A thought-provoking publication looks at the potential for application of machine learning to the non-invasive monitoring of biofilms as a means of increasing their application and enhancing their performance in wastewater treatment.
The paper lies at the intersection of several key areas of interest to many EBNet members, including those in the AI/ML, Aerobic Granulation and ESWS WGs, as well as colleagues in the National Biofilms Innovation Centre (NBIC) and elsewhere. It also ties in with our ongoing interest in microenvironments, as part of Theme 2.
Dr Joe Weaver of Durham University, who coordinated the EBNet-sponsored Biofilm Microenvironments Workshop, said: “Biofilm characteristics critically define how many systems perform, yet directly and non-invasively measuring those characteristics in real-time remains a bottleneck towards progress, reflecting many of the issues and themes discussed at the EBNet Workshop. This paper proposes a machine learning-based approach to that challenge, while remaining aware of the pitfalls of naively applying ML to a problem. Whether the route ultimately leads to robust digital twins, improved process control, or new scientific insights, it points in a direction worth exploring”.
Dr Oliver Fisher of the University of Nottingham, who is Lead author on the paper, also co-Leads EBNet’s AI/ML WG. Outputs from the Microenvironments workshop are available from the Resources page in the form of an Executive Summary, Position Statement, Outcomes & Recommendations, and Full Report.
Making waves: Transforming biofilm-based wastewater treatment using machine learning-driven real-time monitoring. Fisher, O.J., Wang, Y. and Ahmed, A., 2025. Water Research, p.124491. doi: 10.1016/j.watres.2025.124491
