Predicting Water Distribution Pipe Failures Using Machine Learning and Cross-Infrastructure Data
Napoved okvar vodovodnih cevi s strojnim učenjem in podatki o sosednji infrastrukturi
- Avtorji: Daniel Kozelj, David Abert Fernández
- Citat: Acta hydrotechnica, vol. 38, no. 68, pp. 53-64, 2025. https://doi.org/10.15292/acta.hydro.2025.05
- Povzetek: Okvare vodovodnih cevi v urbanih omrežjih so pomemben vzrok komercialnih izgub zaradi neobračunane vode, motenj v oskrbi in visokih stroškov vzdrževanja. Ta študija razvija model strojnega učenja za napoved verjetnosti okvar cevovodov in podporo strategijam vzdrževanja, temelječim na tveganju. Model, izurjen na podatkih o infrastrukturi in geoprostorskih podatkih iz obdobja 2010–2025, vključuje standardne lastnosti cevi – kot so material, starost, premer, vrsta omrežja in zgodovina vzdrževanja – ter prostorsko izpeljane kazalnike infrastrukture v neposredni soseščini. Posebej pomembno je, da model kvantificira napovedno vrednost sosednjih infrastrukturnih sistemov, vključno z električnim omrežjem, plinovodi, daljinskim ogrevanjem, kanalizacijo in cestnim omrežjem, z uporabo prostorske analize soseščine in tehnik prekrivanja. Več teh medinfrastrukturnih značilnosti, zlasti kategorija ceste, napetost električnega omrežja in tip kanalizacije, je pokazalo pomemben napovedni vpliv, kar odraža njihovo posredno, a dosledno povezavo z verjetnostjo okvare cevi. Model strojnega učenja, zasnovan z algoritmom XGBoost in validiran s slojevitim navzkrižnim preverjanjem (K-fold), je dosegel visoko zmogljivost (ROC AUC: 0,9102; priklic: 0,7750; natančnost: 0,8750). Kljub nižji preciznosti zaradi neravnovesja razredov rezultat F1 (0,2261) in LogLoss (0,2500) potrjujeta njegovo zanesljivost. Raziskava predstavlja nov, s prostorskimi podatki obogaten pristop k napovedovanju okvar in prispeva k naprednemu, na podatkih temelječem upravljanju urbane infrastrukture.
- Ključne besede: Vodovodni sistemi, napovedovanje okvar cevovodov, strojno učenje, XGBoost, prostorska analiza, ocenjevanje stanja.
- Polno besedilo: a38dk.pdf
- Viri:
- Akiba, T., Sano, S., Yanase, T., Ohta, T., Koyama, M. (2019). Optuna: A Next-generation Hyperparameter Optimization Framework. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Association for Computing Machinery, New York, NY, United States, Pages 2623–2631. https://doi.org/10.1145/3292500.3330701.
- Asadi, Y. (2024). Employing machine learning in water infrastructure management: predicting pipeline failures for improved maintenance and sustainable operations. Industrial Artificial Intelligence 2(8), 1-13. https://doi.org/10.1007/s44244-024-00022-w.
- Bakhtawar, B., Zayed, T., Elshaboury, N. (2025). Time-to-failure based deterioration factors of water networks: Systematic review and prioritization. Reliability Engineering & System Safety, 263, 111246. https://doi.org/10.1016/j.ress.2025.111246.
- Cabral, M., Gray, D., Brentan, B., Covas, D. (2024). Assessing Pipe Condition in Water Distribution Networks. Water, 16(10), 1318. https://doi.org/10.3390/w16101318.
- Chen, T., Guestrin, C. (2016). "XGBoost: A scalable tree boosting system." Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 785-794. https://doi.org/10.1145/2939672.2939785.
- Chen, T., et al. (2023). XGBoost documentation. XGBoost. https://xgboost.readthedocs.io/.
- European Commission. (2025). European Water Resilience Strategy (COM(2025) 280 final). Brussels: European Commission. https://environment.ec.europa.eu/publications/european-water-resilience-strategy_en.
- Ganjidoost, A., Haghighi, A., Klise, K. A. (2022). Pipe failure prediction in water distribution networks using machine learning models. Journal of Water Resources Planning and Management, 148(5), 04022017. https://doi.org/10.1061/(ASCE)WR.1943-5452.0001557.
- GURS. (2025). Data on public infrastructure = Zbirni kataster gospodarske javne infrastrukture. Republic of Slovenia, Surveying and Mapping Authority of the Republic of Slovenia. Slovenia. https://www.e-prostor.gov.si/.
- MOPE. (2025). Information System of Public Environmental Protection Services = Informacijski sistem za spremljanje gospodarskih javnih služb varstva okolja (IJSVO). Republic of Slovenia, Ministry of Natural Resources and Spatial Planning, Slovenia. (in Slovenian) https://www.ijsvo.si/.
- Jafar, R., Shahrour, I., Juran, I. (2010). Application of artificial neural networks (ANN) to model the failure of urban water mains. Mathematical and Computer Modelling, 51(9–10), 1170–1180. https://doi.org/10.1016/j.mcm.2009.12.033.
