FIDO’s powerful leak detection AI now calculates leak size. This gives water companies a powerful new tool to transform leakage performance and water loss.
The breakthrough makes it easier to prioritise larger leaks for investigation and repair. As a result, this will help tackle the global issue of non-revenue water more effectively.
FIDO AI is already more than 92% accurate at identifying and locating leaks. It works with audio or vibration files from any acoustic sensor or hydrophone, on any pipeline material.
Now developers have used the growing data in FIDO’s neural networks to train it to extract information on the volume of leaks.
Validated, tested and in daily use
FIDO works by applying deep-learning principles to the raw data in sensor files. First, it decides if there is a leak. Second, it calculates leak size. Finally, it ranks leaks as large, medium or small.
Importantly, the volumetric range of each size band can be adjusted according to individual network characteristics.
The function went live in December after three months of validation and testing and is now in daily use. It can also be used on networks without district metered areas (DMAs).
More leak detection breakthroughs to come
Leak sizing is the latest function of FIDO’s leak detection AI. But already our developers are working on the next breakthrough.
For instance, they are currently training the algorithm to narrow leak location to a single defined point between sensors.
FIDO is software-as-a-service (SaaS) and so can be up-and-running in tandem with your existing processes within 24 hours. What’s more, it works with any device because it is sensor, platform and pipeline material agnostic.
Being able to grade leaks by size is a real breakthrough in leak detection. That’s because it provides teams with that missing piece of the puzzle and gets them to their largest leaks. In addition, they get there with fewer wasted visits or expensive dry digs because FIDO is so accurate.Victoria Edwards, FIDO CEO