Neuroflux hacks away at water management problems
Finland’s Neuroflux is utilising big data and machine learning to solve problems in wastewater networks and enable water utilities to reduce the waste of resources.
The Helsinki-based data analytics specialist owes its existence to a hackathon organised in the spring of 2016 by the Helsinki Region Environmental Services Authority (HSY). The participants from Aalto University and the University of Helsinki were tasked with identifying new ways to utilise the amplitude of data generated in the wastewater network of the Finnish capital.
“None of us had any prior knowledge of water management, but we managed to put together a visualisation of multiple data sources over the weekend,” recounts Tuomas Koskinen, the chief executive of Neuroflux.
“HSY was impressed and started asking us if and how we could move forward with the development.”
The hackathon team were preoccupied with other projects at the time but continued the development on a consultancy basis a few months later. During the co-development project, the tool began to shift from data visualisation to data analytics, ultimately metamorphosing into a tool for identifying problems in the wastewater network.
“It generated interest among other water utilities and software vendors, and so we started thinking about how to seize what was starting to look like a market opportunity,” says Koskinen.
Neuroflux was established to that end roughly a year ago, with the mission of helping more utilities to reduce the waste of resources by providing analytic insights into water management. The demand for the tool became obvious as soon as the startup began talks with utilities about piloting it in a real-world environment.
“It was great that we got to trial the tool with several utilities from the get-go,” says Koskinen.
“It kept us grounded on solving real problems and gave us a solid foundation to develop a scalable service due to the various data systems and use cases of different utilities.”
Detecting anomalies, creating savings
The system fundamentally detects hydrodynamic anomalies in water networks. Koskinen highlights that this guarantees greater compatibility and convenience as utilities do not have to install new sensors or modify their data to comply with any specific standard but simply ensure their data is consistent.
“We can work with it as long as the network is continuously producing the same kind of data,” he explains.
It may not sound groundbreaking, but switching to the automated self-learning system offers substantial benefits for water utilities. The system is capable of detecting leaks and blockages before they cause any major damage, facilitating the optimisation of operations by providing real-time analyses and identifying external water sources by combining flow and weather data – melt- and rainwater make up around 40 per cent of water in wastewater systems in Finland, says Koskinen.
“All sorts of stuff end up in wastewater drains – hair, sand, grease, concrete and bicycles – and gradually the drains become clogged. There aren’t too many good tools around for detecting partial blockages,” he tells.
“We are able to detect them – usually before the water spills over, but this also requires quick action from the water utility.”
Water utilities have used the system to successfully detect a fair number of leaks and blockages, consequently saving hundreds of thousands of euros a year, according to Koskinen.
Nordics and hackathons a springboard for growth
The Nordics, he adds, have proven an excellent environment for developing the system due to both the abundance of engineering expertise and the fact that water utilities are willing to share data and provide input on how the problems should be tackled as they are not competing against each other.
The system has also been adopted in Hämeenlinna, Kirkkonummi and Porvoo. Installations are underway in Rovaniemi and Vihti and preliminary negotiations with a handful of municipalities in Finland and Sweden.
“But as a market, the Nordics are rather small,” acknowledges Koskinen. “Our plan is to start expanding overseas more systematically during the course of this and, especially, next year.”
“We’re recruiting more people to make sure we have a solid foundation for growth and we’re ready for the international market as soon as possible,” he adds. “Because if it’s not us who go there, someone else will.”
Neuroflux is confident it can claim a share of the emerging market also abroad, given the accuracy and ease-of-use of its system.
“Because our service has been developed together with water management engineers from the start, it’s specifically oriented to solving their problems. We’ve been working closely with our clients this entire time and will continue to do so also going forward. We want to understand the problems, the data and come up with the solution,” emphasises Koskinen.
The startup has already made an impression in, for example, the US. It was announced as the winner of a hackathon organised in conjunction with the annual Smart Water Networks (SWAN) Forum in Miami, Florida, in May 2019.
Who knows what kind of floodgates this will open.