- Karadirek, I. E., Kaya-Basar, E., Akdeniz, T. (2024). A study on pipe failure analysis in water distribution systems using logistic regression. Water Supply, 24(1), 176–186. https://doi.org/10.2166/ws.2023.335.
- Kleiner, Y., Rajani, B. (2001). Comprehensive review of structural deterioration of water mains: Statistical models. Urban Water, 3(3), 131–150. https://doi.org/10.1016/S1462-0758(01)00033-4.
- Kozelj, D., Gorjup, M., Kramar Fijavž, M. (2017). Uporaba teorije grafov za zasnovo merilnih območij v vodovodnem omrežju = An application of spectral graph partition for designing district metered areas in water supply networks. Acta hydrotechnica, 30(53), 81–96. https://www.dlib.si/details/URN:NBN:SI:doc-AEEWAAH7.
- Latifi, M., Zali, R.B., Javadi, A.A., Farmani, R. (2024). Efficacy of Tree-Based Models for Pipe Failure Prediction and Condition Assessment: A Comprehensive Review. Journal of Water Resources Planning and Management, 150(7), 03124001. https://doi.org/10.1061/JWRMD5.WRENG-6334.
- Large, A., Le Gat, Y., Elachachi, S. M., Renaud, E., Breysse, D., Tomasian, M. (2015). Improved modelling of ‘long-term’ future performance of drinking water pipes. Journal of Water Supply: Research and Technology—AQUA, 64(4), 415–425. https://doi.org/10.2166/aqua.2015.115.
- Le Gat, Y., Curt, C., Werey, C., Caillaud, K., Rulleau, B., Taillandier, F. (2025). Water infrastructure asset management: state of the art and emerging research themes. Structure and Infrastructure Engineering, 21(4), 539-562, https://doi.org/10.1080/15732479.2023.2222030.
- Liu, Z., Kleiner, Y., Rajani, B., Wang, L., Condit, W. (2012). Condition assessment technologies for water transmission and distribution systems (EPA/600/R-12/017). U.S. Environmental Protection Agency. https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=241510&Lab=NRMRL.
- Mackey, T., Cashman, A., Cumberbatch, R. (2014). Identification of factors contributing to the deterioration and losses in the water distribution system in Barbados (CERMES Technical Report No. 68, 73 pp.). The University of the West Indies, Centre for Resource Management and Environmental Studies. https://www.cavehill.uwi.edu/cermes/docs/technical_reports/mackey_et_al_2014_pipe_deterioration_and_water_los.aspx.
- Misiunas, D., Vítkovský, J., Olsson, G., Lambert, M., Simpson, A. (2006). Failure monitoring in water distribution networks, Water Science & Technology, 53 (4-5), 503–511. https://doi.org/10.2166/wst.2006.154.
- Mohammadagha, M., Najafi, M., Kaushal, V., Jibreen, A. (2025). Machine Learning Models for Reinforced Concrete Pipes Condition Prediction: The State-of-the-Art Using Artificial Neural Networks and Multiple Linear Regression in a Wisconsin Case Study. arXiv, cs.LG, 2502.00363. https://doi.org/10.48550/arXiv.2502.00363.
- Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., … others. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830. http://www.jmlr.org/papers/volume12/pedregosa11a/pedregosa11a.pdf.
- Phan, H. C., Dhara, A. S., Hu, G., Sadiq, R. (2019). Managing water main breaks in distribution networks––A risk-based decision making. Reliability Engineering & System Safety, 191, 106581. https://doi.org/10.1016/j.ress.2019.106581.
- QGIS.org, 2025. QGIS Geographic Information System. QGIS Association. http://www.qgis.org.
- Rajani, B., Kleiner, Y. (2001). Comprehensive review of structural deterioration of water mains: Physically based models. Urban Water, 3(3), 151–164. https://doi.org/10.1016/S1462-0758(01)00032-2.
- Rezaei, H., Ryan, B., Stoianov, I. (2015). Pipe failure analysis and impact of dynamic hydraulic conditions in water supply networks. Procedia Engineering, 119, 253–262. https://doi.org/10.1016/j.proeng.2015.08.879.
- SURS. (2025). Public Water Supply - Water supplied from public water supply (1000 m3). Statistical Office of the Republic of Slovenia. https://pxweb.stat.si/SiStat/en.
- VOKAS. (2025). Annual report 2024. Javno podjetje VODOVOD KANALIZACIJA SNAGA d.o.o. (in Slovenian) https://www.vokasnaga.si/sites/www.jhl.si/files/dokumenti/letno_porocilo_2024.pdf.
- Warad, A.A.M., Wassif, K. Darwish, N.R. (2024). An ensemble learning model for forecasting water-pipe leakage. Sci Rep 14, 10683. https://doi.org/10.1038/s41598-024-60840-x.
- Zevnik, J., Kramar Fijavž, M., Kozelj, D. (2019). Generalized normalized cut and spanning trees for water distribution network partitioning. Journal of Water Resources Planning and Management, 145(10), 1–12. https://doi.org/10.1061/(ASCE)WR.1943-5452.0001100.