This study provides a non-exhaustive overview of AI methods, algorithms, and applications, particularly machine learning (ML) in rail transport. Drawing inspiration from the two European directives relating to the development (Directive 2012/34/EU) and interoperability (Directive (EU) 2016/797) of the railway system, this review proposes to classify AI applications according to (1) the “structural elements” of the railway system (infrastructure, rolling stock, energy, control and command, and signaling), (2) the “functional elements” (operation, maintenance, and telematics applications), and (3) both “structural and functional” elements with a view to identifying AI approaches aimed at improving railway safety, in particular the analysis of railway accidents and incidents based on investigation reports. Several “classic” AI techniques are implemented, including ML (supervised, semi-supervised, unsupervised), deep learning such as artificial neural networks (ANN), natural language processing (NLP), case-based reasoning (CBR), etc. However, the lack of interoperability between “classic” AI tools and the inadequacy of these approaches to capitalize, share, and reuse this knowledge have driven research toward the development of new approaches based on ontology’s and knowledge graphs. A study of all AI applications shows that the stages of acquisition, retrieval, analysis, structuring, formalization, modeling, processing, and interpretation of the data produced pose a crucial problem in the rail transport sector. Moreover, with complex models called “black boxes,” it is difficult to understand how and why the internal reasoning mechanisms of the AI system impact the solution and predictions. The new approach of explainable AI (XAI) or explainable machine learning (XML) can provide an element of response to this problem, particularly when it comes to a crucial issue such as railway safety.
ZantalisFKoulourasGKarabetsosS, et al. A review of Machine Learning and IoT in smart transportation. Future Internet2019; 11(4): 94.
2.
IyerLS.AI enabled applications towards intelligent transportation. Transp Eng2021; 5: 100083.
3.
JevingerÅZhaoCPerssonJA, et al. Artificial intelligence for improving public transport: a mapping study. Public Transp2024; 16: 99–158.
4.
McMillanLVargaL.A review of the use of artificial intelligence methods in infrastructure systems. Eng Appl Artif Intell2022; 116: 105472.
5.
TselentisDIPapadimitriouEvan GelderP.The usefulness of artificial intelligence for safety assessment of different transport modes. Accid Anal Prev2023; 186: 107034.
6.
Attoh-OkineN.Big data challenges in railway engineering. In: IEEE international conference on big data (Big Data), Washington, DC, USA, October 2014, pp.27–30. DOI: 10.1109/BigData.2014.7004424.
7.
ThaduriAGalarDKumarU.Railway assets: a potential domain for big data analytics. Procedia Informatique2015; 53: 457–467.
8.
van GulijkCHughesPFigueres-EstebanM, et al. The case for IT transformation and big data for safety risk management on the GB railways. Proc Inst Mech Eng O J Risk Reliab2018; 232(2): 151–163.
9.
GhofraniFHeQGoverdeRMP, et al. Recent applications of big data analytics in railway transportation systems: a survey. Transp Res Part C: Emerg Technol2018; 90: 226–246.
10.
Laiton-BonadiezCBranch-BedoyaJWZapata-CortesJ, et al. Industry 4.0 technologies applied to the rail transportation industry: a systematic review. Sens2022; 22(7): 2491.
11.
DongKRomanovIMcLellanC, et al. Recent text-based research and applications in railways: a critical review and future trends. Eng Appl Artif Intell2022; 116: 105435.
12.
BesinovicNDe DonatoLFlamminiF, et al. Artificial Intelligence in railway transport: taxonomy, regulations, and applications. IEEE Trans Intell Transp Syst2022; 23(9): 14011–14024.
13.
TangRDe DonatoLBešinovićN, et al. A literature review of artificial intelligence applications in railway systems. Transp Res Part C: Emerg Technol2022; 140: 103679.
14.
Directive (EU) 2016/797. Of the European Parliament and of the Council of 11 May 2016 on the interoperability of the rail system within the European Union (recast), http://data.europa.eu/eli/dir/2016/797/oj (2016, accessed 8 July 2024).
ZhaoHChenHDongW, et al. Fault diagnosis of rail turnout system based on case-based reasoning with compound distance methods. In: 29th Chinese control and decision conference (CCDC), Chongqing, China, 28–30 May 2017, pp.4205–4210. DOI: 10.1109/CCDC.2017.7979237.
17.
SanturYKaraköseMKinE. A new rail inspection method based on deep learning using laser cameras. In: International artificial intelligence and data processing symposium (IDAP), Malatya, Turkey, 16–17 September 2017, pp.1-6. DOI: 10.1109/IDAP.2017.8090245.
