How this startup battles water waste with leak-detection algorithms

We are probably all guilty of using too much water. However, most water wastage happens prior to the precious resource entering our homes. That is because national pipe networks are riddled with leaks — and water scarcity is a growing global problem. Could AI be part of the solution?

For instance, the UK wastes more than three billion litres of perfectly good quality drinking water every day. Meanwhile, Italy and its ageing water infrastructure loses 42% of drinking water yearly. Globally, daily water loss is estimated to be 45 billion litres. 

HULO.ai, a startup from Leuwarden, the Netherlands, is looking to reduce water loss by connecting smart sensors with its advanced AI algorithm. Essentially, the cloud-based solution can optimise sensor positioning and detect and localise water leakages within water distribution networks in real-time. 

“If a leak is not being detected in real-time, it could be there for a long time, so you lose a lot of water. That’s one part,” Robbert Lodewijks, HULO.ai’s co-founder and CEO, tells TNW. 

The <3 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!

“Another part is that sometimes when a leak is happening, and you cannot find it soon, the damage that the leak can do gets bigger over time,” he continues, referring to potential sinkholes, railways washing away, or hospitals being flooded — to the detriment of both finances and reputation of water utility providers. 

Improving leak detection accuracy

Essentially, the technology minimises the need for extensive physical infrastructure checks. Normally, the understanding and control of pressure, flow, and leakage throughout a water distribution system is done through what are known as DMAs (District Meter Areas). By utilising an approach of closing certain valves and setting up flowmeters, the utilities’ company can determine how much water is lost before it reaches the consumer. 

While water can also get “lost” due to non-metered consumption, fire hydrants, or metre inaccuracies, approximately 90% of all water loss comes from leaks. HULO.ai’s technology narrows down the highest potential area of a leak without the need for a DMA. 

“In a network of around 20 x 10 kilometres, we could narrow it down to areas of 250 square metres, where we could say ‘Hey, this area of 250 metres is where the leak is happening,” Lodewijks says.  

Additionally, HULO.ai says it can tailor its solutions to the specific client and their needs.

Expanding water loss tech to the world

The startup — a spin-out from the Wetsus sustainable water technology research institute — just raised €800,000 from VC firm LUMO Labs and the Netherlands Enabling Water Technology (NEW) Fund. The money will go towards enhancing the company’s cloud-based solution and making it highly scalable and adaptable to different market needs around the world, including those were water scarcity is the most critical. 

“The growing global demand for sustainable water management solutions is as urgent as it is important,” Andy Lürling, Founding Partner of LUMO Labs commented. “We are drawn by HULO.ai’s potential to lead and transform the water management industry on a global scale with its cutting-edge AI technology.”

While the recent water shortage situations in Cape Town and Mexico City may have drawn significant media attention, the truth is that the water crisis is rapidly escalating across the entire globe. Between two and three billion people worldwide already experience water shortages. This situation is set to worsen significantly over the coming decades — especially in cities. 

Sensor quality matters

The arrival of smart sensors has the potential to drastically reduce both individual water consumption and collective waste. The HULO.ai algorithm is sensor agnostic, meaning it works with any pressure and flow sensor, and as such is not dependent on access to any one supplier. 

However, that also means that the speed with which the algorithm can respond will depend on the quality of the sensor. “Some utilities have sensors that send a measurement every second and some have data every minute or every 15 minutes,” Lodewijks adds. 

“And that also determines how quickly you can react to a leak. So if we have data that comes in every minute, we can tell you if a leak occurred in the last minute. If there’s data every 15 minutes then, naturally, the time period is longer.” 

The increasing threat of water scarcity

In the UN World Water Development Report from this year, UNESCO and UN Water warn that the global urban population facing water scarcity is projected to double from 930 million in 2016 to 1.7-2.4 billion people in 2050. That is, unless we effectively deploy water saving technology and large-scale international cooperation.

With greater water scarcity looming on the horizon for future generations, every drop literally counts. Conceptualised over a number of Friday drinks while its founders worked for another company detecting failure mechanisms of pipelines, HULO.ai’s raison d’etre is quite clear. Lodewijks states: 

“Quite soon, we realised that actually our goal is to be 60 years old and say to our children that this technology, our algorithms, have helped water utilities all over the world, and we reach places where battling water scarcity is really a big problem. And then, the adventure is already a big success.”