Decision support for preventing safety violations

Aim. The aim of the paper is to examine the experience of reducing the effect of the human factor on business processes, to develop the structure and software of the decisionsupport system for preventing safety violations by train drivers using machine learning and to analyse the findings.
Methods. The study presented in the paper uses machine learning, statistical analysis and expert analysis. In terms of machine learning, the following methods were used: logistical regression, random forests, gradient boosting over decision trees with frequency-domain representation of categorical features, neural networks.
Results. A set of indicators characterizing a train driver’s operation were identified and are to be used as part of the system under development. The term “train driver’s reliability” was defined as the ability not to violate train traffic safety over a certain number of trips. Algorithms were designed and examined for predicting violations in a train driver’s operation that are used in defining reliability groups and lists of preventive measures recommended for the reduction of the number of safety violations in a train driver’s operation. Major violations with proven guilt of the driver that may be committed within the following 3, 7, 10, 20, 30, 60 days were chosen as attributes for the purpose of safety violation prediction. Analysis of the results on the test sample revealed that the model based on gradient boosting over decision trees with frequency-domain representation of categorical features shows the best results for binary classification on the prediction horizon of 30 and 60 days. The developed algorithm made a correct prediction in 76% of cases with the threshold value of 0.7 and horizon of 30 days and in 82% of cases with the threshold value of 0.9 and horizon of 60 days. The solution of the problem can be found in the integration of different approaches to predicting safety violations in a train driver’s operation. Additionally, 10 of the most significant indicators of a train driver’s operation were identified with the best of the considered models, i.e., gradient boosting over decision trees with frequency-domain representation of categorical features.
Conclusion. The paper presents an overview of methods and systems of assessing human reliability and the effect of the human factor on the safety of transportation systems. It allowed choosing the most promising directions and methods of predictive analysis of a train driver’s operation, including methods of machine learning. The resulting set of indicators of a train driver’s operation that take into consideration the changes in the quality of such operation allowed obtaining initial data for training the models implemented as part of the system under development. The implemented models enabled the aggregation of information on train drivers and adoption of targeted and temporary preventive measures recommended for improving driver reliability. The resulting approach to the definition of preventive measures has been implemented in three depots of JSC RZD in trial operation mode.

1. Golitsyn A.P., Maslov A.A., Ruchkin D.A. (RU). Certificate 2019612885. [Development of the system for accounting and analysing violations of train movement safety based on the results of automatic decoding of onboard recorder units (ASUT NBD-2)]: Computer program. Rightsholder: Joint Stock Company Russian Railways. No. 2019612885; claimed 22.02.19; published 04.03.2019; 300 MB. (in Russ.)

2. Shubinsky I.B. [Structural dependability of information systems. Analysis methods]. Moscow: Dependability Journal; 2012. (in Russ.)

3. Shubinsky I.B. [Functional dependability of information systems. Analysis methods]. Moscow: Dependability Journal; 2012. (in Russ.).

4. Swain A.D. Human reliability analysis: Need, status, trends and limitations. Reliability Engineering & System Safety 1990;29(3):301-313.

5. Swain A. D., Guttmann H.E. Handbook of humanreliability analysis with emphasis on nuclear power plant applications. Sandia National Labs, Final report no. NUREG/ CR--1278; 1983.

6. Corlett E.N., Wilson J.R. Evaluation of human work. CRC Press; 1995.

7. Forester J.A., Ramey-SmithA., Bley D.C. Discussion of comments from a peer review of a technique for human event analysis (ATHEANA). Sandia National Laboratories; 1998.

8. Holmberg J.E., Bladh K., Oxstrand J. The Application of the Enhanced Bayesian THERP in the HRA Methods Empirical Study Using Simulator Data. Proceedings of PSAM; 2008.

9. Cooper S.E., Ramey-Smith A.M., Wreathall J. A technique for human error analysis. USNRC ed. Washington: DC: NUREG/CR-6350; 1996.

10. Hidayatulloh A. Dampak adaptasi presentasi treeview terhadap niat untuk melakukan pembelian secara online: emosi dan sikap pengguna sebagai mediator (didasarkan pada stimulus-organism-response model). Optimum: Jurnal Ekonomi dan Pembangunan 2015;5(2):147-156.

