Expert: AI will help avoid water losses in the event of a water main failure

Water supply failures result in water losses of approximately 30 percent of resources, sometimes reaching as much as 60 percent. The use of artificial intelligence enables the quick location of areas where failures have occurred or may occur, Dr. Maciej Niemir from Łukasiewicz-PIT told PAP.

The use of AI in water utilities can be based on predictive maintenance, which involves collecting information to prevent failures and identifying anomalies and characteristic patterns that indicate their progression. Data can be collected using sensors, cameras, and even special microphones. Their analysis allows for early incident detection, making it easier to determine where and when a failure will occur.

In addition to traditional equipment condition monitoring solutions, solutions combining artificial intelligence with a distributed network of sensors collecting data in real time are increasingly being implemented. An example is acoustic modules, which—crucially—can be installed non-invasively, without the need for excavation, on any section of the pipeline. The sensors record vibrations and oscillations accompanying water flow, and the collected sound data is analyzed by AI algorithms to quickly detect irregularities.

"It's enough to install a few, or preferably several dozen, sensors in various locations around the city, and that's enough for the system to listen and collect data, and the next day there will be information about a potential leak and its location," said Dr. Maciej Niemir, Chief Artificial Intelligence Specialist at Łukasiewicz - Poznań Institute of Technology.

Listening is conducted in real time. The expert explained that a large amount of this type of data is required for the AI to detect specific signals.

"Of course, the AI must eliminate sounds like the passing of a car or train and distinguish between those that occur during the day and result from normal use. Then, we're left with the 'noise' that indicates there's a malfunction somewhere," he explained.

The specialist noted that frequent measurement and analysis of data is essential because many leaks remain invisible until the next, infrequent infrastructure inspection.

Information about water leaks can also be located using radar satellite images. These provide information about humidity in a given area. Real-time images are compared with previously captured images. This allows for the identification of areas with higher or lower humidity, which in turn can be used to pinpoint the location of the leak.

Sensors that collect data and communicate in real time, as well as systems based on artificial intelligence, are rapidly becoming cheaper while their capabilities are growing, opening up real prospects for modernizing water infrastructure. While full network automation will take time, the potential for improvement is enormous.

"There's practically no water network that couldn't be improved. Average water losses currently reach around 30%, but can sometimes reach as much as 60%. We should be committed to saving more water, and artificial intelligence can be a great help in this regard," the expert said.

Artificial intelligence can also be helpful in controlling water and wastewater treatment processes.

This type of technology was used at the Water and Sewage Company in Tarnowskie Góry, where water losses were reduced from approximately 30% to 10-11%. As expert Włodzimierz Woźniak emphasized in the press release, solutions using automation and AI are expensive, but they deliver long-term savings.

According to a 2021 report by the Central Statistical Office, Poland ranked 24th in the European Union in terms of renewable freshwater resources. It has 1,600 cubic meters of water per capita. Only three countries – the Czech Republic, Cyprus, and Malta – rank behind Poland, meaning that renewable drinking water resources are at a low level. (PAP)

Science in Poland

bch/ bar/ speak/

The PAP Foundation permits free reprinting of articles from the Nauka w Polsce website, provided that you notify us by email once a month of your use of the website and cite the source of the article. On portals and websites, please include the linked address: Source: naukawpolsce.pl, and in journals, please include the annotation: Source: Nauka w Polsce website - naukawpolsce.pl. This permission does not apply to information in the "World" category or any photographs or video materials.