Microplastics enter wastewater treatment plants from households, industrial activities, textile fibers, personal care products, and urban runoff. Although treatment facilities can capture a large proportion of these particles, they were not originally designed to eliminate microplastics, particularly the smallest particles.
A new review published in Artificial Intelligence & Environment examines how artificial intelligence could strengthen the detection, monitoring, removal, and management of microplastics throughout wastewater treatment systems.
"Wastewater treatment plants provide an essential barrier against microplastic pollution, but removal does not necessarily mean elimination," said corresponding author Konstantinos Tsamoutsoglou of the Technical University of Crete. "By combining reliable analytical measurements with artificial intelligence, treatment facilities could move from occasional monitoring toward faster, predictive, and more informed management."
The researchers synthesized studies published between 2019 and 2026, bringing together evidence on microplastic sources, concentrations, analytical methods, treatment technologies, and emerging AI applications.
Reported microplastic concentrations vary widely because treatment plants differ in design and because researchers use different sampling and detection methods. Even so, the review identifies a clear pattern across the treatment process. Preliminary and primary treatment stages remove approximately 72 percent of incoming microplastics. Secondary treatment raises overall removal to about 88 percent, while tertiary processes may achieve rates close to 94 percent.
High percentage removal, however, does not guarantee that treated water is free of plastic particles. Wastewater plants process enormous volumes of water, meaning that even low concentrations in discharged effluent can translate into millions of particles entering rivers, lakes, and coastal waters. Particles smaller than about 150 micrometers are especially difficult to retain because they can remain suspended and pass through conventional treatment units.
The review also highlights a less visible environmental pathway. An estimated 60 to 80 percent of retained microplastics can accumulate in sewage sludge rather than being destroyed. When this sludge is reused as fertilizer or soil amendment, the particles may be transferred to agricultural land, where they can persist, interact with soil microorganisms and plant roots, or move through runoff and terrestrial food webs.
This finding suggests that wastewater treatment may partly relocate microplastic contamination from aquatic environments to land.
Artificial intelligence could help treatment operators understand and manage these complex pathways. Computer vision systems can automatically detect and classify particles in microscope images, reducing manual workload and operator bias. Under controlled conditions, some image based models have achieved classification accuracies above 85 to 95 percent.
Machine learning can also connect particle measurements with operational data such as flow rate, turbidity, suspended solids, and sludge recirculation. These models could help predict removal efficiency, identify likely pollution sources, detect unusual conditions, and guide treatment adjustments or sludge management decisions.
The authors propose a three part framework for intelligent monitoring: collecting laboratory, imaging, spectroscopic, and operational data; using data driven models for particle classification and performance prediction; and integrating model outputs into early warning systems and operational decision support.
Major challenges remain before these tools can be widely deployed. Many models are trained on small laboratory datasets that do not represent the complexity of real wastewater. Monitoring protocols also differ considerably among studies, limiting comparisons and model transfer between facilities. Larger standardized datasets, external validation, transparent algorithms, and full scale testing will therefore be essential.
The authors conclude that AI should complement, rather than replace, chemical analysis and established treatment expertise. Combining these approaches could support more accurate monitoring and more sustainable control of microplastic pollution across both water and land environments.
===
Journal reference: Tsamoutsoglou K; Gkika V; Xu L. AI-assisted microplastics detection, removal and management: Advances and challenges in wastewater. AI Environ. 2026, 1(2): 57-76. DOI: 10.66178/aie-0026-0012
https://www.the-newpress.com/aie/article/doi/10.66178/aie-0026-0012
===
About the Journal:
Artificial Intelligence & Environment is an international multidisciplinary platform for communicating advances in fundamental and applied research on the intersection of environmental science and artificial intelligence (AI). It is dedicated to serving as an innovative, efficient and professional platform for researchers in the cross-discipline fields of earth and environmental sciences, big data science and AI around the world to deliver findings from this rapidly expanding field of science. It is a peer-reviewed, open-access journal that publishes critical review, original research, rapid communication, view-point, commentary and perspective papers.
