Northumbrian Water (NWL) supplies more than 2.7 million people across the North East of England, managing a 17,500 km distribution network spanning both urban and rural communities. With non-revenue water at around 18%, the utility has set ambitious targets to reduce leakage by 8% by 2030 and 55% by 2050.
THE CHALLENGE
After successfully using FIDO AI to support leakage investigations, Northumbrian Water wanted to evaluate whether AI-powered fixed network monitoring could detect leaks earlier and at scale during the 2025–2026 winter breakout season. The goal was to improve leakage performance without adding pressure to already stretched analytics and operational teams.
THE SOLUTION
FIDO analysed DMA leakage data to identify the 17 areas where rising nightlines were most difficult to control and where faster intervention would have the greatest impact. Using network GIS data, FIDO AI designed an optimised deployment of approximately 1,035 acoustic sensors and IoT relay devices.
The sensors automatically upload overnight acoustic recordings to the FIDO AI platform, which identifies likely leaks and generates high-confidence points of interest for field investigation. To further reduce the workload on utility staff, Northumbrian Water adopted FIDO’s Data-as-a-Service (DaaS) model, delivering actionable leak locations directly to leakage teams for faster response and repair.
RESULTS
Despite a relatively mild winter, the programme has reduced leakage by nearly 267 m³/hour (equivalent to 6.4 ML/day), including three significant step-change reductions. The permanent monitoring network now detects leaks soon after they occur, enabling faster investigations and repairs while reducing the duration and impact of leakage events on overall network performance.
Project Snapshot
149km water network covered
130 points of investigation found
267m³ leakage reduction per hour
