The role of artificial intelligence in preventing water leakages from water supply networks

Introduction. One of foundations for the sustainable development and improvement of centralized water supply networks (CWN) and sanitation is the use of artificial intelligence (AI) based on machine learning (ML) algorithms and models: supervised, unsupervised, reinforcement learning. Leaks and unauthorized connections to CWN pose risks, leading to losses of drinking water and reduced pricing in the field of water resource metering. The relevance is associated with solving practical AI problems based on the latest innovations — forecasting and preventing accidents at CWN with optimal planning of repair work and timely maintenance. The purpose of study is to substantiate the role of AI using ML tools in the tasks of predicting pipeline failures and emergency situations in CWN.

Materials and methods. The study of information on the role of AI in preventing water leaks from water supply networks was carried out using the method of literature review of the used AI algorithms for predicting pipe failures in CWN.

Results. The ML models used for diagnostic analysis to predict water leaks from CWN are identified and presented. The review of technologies shows the use of 18 ML algorithms to solve problems related to leaks in CWN. Start of use of Kohonen neural network algorithms (KNN) in Russia indicates the availability of the only neural network software translated into Russian, STATISTICA Automated Neural Networks. Acoustic and ultrasonic methods for monitoring the condition of underground pipeline networks, based on the propagation of volumetric and directional waves (noise), are beginning to develop rapidly.

Conclusions. Among the conclusions — for the Sustainable Development of CWN, water utilities need to ensure reliable and continuous data collection, this is a key practice that will help make reliable decisions based on AI predictions after the ML phase. Databases may include: pipe diameter, length of the section and age of the pipe, pressure, type of soil. The pressure itself (or difference) in the network is not a sign of an accident. This parameter should be considered together with the number of network failures (accidents) in the sections.

1. Robles-Velasco A. A machine learning approach to predict pipe failures in water distribution networks. 2022; 151. URL: https://idus.us.es/handle/11441/131484

2. Weeraddana D., Hapuarachchi H., Kumarapperuma L., Khoa N.L.D., Cai C. Long-term water pipe condition assessment: a semiparametric model using gaussian process and survival analysis. Lecture Notes in Computer Science. 2020; 487-499. DOI: 10.1007/978-3-030-47436-2_37

3. Pouri Z., Heidarimozaffar M. Spatial analysis and failure management in water distribution networks using fuzzy inference system. Water Resources Management. 2022; 36(6):1783-1797. DOI: 10.1007/s11269-022-03104-5

4. Sattar A.M.A., Gharabaghi B., McBean E.A. Prediction of timing of watermain failure using gene expression models. Water Resources Management. 2016; 30(5):1635-1651. DOI: 10.1007/s11269-016-1241-x

5. Sangroula U., Han K.H., Koo K.M., Gnawali K., Yum K.T. Optimization of water distribution networks using genetic algorithm based SOP–WDN program. Water. 2022; 14(6):851. DOI: 10.3390/w14060851

6. Primin O.G., Gromov G.N. Hydraulic calculations improvements of water supply systems by using electronic models. International Journal for Computational Civil and Structural Engineering. 2018; 14(2):141-148. DOI: 10.22337/2587-9618-2018-14-2-141-148. EDN OYBKZZ. (rus.).

7. Primin O.G., Gromov G.N., Ten A.E. Algorithms for constructing and calibrating electronic models of water supply system. Vestnik MGSU [Proceedings of the Moscow State University of Civil Engineering]. 2018; 13(7):847-854. DOI: 10.22227/1997-0935.2018.7.847-854 (rus.).