18.
Faghih-RoohiSHajizadehSNúñezA, et al. Deep convolutional neural networks for detection of rail surface defects. In: International joint conference on neural networks (IJCNN), Vancouver, BC, Canada, 2016, pp.2584–2589. DOI: 10.1109/IJCNN.2016.7727522.
19.
WuYZhuX.Rail defect detection using ultrasonic A-Scan data and deep autoencoder. Transp Res Rec2023; 2677(7): 62–73.
20.
SalehARemenyte-PrescottRPrescottD, et al. Intelligent and adaptive asset management model for railway sections using the iPN method. Reliab Eng Syst Saf2024; 241: 109687.
21.
WangXBaiYLiuX.Prediction of railroad track geometry change using a hybrid CNN-LSTM spatial-temporal model. Adv Eng Inform2023; 58: 102235.
22.
GibertXPatelVMChellappaR.Deep multitask learning for railway track inspection. IEEE Trans Intell Transp Syst2017; 18(1): 153–164.
23.
BergmeirCSáinzGMartínez BertrandC, et al. A study on the use of machine learning methods for incidence prediction in high-speed train tracks. In: AliMBosseTHindriksKV, et al. (eds) Recent trends in Applied Artificial Intelligence. IEA/AIE 2013. Lecture Notes in Computer Science. Vol. 7906. Springer, 2013, pp. 674–683.
24.
HongWCliftonGNelsonJD.A data-driven conceptual framework for understanding the nature of hazards in railway accidents, Media Info, Issue Number: ITLS-WP-23-12, 2023. https://hdl.handle.net/2123/31324
25.
TangYWangYQianY.Railroad crossing surveillance and foreground extraction network: weakly supervised artificial-intelligence approach. Transp Res Rec2023; 2677(9): 525–538.
26.
ZamanAHuangZLiW, et al. Artificial Intelligence-aided grade crossing safety violation detection methodology and a case study in New Jersey. Transp Res Rec2023; 2677(10): 688–706.
27.
LinPZhangLTiongRLK. Multi-objective robust optimization for enhanced safety in large-diameter tunnel construction with interactive and explainable AI. Reliab Eng Syst Saf2023; 234: 109172.
28.
BischofSSchennerG.Rail topology ontology: a rail infrastructure base ontology. In: HothoA (ed.) The semantic web – ISWC 2021. ISWC 2021. Lecture Notes in Computer Science. Vol. 12922. Springer, 2021, pp. 597–612.
29.
BragaJAPAndradeAR. Data-driven decision support system for degrading assets and its application under the perspective of a railway component. Transp Eng2023; 12: 100180.
30.
SkarlatosDKarakasisKTrochidisA.Railway wheel fault diagnosis using a fuzzy-logic method. Appl Acoust2004; 65(10): 951–966.
31.
WanTHTsangCWHuiK, et al. Anomaly detection of train wheels utilizing short-time Fourier transform and unsupervised learning algorithms. Eng Appl Artif Intell2023; 122: 106037.
32.
BezuidenhoutMJoosteJLLuckeD, et al. Leveraging digitilisation and machine learning for improved railway operations and maintenance. Procedia CIRP2023; 120: 702–707.
33.
YinJZhaoW.Fault diagnosis network design for vehicle on-board equipment’s of high-speed railway: a deep learning approach. Eng Appl Artif Intell2016; 56: 250–259.
34.
VarmaARoddyN.ICARUS: design and deployment of a case-based reasoning system for locomotive diagnostics. Eng Appl Artif Intell1999; 12(6): 681–690.
35.
BernalEWuQSpiryaginM, et al. Augmented digital twin for railway systems. Veh Syst Dyn2024; 62(1): 67–83.
36.
LotfiABagheriMAhmadiA.Using machine learning methods for modeling freight train derailment severity. Transp Res Rec2023; 2677(3): 961–973.
37.
EjlaliMArianETaghiyehS, et al. Developing hybrid machine learning models to assign health score to railcar fleets for optimal decision making. Expert Syst Appl2024; 250: 123931.
38.
VerstichelSOngenaeFLoeveL, et al. Efficient data integration in the railway domain through an ontology-based methodology. Transp Res Part C: Emerg Technol2011; 19(4): 617–643.
39.
