Artificial intelligence against leaks and polluting emissions from sewerage systems

Water is one of the elements that is most strongly affected by the climate crisis, and we are seeing proof of this in recent months as never before. Drought, melting glaciers, and desertification are all phenomena that require an immediate commitment from everyone. For example, the Po River District Basin Authority (only italian version) found that at the beginning of September the filling percentage for Lake Como was 5%, 15% for Lake Maggiore and 27% for Lake Garda. These figures are broadly in line with or even below the historical lows since 2003 (7%, 11% and 3% respectively).

At A2A we are working to reduce losses and improve the condition of the water network: flipping through the pages of our 2021 Integrated Report shows that – in the last four years – waste on our networks has decreased by 15%.
System efficiency depends first and foremost on optimal monitoring and proper data collection. This is an area where start-ups can make a significant contribution, so we try to involve them, meeting their speed requirements and simplifying procedures. As a rule, the work on pilot projects involves three phases: identification of innovation needs, selection of new technologies, and initiation and management of experiments.

Adequate monitoring is also a prerequisite for optimal management of the sewage system. Damaged wastewater pipes could lead to polluting emissions into the environment, or to parasitic water entering the system.
Some studies (only italian version) have found that – despite the large amount of time spent on traditional controls involving human intervention – around one in four problems is either not detected or is misclassified. The difficulties in identifying defects are reflected in avoidable costs. 15% (only italian version) of sewage systems are replaced when this is not necessary, while 20% are fixed too late, when the faults have reached the point where a new system has to be installed.
 

Specifically, the goal of making our water and sewage infrastructure more efficient has found the ideal partner in Pallon (only italian version). This start-up – which was founded in 2019 at ETH Zurich – has developed computer vision software that can analyse footage taken by a robot inside sewage pipes, so that imperfections can be detected more quickly and in greater detail than is possible with an operator. 

Among other things, the software can reduce visual inspection times by up to 20% and video analysis times by up to 90%. In addition to a significant reduction in costs, this project will make sewerage systems even safer, further limiting the emission of pollutants into the surrounding area. This is very important work in a critical sector for our country. As explained by Wise (Freshwater information system for Europe), approximately 15.6 million cubic metres of wastewater are generated per day in urban centres in Italy, and only 56% of this is treated in line with European legislation, compared to a target of 95%.

This is not all. In addition to these listed benefits, Pallon is also an excellent data collector. This is a very important issue in business, and it will increasingly be at the heart of the digital transition in the future. It is no coincidence that the European Union is working on a package of reforms aimed at fostering the emergence of an increasingly open and regulated data-driven economy, revolutionising the European market. Therefore, partly due to the introduction of these legislative innovations, the partnership with Pallon represents a step towards becoming a data-driven company.