The growing climate and regulatory demands are driving the need to maximize operational and energy efficiency in integrated water cycle infrastructure, positioning artificial intelligence as a key tool to optimize processes and move towards more resilient and sustainable management models.
In this context, the MASTERY project has demonstrated in a real environment the potential of green artificial intelligence to improve the operability of WWTPs through predictive algorithms capable of optimizing the control of biological processes under sustainability criteria.
The system was implemented at the Ranilla Wastewater Treatment Plant, managed by Facsa and owned by the Metropolitan Water Supply and Sanitation Company of Seville SA (EMASESA), where new sensors were installed and real-time data acquisition and processing solutions were deployed, including an intelligent multispectral spectrometer. Based on this information, a model was developed capable of correlating optical signals with physicochemical variables of the process, incorporating techniques of machine learning y edge AI to detect deviations early and reinforce operational stability.
Experimentation in different scenarios allowed the algorithm to be fine-tuned under real-world conditions and a visual recommendation system to be developed that translates the results into operational guidelines applicable by operations personnel. Furthermore, the project has defined a methodology and specific metrics for the implementation of green artificial intelligence, aimed at minimizing the environmental impact of digital technologies and maximizing their energy efficiency.
Last March, the project held its final meeting with a visit to the Ranilla Wastewater Treatment Plant, attended by the technician from the Centre for the Development of Technology and Innovation (CDTI) and the other partners. During the meeting, the main results achieved were shared, and the challenges posed by the digitalization of the sector and the application of green artificial intelligence solutions were discussed.
