Advances and Applications of Artificial Intelligence in Wastewater Treatment: A Bibliometric Analysis and Systematic Review

Abstract

Artificial intelligence (AI) has emerged as a key technology for optimizing wastewater treatment by improving monitoring, contaminant prediction, and operational efficiency. However, its application remains fragmented due to the variety of models and metrics employed, which limits a comprehensive understanding of its impact and effectiveness. The objective of this paper was to analyze the approaches, models, and metrics used in recent scientific literature. The methodology followed PRISMA guidelines, analyzing 5,487 studies from five databases (Scopus, Web of Science, IEEE Xplore, EBSCOhost, and ProQuest) between 2017 and 2025, and selecting 63 papers based on quality and relevance criteria. The findings revealed that the most common metrics used to assess AI effectiveness are RMSE, MAE, and R²; that the highest-impact journals are concentrated in quartiles Q1 and Q2, reflecting a high level of scientific rigor; and that China, the United States, and India lead the collaboration and co-occurrence networks. Moreover, emerging topics revolve around deep learning, the Internet of Things (IoT), and environmental sustainability. Overall, the review confirms that AI represents a strategic tool to enhance operational efficiency and decision-making in wastewater management, opening new perspectives toward intelligent, sustainable, and adaptive systems.