Is 2022 the year the world takes water leakage seriously?

By Founder and Chief Executive Officer

Home > News > Is 2022 the year the world takes water leakage seriously?.

With water running out and era-defining drought making headlines in 2021, FIDO CEO Victoria Edwards asks: Is 2022 finally going to be the year the world takes water leakage seriously?


2022 has to be the year that utilities and governments face up to water leaks. Or face a backlash from pandemic-weary customers and businesses being told they face even more restrictions.

The rate of world water leakage is way too high. Up to 60% in parts of south America and anything up to 50% in the US. It hovers around 20% in the UK. While in the rest of Europe, we’re not even sure. The EU has given member states until 2026 to find out.

The accepted figure worldwide is about a third. Let that sink in. One glass in every three of expensively and carbon-intensively abstracted, treated and pumped drinking water vanishes before it reaches our taps. It is adding unnecessary pressure to sensitive ecosystems, reducing pipeline life expectancy, undermining property and leading to damage, disruption and loss of public confidence.

No other industry would survive it. So, why is it so high?  

Listening for leaks is hard for humans to do

The answer, in part, is that more than 90% of leaks are underground and it’s always been hard and expensive to find them.  Many are small and others are on customers’ own pipes. This is a moral hazard because, as customers are paying for the water, why waste time finding the leak? The reward for all the effort and expense is lower bill receipts.

The way we find leaks dates back to when pipes were first invented – listening.

At a basic level, it’s the embodiment of having your ‘ear to the ground’. An engineer goes out in the quiet of night with an earpiece fitted to a hollow stick and listens for the sound of running water. It’s laborious and not very accurate.

These days we also use electronic listening devices, like sensors and ground microphones. Sensors too ‘listen’ for sounds in the dead of night and, if a noise reaches a certain threshold, they send a recording of it back to the office. As more than 70 per cent of these sound files are not even leaks, each one needs to be carefully examined before sending an engineer.

The result: it’s laborious, not very accurate and also too expensive to cover an entire network. As well as buying and installing the kit, you need the human resource to manage the vast data it produces.

AI takes human error from leak detection

We need a step change that raises all networks to the level of the best, and fast. What’s more it needs to overcome the current barriers to technology adoption. These are things like capital expense, skills shortages and low existing digital maturity.

This is something the current generation of expensive proprietary fixed loggers are not going to deliver. So, how do we do it?

The answer is by using AI. With FIDO AI, you don’t need to waste your precious skilled talent listening to leak files. It eliminates false positives instantly, tells you leak size and will soon be able to identify lead pipes – all from clever analysis of an acoustic or kinetic file.

Being able to prioritise resources on your biggest leaks is one of the most effective ways to cut leak run time and save water. In one year, one of our clients halved the leak run time of all its leaks by doing just this.

Actionable insight from leakage data

Insight differs from information in that it’s actionable. Accurate AI data allows confident decision-making for better operations, performance and even future outcomes.

By tracking leaks, AI provides auditable metrics for performance reporting but, more importantly, gives insight into where leaks are degrading or recurring on pipes in real time. With this, utilities build targeted infrastructure maintenance plans based on certainty rather than educated guesswork.

FIDO AI is an example of this. It not only finds leaks to more than 92% accuracy by removing human error and false positives, it’s also the world’s first technology to give leak size from data analysis.

In fact, we take water leakage and day zero so seriously that we make our AI analysis available for use on all audio files from any 3rd party sensors, whether smart pipes, smart meters, loggers or hydrophones.

But we don’t need external sensors because FIDO includes sensors for free. This makes instant smart detection is available on any network.

These same FIDO sensors can also be used to correlate leak noises to give exact leak location with no technical input or human analysis. Or, you can move them between locations to sweep an entire area and generate reliable heatmaps of leakage points of interest as you go. Because FIDO’s leak size calculations are completely independent of flow data, there’s no need for district metering.

2022 must be the year we get serious about water leakage?

FIDO works because we built its unique algorithms on the cleanest of data foundations. It maps and tracks leaks throughout the leak process and self-verifies its accuracy, learning as it goes.

Having established a winning formula, 2022 will be the year we go mainstream. We’ll add even more world-first functionality and, even more critically, we’ll be making it easy for small companies and single-person teams to get access to our tech.  FIDO AI with free hardware will soon be available in small user bundles with no capital outlay.

This will further democratise leak detection. FIDO is sensor agnostic and works with any legacy acoustic estate, so every leak it finds anywhere in the world improves its performance for everyone who uses it.

By giving the millions of small teams equal access to smart leak detection we can all do our bit to save the world’s water one megalitre (or gallon) at a time.