TutcherJEastonJMRobertsC.Enabling data integration in the rail industry using RDF and OWL: the RaCoOn ontology. ASCE-ASME J Risk Uncert Eng Syst Part A: Civil Eng2017; 3(2): 401–501. DOI: 10.1061/ajrua6.0000859
40.
Rojas MelendezJAAguadoMVasilopoulouP, et al. Leveraging semantic technologies for digital interoperability in the European railway domain. In: HothoA (ed.) The Semantic Web – ISWC 2021. ISWC 2021. Lecture Notes in Computer Science. Vol. 12922. Springer, 2021; pp. 648–664.
BelliniPNesiPZazaI. RAISO: railway infrastructures and signaling ontology for configuration management, verification and validation. In: 2016 IEEE tenth international conference on semantic computing (ICSC), Laguna Hills, CA, USA, 2016, pp.350–353. DOI: 10.1109/ICSC.2016.94.
44.
FayA.A fuzzy knowledge-based system for railway traffic control. Eng Appl Artif Intell2000; 13(6): 719–729.
45.
HoinaruCGransartMGLemaireE.An ontology for ERTMS/ETCS”, in Transport Research Arena, Paris, 2014. DOI: 10.1016/j.eswa.2012.12.028.
46.
SangoMOlimpiaHGransartC, et al. A temporal QoS ontology for ERTMS/ETCS, World Academy of Science, Engineering and Technology. Int J Ind Manuf Eng2015; 9(1): 1–7.
47.
De SouzaVDMBorgesAPSatoDMV, et al. Automatic knowledge learning using case-based reasoning: a case study approach to automatic train conduction. In: International joint conference on neural networks (IJCNN), Vancouver, BC, Canada, 2016, pp.4579–4585. DOI: 10.1109/IJCNN.2016.7727800.
48.
JiaoYChenXSunZ, et al. Data-driven detection and assessment for urban railway transit driver fatigue in real work conditions. Transp Res Rec2023; 2677(1): 1367–1375.
49.
LiJLinBWangZ, et al. A pragmatic optimization method for motor train set assignment and maintenance scheduling problem. Discrete Dyn Nat Soc2016; 2016(1): 1–13.
50.
ConsilvioAVignolaGLópez ArévaloP, et al. A data-driven prioritisation framework to mitigate maintenance impact on passengers during metro line operation. Eur Transp Res Rev2024; 16: 6.
51.
LewisRFuchsFPirker RobertsM, et al. Using ontology to integrate railway condition monitoring data. In: 2006 IET international conference on railway condition monitoring, Birmingham, UK, 2006, pp.149–155.
52.
UmiliacchiPLaneDRomanoF, et al. Predictive maintenance of railway subsystems using an Ontology based modelling approach. In: Proceedings of 9th world conference on railway research, May 2011, pp.22–26.
53.
RahmigCDebbechSErrandoneaI, et al. Ontology-based conceptual model development for the railway domain: a maintenance case study. In: 4th smart racon workshop, San Sebastian, Spain, 2022. https://elib.dlr.de/189328/
54.
LouadahHPapadakisEMcCluskeyTL, et al. Supporting the management of rolling stock maintenance with an ontology-based Virtual Depot. Appl Sci2024; 14(3): 1220.
55.
de OliveiraKMBachaFMnasserH, et al. Transportation ontology definition and application for the content personalization of user interfaces. Expert Syst Appl2013; 40(8): 3145–3159.
56.
BenvenutiFDiamantiniCPotenaD, et al. An ontology-based framework to support performance monitoring in public transport systems. Transp Res Part C: Emerg Technol2017; 81: 188–208.
57.
NabianMAAlemazkoorNMeidaniH.Predicting near-term train schedule performance and delay using bi-level random forests. Transp Res Rec2019; 2673(5): 564–573.
58.
KlumpenhouwerWShalabyA.Using delay logs and machine learning to support passenger railway operations. Transp Res Rec2022; 2676(9): 134–147.
59.
ChenXMaZSunW.Incident delay prediction in urban railway systems: methodology review and exploratory comparative analysis. Transp Res Rec2024; 2678(12): 1629–1638.
60.
ShynkarenkoVZhuchyiL. Ontological harmonization of railway transport information systems. In: CEUR workshop proceedings. 2021. Vol. 2870: 5th international conference on computational linguistics and intelligent systems. Vol. I: main conference (COLINS 2021), Lviv, Ukraine, 22–23 April 2021, pp.541–554.
61.
BouhanaAZidiAFekihA, et al. An ontology-based CBR approach for personalized itinerary search systems for sustainable urban freight transport. Expert Syst Appl2015; 42(7): 3724–3741.
