Skip to main content
Springer Nature Link
Log in

A Formal Model of Train Control with AI-Based Obstacle Detection

  • Conference paper
  • First Online:
Reliability, Safety, and Security of Railway Systems. Modelling, Analysis, Verification, and Certification (RSSRail 2023)

Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 14198))

Abstract

The research project KI-LOK aims to develop a certification methodology for incorporating AI components into rail vehicles. In this work, we study how to safely incorporate an AI for obstacle detection into an ATO (automatic train operation) system for shunting movements. To analyse the safety of our system we present a formal B model comprising the steering and AI perceptions subsystems as well as the shunting yard environment. Classical model checking is applied to ensure that the complete system is safe under certain assumptions. We use SimB to simulate various scenarios and estimate the likelihood of certain errors when the AI makes mistakes.

This research is part of the KI-LOK project funded by the “Bundesministerium für Wirtschaft und Energie”; grant # 19/21007E.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

Notes

  1. 1.

    https://ki-lok.itpower.de.

  2. 2.

    But we are also investigating systems with two cameras or with LiDAR sensors.

  3. 3.

    https://github.com/ultralytics/ultralytics.

  4. 4.

    https://universe.roboflow.com/kilok/sign-detection-4oqe4/dataset/2.

  5. 5.

    https://automaticaddison.com/how-to-determine-the-orientation-of-an-object-using-opencv/.

  6. 6.

    We hope to obtain such precise figures from industrial partners in our project.

  7. 7.

    For now, we define that the probability of a safe drive from 347a to 855b must be \(\ge \) 99.9%.

  8. 8.

    For now, we define that the probability of achieving the mission order must be \(\ge \) 99.9%.

  9. 9.

    The models can be found at https://github.com/hhu-stups/kilok_shunting_model/tree/14c2ecdb6e32ba593cac64e5868c94773139b391.

  10. 10.

    Version: 1.12.0-final (fef4b935b59d76e353ab67230f6206b15f903f4b, 05.04.2023).

  11. 11.

    Some of the traces can be accessed as an interactive HTML document at

    https://stups.hhu-hosting.de/models/kilok/HTML_Traces.

  12. 12.

    https://data.fid-move.de/dataset/osdar23.

References

  1. Abrial, J., Hoare, A.: The B-Book: Assigning Programs to Meanings. Cambridge University Press (2005)

    Google Scholar 

  2. Abrial, J.-R.: Modeling in Event-B: System and Software Engineering. Cambridge University Press (2010)

    Google Scholar 

  3. ClearSy. Atelier B, User and Reference Manuals. Aix-en-Provence, France (2016). https://www.atelierb.eu/

  4. Comptier, M., Déharbe, D., Perez, J.M., Mussat, L., Thibaut, P., Sabatier, D.: Safety analysis of a CBTC system: a rigorous approach with Event-B. In Proceedings RSSRail, LNCS, vol. 10598, pp. 148–159 (2017)

    Google Scholar 

  5. Comptier, M., Leuschel, M., Mejia, L., Perez, J.M., Mutz, M.: Property-based modelling and validation of a CBTC zone controller in Event-B. In: Proceedings RSSRail. LNCS, vol. 11495, pp. 202–212 (2019)

    Google Scholar 

  6. Cummings, M.L.: Rethinking the maturity of artificial intelligence in safety-critical settings. AI Mag. 42(1), 6–15 (2021)

    Google Scholar 

  7. dos Santos, C.F.G., Papa, J.P.: Avoiding overfitting: a survey on regularization methods for convolutional neural networks. CoRR, abs/2201.03299 (2022)

    Google Scholar 

  8. A. R. Fayjie, S. Hossain, D. Oualid, and D.-J. Lee. Driverless car: Autonomous driving using deep reinforcement learning in urban environment. In 2018 15th international conference on ubiquitous robots (ur), pages 896–901. IEEE, 2018

