Media Centre
Case studies

Using big data to detect and remedy costly pipe leaks

“By working together on this project with experts from both our company and from CGI, we achieved outstanding results. Thanks to their expertise with big data and their hands-on approach, CGI helped us ‘fixthe leak’.”

Erik Driessen, Innovation Manager at Vitens

Guaranteeing a reliable supply of drinking water to 5.5 million customers in the Netherlands is the job of the Vitens water supply company. Before flowing out of a customer’s faucet, water is first purified at a production facility and then sent through thousands of kilometers of pipes. With the help of big data and predictive analytics, CGI has made it possible for Viten to detect and remedy leaks in these pipes more quickly and easily.

The challenge
Pipe leaks present problems for customers and have a major impact on Vitens’ operations. A leak can force customers to go hours without running water and require dozens of workers to be deployed to repair the leak. During what has come to be known as the “Leak of Leeuwarden” in March 2013, it took 3 hours and 30 workers to locate and address a massive leak. To resolve incidents like these more efficiently and to simplify the maintenance of 96 water production facilities and 49,000 kilometers of pipes, Vitens has a great deal of data at its disposal that can be put to use. The question is how.

The solution
CGI proposed the use of big data and predictive analytics to address pipe leaks and improve overall maintenance. Workshops involving a multidisciplinary team of Vitens and CGI professionals were first conducted to study the details of real-world leaks, such as the Leak of Leeuwarden. Next, CGI developed a proof of concept solution. Data on variables such as pressure, flow, temperature, conductivity, physical location of the pipe network and land registration were loaded into our big data lab and analyzed by a data scientist and domain expert. Using predictive analytics and visualization software, we searched for data patterns that could be used to detect or predict incidents.

The results
The approach of the multidisciplinary team from Vitens and CGI made it possible to detect leaks within a 2.5 kilometer radius in 50 percent of cases. This demonstrated that, by using big data and predictive analytics, leaks can be detected and repaired faster. As a result, damage caused by leaks can be minimized, and customers can be assisted more promptly when incidents occur.  With this new model, Vitens is expected to save millions of euros, prevent interruptions in the supply of water and eventually be able to diagnose and repair leaks before they actually occur.

Why CGI
CGI has taken a lead role in the use of big data and predictive analytics. We strongly believe in the profitable opportunities that large volumes of information can generate for businesses. Water supply companies in Europe invest 20 billion euros each year to keep their networks up to date, and substantial cost savings can be achieved by leveraging big data and predictive analytics to detect leaks and maintain these networks. We work with clients to turn these opportunities into reality, driving performance and results from big data.

Contact us for more information