Machine Learning in Wastewater Treatment: A Comprehensive Bibliometric Review

Abstract

Accurate identification and control of wastewater treatment processes are critical for the efficient use of water resources. Advances in online monitoring and computational capabilities have facilitated the integration of artificial intelligence (AI), particularly machine learning (ML), into wastewater treatment systems. This review analyzes 433 studies on ML applications in wastewater treatment from 2000 to 2022 using bibliometric methods, examining research trends, hotspots, and future directions. Since 2015, the field has experienced a significant surge in publications. The United States and Spain are notable for their long-standing contributions, while China, despite entering the field late in 2012, has emerged as the leading contributor in publication volume. Keyword analysis reveals “neural networks” and “artificial neural networks” as the most frequently applied ML techniques, alongside terms like “prediction”, “optimization”, “fault detection”, and “design”. Our comprehensive review further shows that ML applications in wastewater treatment primarily focus on feature identification, parameter prediction, anomaly detection, and optimized control with key application scenarios including systems, wastewater, waste gas, and sludge. As the demand for AI in wastewater treatment continues to grow, multimodel integration and in-depth development may become the focus of future research to address multiobjective challenges in wastewater treatment more effectively.