62.
Regulation (EU). 2019/777. Commission Implementing Regulation (EU) 2019/777 of 16 May 2019 on the common specifications for the register of railway infrastructure and repealing Implementing Decision 2014/880/EU, http://data.europa.eu/eli/reg_impl/2019/777/oj (accessed 8 July 2024).
63.
Projet européen InteGRail (Intelligent Integration of Railway Systems), European research project, [online], https://cordis.europa.eu/project/id/12526 (2024, accessed 8 July 2024).
64.
KatsumiMFoxM.Ontologies for transportation research: a survey. Transp Res Part C: Emerg Technol2018; 89: 53–82.
65.
YangLCormicanKYuM.Ontology-based systems engineering: a state-of-the-art review. Comput Ind2019; 111: 148–171.
66.
TiddiISchlobachS.Knowledge graphs as tools for explainable machine learning: a survey. Artif Intell2022; 302: 103627.
67.
Hadj-MabroukH. Literature review on applications of ontologies and knowledge graphs in railway transport safety. In: Book “Railway Transport and Engineering”. IntechOpen, 2024. https://www.intechopen.com/chapters/1203650
68.
CuiYTangZDaiH. Case-based reasoning and rule-based reasoning for railway incidents prevention. In: Proceedings of ICSSSM '05. International Conference on Services Systems and Services Management, Chongquing, China, Vol. 2, June 2005, pp.1057–1060. DOI: 10.1109/ICSSSM.2005.1500156.
69.
MirabadiASharifianS.Application of association rules in Iranian Railways (RAI) accident data analysis. Saf Sci2010; 48(10): 1427–1435.
70.
BrownDE.Text mining the contributors to rail accidents. IEEE Trans Intell Transp Syst2016; 17(2): 346–355.
71.
LiJWangJXuN, et al. Importance degree research of safety risk management processes of urban rail transit based on text mining method. Journal Information2018; 9(2): 26.
72.
LiuJSchmidFLiK, et al. A knowledge graph-based approach for exploring railway operational accidents. Reliab Eng Syst Saf2021; 207: 107352.
73.
LiuJChenKDuanH, et al. A knowledge graph-based hazard prediction approach for preventing railway operational accidents. Reliab Eng Syst Saf2024; 247: 110126.
74.
LaiJZhuJGuoY, et al. Dynamic data-driven railway bridge construction knowledge graph update method. Trans GIS2023; 27: 2099–2117.
75.
RagalaZRetbiABennaniS. Construction and Application of Knowledge Graph for railway rolling stock failure tree. In: 2022 IEEE international conference on technology management, operations and decisions (ICTMOD), Marrakech, Morocco, 2022, pp.1–6. DOI: 10.1109/ICTMOD55867.2022.10041881.
76.
YanHMaXChenF, et al. Knowledge modeling and analysis for railway fire accident using ontology-based knowledge graph. In: Proceedings of the 5th international conference on electrical engineering and information technologies for rail transportation (EITRT) 2021. EITRT 2021. Lecture notes in electrical engineering (eds LiangJJiaLQinY, et al.), Springer, Singapore, vol. 867, 2022. 10.1007/978-981-16-9909-2_59
77.
WangNYangXChenJ, et al. Hazards correlation analysis of railway accidents: a real-world case study based on the decade-long UK railway accident data. Saf Sci2023; 166: 106238.
78.
ZhaoRMaXDongH, et al. Data-driven ontology construction method for railway derailment accidents. In: Proceedings of the 5th international conference on electrical engineering and information technologies for rail transportation (EITRT) 2021. EITRT 2021. Lecture notes in electrical engineering (eds QinYJiaLLiangJ, et al.), Springer, Singapore, vol. 868, 2022. DOI: 10.1007/978-981-16-9913-9_59.
79.
KaeeniSKhalilianMMohammadzadehJ.Derailment accident risk assessment based on ensemble classification method. Saf Sci2018; 110(Part B): 3–10.
80.
HeidarysafaMKowsariKBarnesL, et al. Analysis of railway accidents’ narratives using deep learning. In: International conference on machine learning and applications ((IEEE ICMLA), Orlando, FL, USA, 2018, pp.1446–1453. DOI: 10.1109/ICMLA.2018.00235.
WilliamsTBetakJ.A comparison of LSA and LDA for the analysis of railroad accident text. Procedia Comput Sci2018; 130: 98–102.
83.
