{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,15]],"date-time":"2026-05-15T06:11:45Z","timestamp":1778825505749,"version":"3.51.4"},"reference-count":52,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,5,15]],"date-time":"2026-05-15T00:00:00Z","timestamp":1778803200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Railway track failures such as cracks, displacements, corrosion, and missing fasteners pose serious safety risks and require timely detection and prioritization. Manual inspection techniques are time-consuming, prone to mistakes, and unsuitable for extensive monitoring. This paper suggests DeepTrackSecure, a multi-stage deep learning framework for railway track fault detection and severity-aware risk prediction, to overcome these constraints. The suggested system uses YOLOv5 for accurate fault region localization in defective images after using ResNet50 as an initial screening model to categorize railway track images as defective or non-defective. By combining defect area, detection confidence, and defect frequency, a quantitative severity score is created that goes beyond traditional detection and allows for well-informed risk assessment. Random Forest and XGBoost models are assessed for classifying track conditions into low, medium, and high-risk levels based on this severity representation, exhibiting dependable and consistent decision-making. Ablation experiments to compare YOLO-only detection, classification-detection fusion, and the full severity-based risk prediction model confirm the role of each of the system components. Also, runtime and memory analysis are used to ensure that the performance is near to what is required in edge and cloud deployment and that the robustness is tested in harsh conditions like blur, noise, and low-light conditions. The findings of the experiments indicate that DeepTrackSecure functions as a decision-support and maintenance prioritization tool that produces severity-derived risk proxies (Low\/Medium\/High) validated against expert opinion. The system is intended to assist maintenance scheduling decisions rather than serve as a directly deployable operational safety system; broader deployment should be preceded by recall improvement and validation against real failure-event ground truth.<\/jats:p>","DOI":"10.3389\/frai.2026.1800348","type":"journal-article","created":{"date-parts":[[2026,5,15]],"date-time":"2026-05-15T05:42:34Z","timestamp":1778823754000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["DeepTrackSecure: an integrated classification-detection system with predictive risk analytics for proactive railway safety management"],"prefix":"10.3389","volume":"9","author":[{"given":"P.","family":"Balakrishnan","sequence":"first","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"A. Anny","family":"Leema","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"S.","family":"Suresh Nagarajan","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"K. S.","family":"Rashmika","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"J.","family":"Rahul Kumar","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2026,5,15]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"140295","DOI":"10.1016\/j.conbuildmat.2025.140295","article-title":"A review on railway track buckling prediction methods","volume":"466","author":"Agustin","year":"2025","journal-title":"Constr. Build. Mater."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"134846","DOI":"10.1109\/ACCESS.2023.3334973","article-title":"The accelerated inference of a novel optimized YOLOv5-LITE on low-power devices for railway track damage detection","volume":"11","author":"Dang","year":"2023","journal-title":"IEEE Access"},{"key":"ref3","author":"Dong","year":"2024"},{"key":"ref4","volume-title":"Railway fastener fault detection using YOLOv5","author":"Efraimsson","year":"2022"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"482","DOI":"10.3390\/ai5020024","article-title":"ECARRNet: an efficient LSTM-based ensembled deep neural network architecture for railway fault detection","volume":"5","author":"Eunus","year":"2024","journal-title":"AI"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"19","DOI":"10.3934\/ElectrEng.2020.1.19","article-title":"Analysis and prediction of railway accident risks using machine learning","volume":"4","author":"Hadj-Mabrouk","year":"2020","journal-title":"AIMS Electron. Electr. Eng."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"1983","DOI":"10.3390\/s22051983","article-title":"Railway track inspection using deep learning based on audio to spectrogram conversion: an on-the-fly approach","volume":"22","author":"Hashmi","year":"2022","journal-title":"Sensors"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"1245","DOI":"10.1109\/access.2024.3412425","article-title":"EHA-YOLOv5: an efficient and highly accurate improved YOLOv5 model for workshop bearing rail defect detection application","volume":"12","author":"Hu","year":"2024","journal-title":"IEEE Access"},{"key":"ref10","author":"James","year":"2018"},{"key":"ref11","author":"Jiang","year":"2023"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1097","DOI":"10.1142\/S0219455422501097","article-title":"Machine learning to identify dynamic properties of railway track components","volume":"22","author":"Kaewunruen","year":"2022","journal-title":"Int. J. Struct. Stab. Dyn."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"11458","DOI":"10.17148\/IARJSET.2024.11458","article-title":"Review on various railway track fault detection systems and methods","volume":"11","author":"Karve","year":"2024","journal-title":"IARJSET."},{"key":"ref14","author":"Kaur","year":"2024"},{"key":"ref15","article-title":"A systematic literature review of defect detection in railways using machine vision-based inspection methods","author":"Kumar","year":"2024","journal-title":"Int. J. Transp. Sci. Technol."