Artificial Intelligence in Water Infrastructure Management: A Systematic Review and Cognitive Architecture Proposal

Authors:
Open Access
Publisher website
Google Scholar
Abstract
This paper presents a Systematic Literature Review (SLR) on the application of Artificial Intelligence (AI) techniques in smart water networks, analysing scientific production from 2015 to 2025. The primary objective is to identify predominant algorithms and existing technological gaps to propose a reference architecture. The methodology adhered to PRISMA guidelines, filtering an initial pool of 607 records down to 15 primary studies of high technical relevance. The results evidence a predominance of Convolutional Neural Networks (CNN) for automated meter reading and Reinforcement Learning algorithms for distribution optimization. However, a significant lack of ethical and regulatory frameworks for autonomous decision-making was detected. Finally, a conceptual “Cognitive Water Intelligence” architecture is proposed, integrating prediction and anomaly detection modules, offering a roadmap for the transition towards autonomous systems.

This publication has 50 references indexed in Scilit: