The primary aim of the DARROW project is to effectively utilise the data from a wastewater treatment plant to develop various tools based on artificial intelligence, assisting operators in making more informed decisions. In this expert interview, Itxaro Errandonea Aranibar, a researcher at the Ceit Research Center and project coordinator of DARROW, shares valuable insights on how AI can revolutionise wastewater treatment and enhance operational efficiency.
How is AI currently being used in the water treatment/filtration industry?
Currently, AI is applied in a very ad hoc manner. AI-based solutions tend to be very focused on solving specific functionalities such as estimating a particular component at a specific point in the plant. However, it is not applied in a holistic way to optimise the operation of a plant.
Could you please tell me about your DARROW project? Where are you at with the project, what are its aims?
The Darrow project is an EU co-funded project that applies AI solutions in wastewater treatment plants.
The main objective of Darrow project is to use the available data in a wastewater treatment plant efficiently to develop different tools. These are based on artificial intelligence to help operators make better decisions.
The project is currently at the mid-term point. The different AI-based models are being developed and we will soon start working on the integration part of these. This will make it possible to deploy the Darrow platform.
If proven successful, what will the key benefits of using AI in this way be?
There are several benefits that Darrow (AI-based solution) can offer. From the plant operator’s point of view, it can make a big difference to their day-to-day work. Darrow will provide him with the information he needs about the state of the plant and process control recommendations for today as well as recommendations for the future, to assess and make better decisions. From a more general point of view, the Darrow solution will enable you to be more energy efficient, reduce greenhouse gas emissions, contribute to the circular economy and reduce waste.
How might AI be used to improve water treatment/filtration in the future?
Good data quality is key for the application of AI. Therefore, it is necessary to have AI-based models for the treatment, augmentation and enrichment of available plant data (laboratory data, influent data, sensor data,…), as proposed in the Darrow solution. It is then that further AI models can be applied to support more advanced functionalities such as process and control recommendations. Having a platform such as Darrow offers that allows the integration and interaction between the different AI models is key.
Is there anything you’d like to add on the topic?
Another aspect that is addressed in the project, given its high importance, is trying to reach the end-user with the proposed solution. End-users may be reticent to use AI-based solutions, depending on which sectors and even more so when it concerns critical infrastructure. For this reason, Darrow has developed different activities oriented to the social analysis of human interaction with AI tools and the application of Explainable AI methods. This way, we achieve a better understanding of the performance of the models by the user.
This interview was conducted by Keri Allan for an article in the July 2024 issue of Filtration + Separation magazine.
Interested in knowing more about intelligent water treatment?