},{"key":"ref16","author":"La","year":"2022"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"7275","DOI":"10.3390\/s22197275","article-title":"Prediction models for railway track geometry degradation using machine learning methods: a review","volume":"22","author":"Liao","year":"2022","journal-title":"Sensors"},{"key":"ref18","author":"Lin","year":"2019"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"2853","DOI":"10.1109\/TPWRD.2020.3022750","article-title":"Real-time hierarchical neural network based fault detection and isolation for high-speed railway system under hybrid AC\/DC grid","volume":"35","author":"Liu","year":"2020","journal-title":"IEEE Trans. Power Deliv."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"125109","DOI":"10.1109\/ACCESS.2020.3007603","article-title":"SCueU-net: efficient damage detection method for railway rail","volume":"8","author":"Lu","year":"2020","journal-title":"IEEE Access"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"106622","DOI":"10.1016\/j.engappai.2023.106622","article-title":"TrackSafe: a comparative study of data-driven techniques for automated railway track fault detection using image datasets","volume":"125","author":"Minguell","year":"2023","journal-title":"Eng. Appl. Artif. Intell."},{"key":"ref22","article-title":"Vision based railway track monitoring using deep learning","author":"Mittal","year":"2017","journal-title":"arXiv preprint"},{"key":"ref23","author":"Mukhtiar","year":"2023"},{"key":"ref24","author":"Nandan","year":"2023"},{"key":"ref25","author":"Nigam","year":"2024"},{"key":"ref26","author":"Owais","year":"2023"},{"key":"ref27","author":"Parvathy","year":"2017"},{"key":"ref28","author":"Poornima","year":"2024"},{"key":"ref29","author":"Prakash","year":"2024"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"37678","DOI":"10.1109\/JSEN.2024.3447730","article-title":"TOSS: deep learning- based track object detection using smart sensor","volume":"24","author":"Rajeswari","year":"2024","journal-title":"IEEE Sensors J."},{"key":"ref31","author":"Rakshit","year":"2022"},{"key":"ref32","author":"Reddy","year":"2025"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"3","DOI":"10.3390\/infrastructures10010003","article-title":"Automatic detection of railway faults using neural networks: a comparative study of transfer learning models and YOLOv11","volume":"10","author":"Rodr\u00edguez-Abreo","year":"2024","journal-title":"Infrastructures"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"7018","DOI":"10.3390\/s23167018","article-title":"Railway track fault detection using selective MFCC features from acoustic data","volume":"23","author":"Rustam","year":"2023","journal-title":"Sensors"},{"key":"ref35","author":"Samantaray","year":"2024"},{"key":"ref36","author":"Sangeetha","year":"2023"},{"key":"ref37","author":"Saritas","year":"2023"},{"key":"ref38","doi-asserted-by":"publisher","first-page":"6221","DOI":"10.3390\/s21186221","article-title":"A novel approach to railway track faults detection using acoustic analysis","volume":"21","author":"Shafique","year":"2021","journal-title":"Sensors"},{"key":"ref39","author":"Shah","year":"2022"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"106520","DOI":"10.1109\/ACCESS.2022.3210326","article-title":"IoT based railway track faults detection and localization using acoustic analysis","volume":"10","author":"Siddiqui","year":"2022","journal-title":"IEEE Access"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"23883","DOI":"10.1109\/TITS.2022.3207490","article-title":"Geometric constraint and image inpainting-based railway track fastener sample generation for improving defect inspection","volume":"23","author":"Su","year":"2022","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref42","author":"Su","year":"2022"},{"key":"ref43","author":"Thinakaran","year":"2024"},{"key":"ref44","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TIM.2021.3117357","article-title":"Real-time defect detection of track components: considering class imbalance and subtle difference between classes","volume":"70","author":"Tu","year":"2021","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref1001","doi-asserted-by":"publisher","first-page":"107405","DOI":"10.1016\/j.soildyn.2022.107405","article-title":"Research on dynamic response and fatigue life of tunnel bottom structure under coupled action of train load and groundwater","volume":"161","author":"Wang","year":"2022","journal-title":"Soil. Dyn. Earthquake Eng."},{"key":"ref45","doi-asserted-by":"publisher","first-page":"805","DOI":"10.3390\/app14020805","article-title":"Automated identification and localization of rail internal defects based on object detection networks","volume":"14","author":"Wang","year":"2024","journal-title":"Appl. Sci."},{"key":"ref46","doi-asserted-by":"publisher","first-page":"61973","DOI":"10.1109\/ACCESS.2020.2984264","article-title":"Multi-target defect identification for railway track line based on image processing and improved YOLOv3 model","volume":"8","author":"Wei","year":"2020","journal-title":"IEEE Access"},{"key":"ref47","author":"Welankiwar","year":"2018"},{"key":"ref48","author":"Yang","year":"2024"},{"key":"ref49","author":"Yilmazer","year":"2023"},{"key":"ref50","author":"Yuan","year":"2019"},{"key":"ref51","doi-asserted-by":"publisher","first-page":"114061","DOI":"10.1109\/ACCESS.2023.3322444","article-title":"An improved lightweight YOLOv5 algorithm for detecting railway catenary hanging string","volume":"11","author":"Zhang","year":"2023","journal-title":"IEEE Access"},{"key":"ref52","doi-asserted-by":"publisher","first-page":"6409","DOI":"10.3390\/s22176409","article-title":"A review on rail defect detection systems based on wireless sensors","volume":"22","author":"Zhao","year":"2022","journal-title":"Sensors"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1800348\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,15]],"date-time":"2026-05-15T05:42:35Z","timestamp":1778823755000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1800348\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,15]]},"references-count":52,"alternative-id":["10.3389\/frai.2026.1800348"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1800348","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,15]]},"article-number":"1800348"}}