    Google Scholar 

  9. K. P. F.R.S. LIII. On lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 2(11), 559–572 (1901)

    Google Scholar 

  10. Gehr, T., Mirman, M., Drachsler-Cohen, D., Tsankov, P., Chaudhuri, S., Vechev, M.: Ai2: safety and robustness certification of neural networks with abstract interpretation. In: 2018 IEEE Symposium on Security and Privacy (SP), pp. 3–18. IEEE (2018)

    Google Scholar 

  11. Hansen, D., Leuschel, M., Schneider, D., Krings, S., Körner, P., Naulin, T., Nayeri, N., Skowron, F.: Using a formal B model at runtime in a demonstration of the ETCS hybrid level 3 concept with real trains. In: Butler, M., Raschke, A., Hoang, T.S., Reichl, K. (eds.) ABZ 2018. LNCS, vol. 10817, pp. 292–306. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-91271-4_20

    Chapter  Google Scholar 

  12. Huang, X., Kwiatkowska, M., Wang, S., Wu, M.: Safety verification of deep neural networks. In: Majumdar, R., Kunčak, V. (eds.) CAV 2017. LNCS, vol. 10426, pp. 3–29. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-63387-9_1

    Chapter  Google Scholar 

  13. Huang, X., Ruan, W., Tang, Q., Zhao, X.: Bridging formal methods and machine learning with global optimisation. In: Riesco, A., Zhang, M. (eds) ICFEM 2022. LNCS, vol. 13478, pp. 1–19. Springer, Cham (2022)

    Google Scholar 

  14. Itseez. Open source computer vision library (2015). https://github.com/itseez/opencv

  15. Jackson, D., et al.: Certified control: An architecture for verifiable safety of autonomous vehicles. CoRR, abs/2104.06178 (2021)

    Google Scholar 

  16. Katz, G., Barrett, C., Dill, D.L., Julian, K., Kochenderfer, M.J.: Reluplex: an efficient SMT solver for verifying deep neural networks. In: Proceedings CAV, LNCS, vol. 10426, pp. 97–117 (2017)

    Google Scholar 

  17. Leurent, E.: An environment for autonomous driving decision-making (2018). https://github.com/eleurent/highway-env

  18. Leuschel, M.: Operation caching and state compression for model checking of high-level models - how to have your cake and eat it. In: Proceedings iFM. LNCS, vol. 13274, pp. 129–145 (2022)

    Google Scholar 

  19. Leuschel, M., Butler, M.: ProB: a model checker for B. In: Proceedings FME, LNCS, vol. 2805, pp. 855–874 (2003)

    Google Scholar 

  20. Leuschel, M., Butler, M.J.: ProB: an automated analysis toolset for the B method. STTT 10(2), 185–203 (2008)

    Article  Google Scholar 

  21. Nonami, K., Kendoul, F., Suzuki, S., Wang, W., Nakazawa, D.: Autonomous flying robots: unmanned aerial vehicles and micro aerial vehicles. Springer Science & Business Media (2010)

    Google Scholar 

  22. Peleska, J., Haxthausen, A.E., Lecomte, T.: Standardisation considerations for autonomous train control. In: Proceedings ISoLA. LNCS, vol. 13704, pp. 286–307 (2022)

    Google Scholar 

  23. Redmon, J., Divvala, S.K., Girshick, R.B., Farhadi, A.: You Only Look Once: Unified, Real-Time Object Detection. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779–788, Los Alamitos, CA, USA, June 2016. IEEE Computer Society

    Google Scholar 

  24. Ruan, W., Huang, X., Kwiatkowska, M.: Reachability analysis of deep neural networks with provable guarantees. In: Proceedings IJCAI International Joint Conferences on Artificial Intelligence Organization, pp. 2651–2659, 7 2018

    Google Scholar 

  25. Sabatier, D.: Using formal proof and B method at system level for industrial projects. In: Lecomte, T., Pinger, R., Romanovsky, A. (eds.) RSSRail 2016. LNCS, vol. 9707, pp. 20–31. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-33951-1_2