8. Kleiner Y., Rajani B. Comparison of four models to rank failure likelihood of individual pipes. Journal of Hydroinformatics. 2012; 14(3):659-681. DOI: 10.2166/hydro.2011.029

9. Wang R., Dong W., Wang Y., Tang K., Yao X. Pipe failure prediction: A data mining method. 2013 IEEE 29th International Conference on Data Engineering (ICDE). 2013; 1208-1218. DOI: 10.1109/ICDE.2013.6544910

10. Francis R.A., Guikema S.D., Henneman L. Bayesian belief networks for predicting drinking water distribution system pipe breaks. Reliability Engineering & System Safety. 2014; 130:1-11. DOI: 10.1016/j.ress.2014.04.024

11. Aydogdu M., Firat M. Estimation of failure rate in water distribution network using fuzzy clustering and LS-SVM methods. Water Resources Management. 2015; 29(5):1575-1590. DOI: 10.1007/s11269-014-0895-5

12. Komba G.M., Mathonsi T.E., Owolawi P.A. Water pipeline leak detection and localisation in water distribution networks. 2023 International Conference on Emerging Trends in Networks and Computer Communications (ETNCC). 2023; 140-145. DOI: 10.1109/ETNCC59188.2023.10284934

13. Komba G.M., Mathonsi T.E., Owolawi P.A. Optimizing leak detection and location in water distribution networks using SVM-RF algorithm. 2024 15th International Conference on Mechanical and Intelligent Manufacturing Technologies (ICMIMT). 2024; 27-33. DOI: 10.1109/ICMIMT61937.2024.10585935

14. Amaitik N.M., Buckingham C.D. Developing a hierarchical fuzzy rule-based model with weighted linguistic rules: A case study of water pipes condition prediction. 2017 Computing Conference. 2017; 30-40. DOI: 10.1109/SAI.2017.8252078

15. Farmani R., Kakoudakis K., Behzadian K., Butler D. Pipe failure prediction in water distribution systems considering static and dynamic factors. Procedia Engineering. 2017; 186:117-126. DOI: 10.1016/j.proeng.2017.03.217

16. Rifaai T.M., Abokifa A.A., Sela L. Integrated approach for pipe failure prediction and condition scoring in water infrastructure systems. Reliability Engineering & System Safety. 2022; 220:108271. DOI: 10.1016/j.ress.2021.108271

17. Fan X., Wang X., Zhang X., Yu X. Machine learning based water pipe failure prediction: The effects of engineering, geology, climate and socio-economic factors. Reliability Engineering & System Safety. 2022; 219:108185. DOI: 10.1016/j.ress.2021.108185

18. Snider B., McBean E.A. Watermain breaks and data: The intricate relationship between data availability and accuracy of predictions. Urban Water Journal. 2020; 17(2):163-176. DOI: 10.1080/1573062X.2020.1748664

19. Giraldo-González M.M., Rodríguez J.P. Comparison of statistical and machine learning models for pipe failure modeling in water distribution networks. Water. 2020; 12(4):1153. DOI: 10.3390/w12041153

20. Almheiri Z., Meguid M., Zayed T. Intelligent approaches for predicting failure of water mains. Journal of Pipeline Systems Engineering and Practice. 2020; 11(4). DOI: 10.1061/(ASCE)PS.1949-1204.0000485

21. Tang K., Parsons D.J., Jude S. Comparison of automatic and guided learning for Bayesian networks to analyse pipe failures in the water distribution system. Reliability Engineering & System Safety. 2019; 186:24-36. DOI: 10.1016/j.ress.2019.02.001

22. Shekofteh M., Jalili Ghazizadeh M., Yazdi J. A methodology for leak detection in water distribution networks using graph theory and artificial neural network. Urban Water Journal. 2020; 17(6):525-533. DOI: 10.1080/1573062X.2020.1797832

23. Peng S., Cheng J., Wu X., Fang X., Wu Q. Pressure sensor placement in water supply network based on graph neural network clustering method. Water. 2022; 14(2):150. DOI: 10.3390/w14020150

24. Tavakoli R., Sharifara A., Najafi M. Artificial neural networks and adaptive neuro-fuzzy models to predict remaining useful life of water pipelines. World Envi-ronmental and Water Resources Congress 2020. 2020; 191-204. DOI: 10.1061/9780784482988.019

