Smart Leak Detection in Water Distribution Networks: A Systematic Review and D-A-T Taxonomy

Leaks in water distribution networks present a significant challenge for water resource management, necessitating timely and reliable detection methods. However, the literature remains fragmented across sensing technologies, data-processing approaches, and diagnostic capabilities. This article presents a systematic review of technologies and methods for water leak detection, localisation, and characterisation, providing a structured classification and critical analysis of the available evidence. The review was conducted and reported in accordance with PRISMA 2020 and included 91 publications, comprising 70 primary studies and 21 review articles. Literature searches were performed in June 2026 across Scopus, Web of Science, IEEE Xplore, ScienceDirect, SpringerLink, MDPI, ASCE, and Taylor & Francis. The studies were classified according to the D-A-T taxonomy, encompassing three dimensions: data acquisition, processing architecture, and diagnostic task. Among the 70 primary studies, hydraulic sensing (D1) was identified in 37 studies (52.9%), signal and statistical processing (A2) in 48 (68.6%), and leak detection (T1) in 67 (95.7%). The principal challenges identified concerned validation under real-world operating conditions, generalisation across networks, and the comparability of findings between studies. The review further identifies gaps in the availability of real-world data and highlights multimodal data acquisition, hybrid or fusion-based processing, and multitask diagnosis as potential avenues for future research.