    Chapter  Google Scholar 

  26. Sabatier, D., Burdy, L., Requet, A., Guéry, J.: Formal proofs for the NYCT line 7 (flushing) modernization project. In: Derrick, J., Fitzgerald, J., Gnesi, S., Khurshid, S., Leuschel, M., Reeves, S., Riccobene, E. (eds.) ABZ 2012. LNCS, vol. 7316, pp. 369–372. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-30885-7_34

    Chapter  Google Scholar 

  27. Seshia, S.A., Sadigh, D., Sastry, S.S.: Toward verified artificial intelligence. Commun. ACM 65(7), 46–55 (2022)

    Article  Google Scholar 

  28. Stock, S., Vu, F., Geleßus, D., Leuschel, M., Mashkoor, A., Egyed, A.: Validation by abstraction and refinement. In: Proceedings ABZ. LNCS, vol. 14010, pp. 160–178 (2023.) https://doi.org/10.1007/978-3-031-33163-3_12

  29. Sun, P., et al.: Scalability in perception for autonomous driving: Waymo open dataset. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 2446–2454 (2020)

    Google Scholar 

  30. Sun, Y., Wu, M., Ruan, W., Huang, X., Kwiatkowska, M., Kroening, D.: Concolic testing for deep neural networks. CoRR, abs/1805.00089 (2018)

    Google Scholar 

  31. Suzuki, S., Abe, K.: Topological structural analysis of digitized binary images by border following. Comput. Vis. Graph. Image Process. 30(1), 32–46 (1985)

    Article  MATH  Google Scholar 

  32. Vu, F., Leuschel, M., Mashkoor, A.: Validation of formal models by timed probabilistic simulation. In: Raschke, A., Méry, D. (eds.) ABZ 2021. LNCS, vol. 12709, pp. 81–96. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-77543-8_6

    Chapter  Google Scholar 

  33. Werth, M., Leuschel, M.: VisB: a lightweight tool to visualize formal models with SVG graphics. In: Raschke, A., Méry, D., Houdek, F. (eds.) ABZ 2020. LNCS, vol. 12071, pp. 260–265. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-48077-6_21

    Chapter  Google Scholar 

Download references

Acknowledgements

Infrastructure for model checking benchmarks was provided by the “Centre for Information and Media Technology” (ZIM) at the University of Düsseldorf (Germany). We thank anonymous reviewers for their very helpful comments and links to related work.

Author information

Authors and Affiliations

  1. Institut für Informatik, Universität Düsseldorf, Universitätsstr. 1, 40225, Düsseldorf, Germany

    Jan Gruteser, David Geleßus, Michael Leuschel, Jan Roßbach & Fabian Vu

Authors
  1. Jan Gruteser
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. David Geleßus
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Michael Leuschel
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jan Roßbach
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Fabian Vu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Michael Leuschel .

Editor information

Editors and Affiliations

  1. Technische Universität Berlin, Berlin, Germany

    Birgit Milius

  2. Université Gustave Eiffel, Villeneuve d’Ascq, France

    Simon Collart-Dutilleul

  3. CLEARSY Systems Engineering, Aix en Provence, France

    Thierry Lecomte

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Gruteser, J., Geleßus, D., Leuschel, M., Roßbach, J., Vu, F. (2023). A Formal Model of Train Control with AI-Based Obstacle Detection. In: Milius, B., Collart-Dutilleul, S., Lecomte, T. (eds) Reliability, Safety, and Security of Railway Systems. Modelling, Analysis, Verification, and Certification. RSSRail 2023. Lecture Notes in Computer Science, vol 14198. Springer, Cham. https://doi.org/10.1007/978-3-031-43366-5_8

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-43366-5_8

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-43365-8

  • Online ISBN: 978-3-031-43366-5

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics