Assessment of the Implications and Challenges of Using Artificial Intelligence for Urban Water Networks in the Context of Climate Change When Building Future Resilient and Smart Infrastructures

Urban water networks are more than obligatory infrastructures that help to develop and sustain cities; they are integral to urban resilience. Water networks have undergone an evolutionary process, yet persistent challenges continue, accompanied by complex climate dynamics that necessitate new thinking. This paper focuses on artificial intelligence—specifically, machine learning—with respect to transforming urban infrastructures into smart cities. The following is an examination of two key impacts that machine learning has in water infrastructure: detection and localization of leaks and prediction of failure in pipes due to physical factors. The aim of this paper is to combine the latest research that has been carried out up to 2023, along with the implementation of machine learning models and their scenarios and impacts. The results are promising with the most used support vector machines and neural networks, although effectiveness varies due to the physical and external socioeconomic or climatic conditions related to water. Major persistent problems include user demand uncertainty, unbalanced data, measurement uncertainties, noise, and calibration drifts. Future research is encouraged in areas like water demand forecasting and quality management through machine learning. There is also the potential for integration of this technology with other techniques, such as geotechnical modeling and/or experimental building methods to help overcome the current challenges and contribute to building resilient, smart urban infrastructures.