25. Romanov R.V., Kochetkova S.S. Application of a neural network approach for localization of problem areas of a centralized water supply system. Third International Conference on Optics, Computer Applications, and Materials Science (CMSD-III 2023). 2024; 30. DOI: 10.1117/12.3025000

26. Mehran Jafari S., Bozorg-Haddad O., Reza Nikoo M. Application of artificial neural network and fuzzy logic in the urban water distribution networks pipe failure modelling. Studies in Computational Intelligence. 2022; 333-354. DOI: 10.1007/978-981-19-2519-1_16

27. Rashid A., Kumari S. Optimal design of water distribution network using neuro – fuzzy technique. 2021. DOI: 10.21203/rs.3.rs-915481/v1

28. Gheibi M., Moezzi R., Taghavian H., Wacła-wek S., Emrani N., Mohtasham M. et al. A risk-based soft sensor for failure rate monitoring in water distribution network via adaptive neuro-fuzzy interference systems. Scientific Reports. 2023; 13(1). DOI: 10.1038/s41598-023-38620-w

29. Yang Y., Hu Y., Zheng J. A decision tree approach to the risk evaluation of urban water distribution network pipes. Safety. 2020; 6(3):36. DOI: 10.3390/safety6030036

30. Lučin I., Lučin B., Čarija Z., Sikirica A. Data-driven leak localization in urban water distribution networks using big data for random forest classifier. Mathematics. 2021; 9(6):672. DOI: 10.3390/math9060672

31. Lucin I., Carija Z., Druzeta S., Lucin B. Detailed leak localization in water distribution networks using random forest classifier and pipe segmentation. IEEE access. 2021; 9:155113-155122. DOI: 10.1109/ACCESS.2021.3129703

32. Marvin G., Grbčić L., Družeta S., Kranjčević L. Water distribution network leak localization with histogram-based gradient boosting. Journal of Hydroinformatics. 2023; 25(3):663-684. DOI: 10.2166/hydro.2023.102

33. Parajuli U., Shin S. Identifying failure types in cyber-physical water distribution networks using machine learning models. AQUA — Water Infrastructure, Ecosystems and Society. 2024; 73(3):504-519. DOI: 10.2166/aqua.2024.264

34. Timashev S.A., Makeeva T.V. Assessment and prediction of water supply network reliability under information shortage using artificial neural networks. ASCE Inspire 2023. 2023; 733-741. DOI: 10.1061/978078-4485163.085

35. Timashev S.A., Makeeva T.V. Assessing the reliability of the urban water supply network under information deficit using artificial neural networks. Ekaterinburg, Publishing house of the Ural University, 2023; 78. (rus.).

36. Karambirov S.N., Umansky P.M. Cluster analysis of parts of water supply network. Prirodoobustrojstvo. 2016; 1:23-27. EDN VVWITX. (rus.).

37. Scussel O., Brennan M.J., Almeida F.C., Muggleton J.M., Rustighi E., Joseph P.F. Estimating the spectrum of leak noise in buried plastic water distribution pipes using acoustic or vibration measurements remote from the leak. Mechanical Systems and Signal Processing. 2021; 147:107059. DOI: 10.1016/j.ymssp.2020.107059

38. Liu R., Zayed T., Xiao R. Advanced acoustic leak detection in water distribution networks using integrated generative model. Water Research. 2024; 254:121434. DOI: 10.1016/j.watres.2024.121434

39. Yussif A.M., Sadeghi H., Zayed T. Application of machine learning for leak localization in water supply networks. Buildings. 2023; 13(4):849. DOI: 10.3390/buildings13040849

40. Fares A., Tijani I.A., Rui Z., Zayed T. Leak detection in real water distribution networks based on acoustic emission and machine learning. Environmental Technology. 2023; 44(25):3850-3866. DOI: 10.1080/09593330.2022.2074320