11. Gorelik A.V., Taradin N.A., Zhuravlev I.A. [Methods of functional safety analysis of railway signalling systems]. Dependability 2011;1:40-46. (in Russ.)

12. Lisenkov V.M. [Safety and efficiency of transportation processes]. Railway Economics 2008;4:33-42.

13. Popov Yu.I., Roizner A.G., Zelikman B.L., Pevzner M.A., Yarkovsky F.V. Patent 133960 Russian Federation. [Mobile training and demonstration system of railway transportation safety devices]. Applicant: Joint Stock Company Russian Railways. No. 2013125124/11; claimed 30.05.13; published 27.10.13. (in Russ.)

14. Kuchumov V.A., Nikiforova N.B., Murzin R.V. et al. Forecasting methods of electricity consumption for traction of trains. Science and Technology in Transport 2015;3:104-110.

15. Telpov B.V., Borisenkov S.S. [Comprehensive automated passenger train operation system]. Zheleznodorozhny transport 2011;3:48-52. (in Russ.)

16. Kolmakov V.O., Zubkov V.V., Novikov A.V. [SAUT automatic brake control system]. Innovatsii. Nauka. Obrazovanie 2020;22:545-549. (in Russ.)

17. Zorin V.I., Perevozchikov S.A., Rychkov A.S. Patent 2420418 Russian Federation. [Integrated on-board train protection system]. Applicant: Izhevskiy Radiozavod AO. No. 2007145632/11; claimed 11.12.07; published 10.06.11. (in Russ.)

18. Bugaev A.S., Gerus S.V., Dementienko V.V. et al. [Remote Driver Vigilance Supervision System]. Bulleten obiedinennogo uchionogo soveta OAO “RZhD” 2017;2:21- 41. (in Russ.)

19. Shikher Ya.G., Boveh Yu.E., Meerzon Yu.M., Oreshkin E.V., Shakhnarovich V.M. Certificate of authorship 990573 Russian Federation. [Driver vigilance supervision device]. Applicant: Design Bureau of the Main Locomotive Directorate of the Ministry of Railways of the USSR. No. 3337528; claimed 11.08.81; published 23.01.83. (in Russ.)

20. Bodrov V.A., Orlov V.Ya. [Psychology and dependability: human being in control systems]. Moscow: Institute of Psychology of the RAS; 1998. (in Russ.)

21. Apattsev V.I., Zavyalov A.M., Sinyakina I.N. et al. Safety of train operation on the basis of decrease in influence of human factor. Science and Technology in Transport 2014;2:75-78.

22. Voronkova E.A., Medvedeva V.M. Assessment of the professional risks of machinists and assistants of railway and construction machine operators. Security problems of the Russian society 2019;4:42-48. (in Russ.)

23. Baranov L.A., Sidorenko V.G., Balakina E.P. et al. Intelligent centralized traffic management of a rapid transit system under heavy traffic. Dependability 2021;21(2):17-23. DOI: 10.21683/1729-2646-2021-21-2-17-23.

24. Kharin O.V., Yakimov S.M., Kulagin M.A. et all. (RU). Certificate 2020613754. [Automated System Trusted Environment of the Locomotive Service]: Computer program. Rightsholder: Joint Stock Company Russian Railways. No. 2020613754; claimed. 11.03.2020; published 23.03.2020; 490 Kb. (in Russ.)

25. Sidorenko V.G., Kulagin M.A. The approach to the formation of a driver’s rating using different comparison metrics. Electronics and electrical equipment of transport 2018;1:14-17. (in Russ.)

26. Kulagin M.A., Sidorenko V.G. Qualification of drivers as a factor of increasing reliability of electric rolling stock. Electronics and electrical equipment of transport 2018;4:70-76. (in Russ.)

27. Dorogush A.V., Ershov V., Gulin A. CatBoost: gradient boosting with categorical features support. arXiv preprint arXiv:1810.11363; 2018.

28. Kulagin M., Sidorenko V. A Recommender Subsystem Construction for Calculating the Probability of a Violation by a Locomotive Driver using Machine-learning Algorithms. IEEE East-West Design & Test Symposium (EWDTS); 2020. p. 1-5.