Algorithm assesses thermal imaging and high-resolution photogrammetry visualised in Sensat to predict where water leakage is present.
Sensat, the leading visualisation and collaboration software company for civil infrastructure, today announced a pilot study to automate the detection of leakage in the water network to support United Utilities’ commitment to reduce leakage by another 15% by 2025.
It is the latest innovation to be trialled by North West water firm United Utilities as part of its strategy to reduce leakage and maintain resilient water supplies in the future.
Greg Larner, United Utilities’ leakage technical officer, said: “With shifting weather patterns, due to climate change, there is an increased risk of periods of drought, so it is important for us to explore the latest technologies to reduce water wastage and protect the environment. We’re increasingly collecting vast amounts of data about leakage, so artificial intelligence is becoming a key tool in helping to make sense of it.”
The water firm is no stranger to trying new technology in its battle to beat water leaks. As well as using specially-trained sniffer dogs to sniff out leaks, United Utilities has also teamed up with inventors, as part of its Innovation Lab challenge, to develop the FIDO artificial intelligence solution which is now in use around the globe. FIDO uses machine-learning and AI to pinpoint the location and even the size of the leak, just by the noise it makes.
By utilising unmanned aerial vehicle (UAV) technology to remotely map thermal data, combined with high-resolution photogrammetry, Sensat has developed a probability-based algorithm that assesses vegetation stress, terrain relief and thermal scans to predict where leaks might be along trunk mains in rural areas. This new approach to identifying leak locations more efficiently over a wider survey area is enabling United Utilities to take remedial action faster and more cost-effectively. Working closely with the United Utilities' leak detection experts in an invaluable feedback loop, the validation of results will further inform the models, meaning the algorithms will improve their accuracy over time at predicting the leaks, providing further efficiencies in the future.
“When you consider that the reliable supply of water underpins almost every aspect of human life and the threat of climate change will mean we will have to prepare for a drier future, it’s vital that we work together to tackle the challenges that are affecting the water industry,” said James Dean, CEO at Sensat. “Our visualisation platform identifies leaks through geospatial mapping and its probability model is enabling a collaborative approach for analysis, removing the need for manual inspections to significantly reduce the time and cost for detecting leaks. This supports United Utilities’ commitment to securing resilient water supplies on track and our commitment to helping companies be more sustainable by reducing wastage.”
The successful completion of this trial will open up the exploration of further use cases for data beyond leak detection to help deliver the appropriate level of resilience that the water network needs for the long term.