 372.
58.
DabbaATariAMeftaliS. A new multi-objective binary Harris Hawks optimization for gene selection in microarray data. J Ambient Intell Hum Comput2023; 14(4): 3157–3176.
59.
BelgiuMDrăguţL. Random forest in remote sensing: a review of applications and future directions. ISPRS J Photogrammetry Remote Sens2016; 114: 24–31.
60.
RamayahTAhmadNHHalimHA, et al.Discriminant analysis: an illustrated example. Afr J Bus Manag2010; 4(9): 1654.
61.
KandiriAGhiasiRNogalM, et al.Travel time prediction for an intelligent transportation system based on a data-driven feature selection method considering temporal correlation. Transport Eng2024; 18: 100272.
62.
KazemitabarJAminiABloniarzA, et al.Variable importance using decision trees. Advances in neural information processing systems (NIPS2017), 2017. 30.
63.
AfsharmovahedMHGhodrati AmiriGDarvishanE. A novel damage detection approach based on feature extraction and selection using machine learning without signal processing: a case study on the tianjin Yonghe Bridge. Iran J Sci Technol Trans Civ Eng2023; 47(6): 3649–3661.
64.
MontenegroPANevesSGMCalçadaR, et al.Wheel-rail contact formulation for analyzing the lateral train-structure dynamic interaction. Comput Struct2015; 152: 200–214.
65.
MontenegroPAHelenoRCarvalhoH, et al.A comparative study on the running safety of trains subjected to crosswinds simulated with different wind models. J Wind Eng Ind Aerod2020; 207: 104398.
66.
HertzH. Ueber die Berührung fester elastischer Körper [On the contact of elastic solids]. J für die Reine Angewandte Math (Crelle's J)1882; 92: 156–171.
67.
KalkerJJ. Book of tables for the Hertzian creep-force law. 2nd Mini Conference on Contact Mechanics and Wear of Wheel/Rail Systems, Budapest, Hungary, 1996.
68.
MATLAB®. 2018, Natick, MA, USA: release R2018a, The MathWorks Inc.
69.
ANSYS®. 2018, Academic Research, Release 19.2, ANSYS Inc.
70.
MohammadiMMoslehAValeC, et al.An unsupervised learning approach for wayside train wheel flat detection. Sensors2023; 23(4): 1910.
71.
MoslehAMontenegroPACostaPA, et al.Approaches for weigh-in-motion and wheel defect detection of railway vehicles. Rail Infrastructure Resilience. Elsevier, 2022, pp. 183–207.
72.
SilvaRGuedesARibeiroD, et al.Early identification of unbalanced freight traffic loads based on wayside monitoring and artificial intelligence. Sensors2023; 23(3): 1544.
73.
GuedesASilvaRRibeiroD, et al.Detection of wheel polygonization based on wayside monitoring and artificial intelligence. Sensors2023; 23(4): 2188.
74.
GonçalvesVMoslehAValeC, et al.Wheel out-of-roundness detection using an envelope spectrum analysis. Sensors2023; 23(4): 2138.
75.
PintãoBMoslehAValeC, et al.Development and validation of a weigh-in-motion methodology for railway tracks. Sensors2022; 22(5): 1976.
76.
LourençoAFerrazCRibeiroD, et al.Adaptive time series representation for out-of-round railway wheels fault diagnosis in wayside monitoring. Eng Fail Anal2023; 152: 107433.
77.
ZhaiWMWangQCLuZW, et al.Dynamic effects of vehicles on tracks in the case of raising train speeds. Proc Inst Mech Eng F J Rail Rapid Transit2001; 215(2): 125–135.
78.
PieringerAKroppWNielsenJCO. The influence of contact modelling on simulated wheel/rail interaction due to wheel flats. Wear2014; 314(1–2): 273–281.
79.
WuTXThompsonDJ. A hybrid model for the noise generation due to railway wheel flats. J Sound Vib2002; 251(1): 115–139.
80.
IşikFInderyasOSuzukiT, et al.Enhancing acoustic emission analysis for corrosion damage detection in reinforced concrete by KNearest neighbors method in EWSHM 2024- 11th European Workshop on Structural Health Monitoring2024. https://doi.org/10.58286/29665
81.
European Standard - EN 15313, Railway applications - In-service wheelset operation requirements - In-service and off - vehicle wheelset maintenance. 2016: Brussels.
82.
BelottiVCrennaFMicheliniRC, et al.Wheel-flat diagnostic tool via wavelet transform. Mech Syst Signal Process2006; 20(8): 1953–1966.
83.
NowakowskiTKomorskiPSzymańskiGM, et al.Wheel-flat detection on trams using envelope analysis with Hilbert transform. Lat Am J Solid Struct2019; 16: e148.
84.
MohammadiMMoslehAValeC, et al.Smart railways: AI-based track-side monitoring for wheel flat identification. Proc Inst Mech Eng, Part F2025; 239(3): 272–289.
85.
SalehiMBagherzadehSAFakhariM. Experimental detection of train wheel defects using wayside vibration signal processing. Struct Health Monit2023; 22(5): 3286–3301.
86.
JorgeTMagalhãesJSilvaR, et al.Early identification of out-of-roundness damage wheels in railway freight vehicles using a wayside system and a stacked sparse autoencoder. Veh Syst Dyn2024: 1–26.
87.
FilogranoMLGuillénPCRodríguez-BarriosA, et al.Real-time monitoring of railway traffic using fiber Bragg grating sensors. IEEE Sens J2011; 12(1): 85–92.
88.
MishraSSharanPSaaraK. Real time implementation of fiber Bragg grating sensor in monitoring flat wheel detection for railways. Eng Fail Anal2022; 138: 106376.
89.
KomorskiPSzymanskiGMNowakowskiT, et al.Advanced acoustic signal analysis used for wheel-flat detection. Lat Am J Solid Struct2021; 18: e338.
