SmartWaterGrid: state of the art technology to detect water leaks up to street level
As the climate changes, and we are experiencing longer periods of drought, water starts to become scarcer. This makes efficient water provision even more critical. Each year, an estimated 35% of the entire water supply in Europe gets lost, which corresponds to the yearly consumption of 200 million people or 2 billion € per year!
Detecting and fixing water leaks in an efficient way has therefore become a key challenge for Water Utilities everywhere.
That is why a consortium of research and industry partners, among which Itineris, joined forces in the SmartWaterGrid project. We pooled our knowledge and expertise to conduct research on smart technologies to detect small and medium water leaks up to street level.
Creating a hybrid AI twin of the water supply network
Currently, when a water leak is identified in a certain zone, dozens – sometimes even hundreds – of kilometers of pipes have to be looked at to find the leak. It could take weeks, or even months, to locate it. In most cases, the Water Utility just has to wait until a customer calls to report a leak or incident.
That is why it is so important to find new state-of-the-art solutions to detect and deal with leaks in an efficient and smart way.
Within the SmartWaterGrid project, we created a hybrid AI twin of the water supplynetwork, using input from Internet of Things (IoT) sensors, Geographical Information System (GIS) data, hydraulic models, and customer and expert feedback. By using machine learning models, this input is translated into leak alerts and the localization of the leak in the water network.
As a result, leak localization is improved from 70 km to street level, and from weeks to a few hours.
One holistic view for the customer service agent
Itineris’ role within this project was to help translate the expert and customer feedback data into input for the hybrid model. This was very important because research showed that the more knowledge we included in the machine learning models, the better the hybrid AI twin succeeded in locating the leaks.
We worked on this together with De Watergroep, one of the largest drinking water companies in the Benelux region, and one of our customers.
We created a dashboard in a Microsoft Customer Service demo environment, displaying customer service feedback, planned service orders, and the hybrid model output. This way customer service agents have a holistic view, allowing them to better inform their customers, which significantly increases customer engagement and satisfaction.
How a smart water grid solution can benefit Water Utilities
The implementation of a smart water grid solution has many benefits:
First and foremost, it decreases water losses and protects the water supply.
As leaks can be detected much faster, it also creates a positive business case for lower leakage rates.
Last but not least: it decreases customer hindrance and friction. With the correct parameters, we can e.g. set up a proactive alerting system, allowing Utilities to keep their customers informed about the status of a water leak.
Discover why De Watergroep joined this project
Want to learn more?
Please do not hesitate to get in touch and we’ll gladly share what we have learned from this project.
About the SmartWaterGrid project
SmartWaterGrid was an imec.icon research project funded by imec and Agentschap Innoveren & Ondernemen. It started on October 1st, 2019, and ran until September 30th, 2021. The project consortium consisted of various research and industry partners: Aloxy, De Watergroep, Hydroscan, Itineris, IDLab UGent, and IDLab UAntwerpen.
We are currently looking into an extension of the project to further investigate topics such as data quality, optimizing water processes and procedures, smart asset management systems, …