AwadFAGrahamDJSinghR, et al. Predicting urban rail transit safety via artificial neural networks. Saf Sci2023; 167: 106282.
84.
ChenYSongBZengY, et al. A deep learning-based approach for fault diagnosis of current-carrying ring in catenary system. Neural Comput Appl2023; 35: 23725–23737.
85.
LuYLiQXiaoW.Case-based reasoning for automated safety risk analysis on subway operation: case representation and retrieval. Saf Sci2013; 57: 75–81.
86.
DengYZhangYYuanZ, et al. Analyzing subway operation accidents causations: apriori algorithm and network approaches. Int J Environ Res Public Health2023; 20(4): 3386.
87.
Hadj-MabroukH.Contribution of artificial intelligence to risk assessment of railway accidents. Urban Rail Transit2019; 5: 104–122.
88.
Hadj-MabroukH., et al. Contribution of machine learning to rail transport safety. chapter 10. In: VasantP (ed.) Advances of machine learning in clean energy and the transportation industry. Nova Science Publishers, 2021, pp.277–312.
89.
Hadj-MabroukH. Approach to assist in the discovery of railway accident scenarios based on supervised learning. In: SharmaSKUpadhyayRKKumarV, et al. (eds) Transportation energy and dynamics. Energy, environment, and sustainability. Springer, 2023, pp. 129–156.
90.
Hadj-MabroukH.Application of case-based reasoning to the safety assessment of critical software used in rail transport. Saf Sci2020; 131: 104928.
91.
QiJWangJ.Bridging artificial intelligence and railway cybersecurity: a comprehensive anomaly detection review. Transp Res Rec2025; 2679(5): 232–255.
92.
PardoCPinoFJGarcíaF, et al. An ontology for the harmonization of multiple standards and models. Comput Stand Interfaces2012; 34(1): 48–59.
93.
AnsaldiSMontiMAgnelloP, et al. An ontology for the identification of the most appropriate risk management methodology. In: HerreroPPanettoHMeersmanR, et al. (eds) On the move to meaningful Internet systems: OTM 2012 workshops. OTM 2012. Lecture Notes in Computer Science. Vol. 7567. Springer, 2012, pp. 444–453.
94.
Henderson-SellersBGonzalez-PerezCMcBrideT, et al. An ontology for ISO software engineering standards: 1) Creating the infrastructure. Comput Stand Interfaces2014; 36(3): 563–576.
95.
Gonzalez-PerezCHenderson-SellersBMcBrideT, et al. An ontology for ISO software engineering standards: 2) proof of concept and application. Comput Stand Interfaces2016; 48: 112–123.
96.
Chenariyan-NakhaeeMCHiemstraDStoelingaM, et al. The recent applications of machine learning in rail track maintenance: a survey. In: Collart-DutilleulSLecomteTRomanovskyA (eds) Reliability, safety, and security of railway systems. Modelling, analysis, verification, and certification. rssrail 2019. Lecture Notes in Computer Science. Vol. 11495. Springer, pp. 91–105.
97.
Hadj-MabroukH.A literature review on the applications of artificial intelligence to European rail transport safety. IET Intell Transp Syst2024; 18(12): 2291–2324.
98.
CoorayS.The subjectivity of data scientists in machine learning design. J Comput Inf Syst2024; 64: 665–682.
99.
NiuYFanYJuX.Critical review on data-driven approaches for learning from accidents: comparative analysis and future research. Saf Sci2024; 171: 106381.
100.
TamascelliNCampariAParhizkarT, et al. Artificial Intelligence for safety and reliability: a descriptive, bibliometric and interpretative review on machine learning. J Loss Prev Process Ind2024; 90: 2024–105343.
101.
RichardsonS.Exposing the many biases in machine learning. Bus Inf Rev2022; 39(3): 82–89.
102.
XuYKohtzSBoakyeJ, et al. Physics-informed machine learning for reliability and systems safety applications: State of the art and challenges. Reliab Eng Syst Saf2023; 230: 108900.
RohlfingKJCimianoPScharlauI, et al. Explanation as a social practice: toward a conceptual framework for the social design of AI systems. IEEE Trans Cogn Dev Syst2021; 13(3): 717–728.
105.
LongoLBrcicMCabitzaF, et al. Explainable Artificial Intelligence (XAI) 2.0: a manifesto of open challenges and interdisciplinary research directions. Inf Fusion2024; 106: 102301.
106.
ThekdiSAvenT.Understanding explainability and interpretability for risk science applications. Saf Sci2024; 176: 106566.
