With digital technology transforming the way water companies manage their networks, Ovarro’s data innovation subject matter experts Alan Cunningham and George Heywood discuss an advanced solution that is helping the sector address key environmental and operational challenges.
Ovarro’s long-term partnerships with water companies has enabled a deep understanding of the issues facing the sector day-to-day. What are the biggest challenges?
George Heywood: The water industry is facing unprecedented challenges. Firstly, there are increased expectations from regulators, customers and the public to improve environmental performance. There is currently particular focus on the quality of the UK’s bathing waters and rivers, and water companies can expect to be held to account for not fulfilling their environmental obligations.
The second is the perennial challenge of funding constraints, particularly in heavily regulated water industries like we have in the UK. I see digital transformation as the third challenge. The tricky part is for the water industry to keep up with the pace of change – but a huge opportunity lies in utilising new technologies to solve some of the difficulties.
As an example, one of our newest products, BurstDetect, is a cloud-based, early warning system that captures existing pumping station data to detect rising main sewer bursts.
What can water companies gain from these technological advances?
Alan Cunningham: What water companies have told us they’ve been data-rich and information-poor. They’ve been collecting lots of information from their networks through logging devices and telemetry units, but they haven't had a way of bringing that data together to look at the big picture.
Through digital transformation, companies can now start interpreting this data, understanding in more detail what's going on in their systems.
How can your software help water companies reduce sewer pollution?
Alan Cunningham: BurstDetect was developed in response to a water company customer coming to us with the problem of bursts from rising mains. Because these are pressurised pipes, bursts can cause severe pollution incidents. Often, they are not detected by current monitoring systems and the company is alerted by a member of the public – perhaps a dog walker. Water companies do not want to find out about bursts that way – they want to find out before any damage to the environment occurs.
We worked on developing a machine-learning algorithm that would look at pump on-off signals, and pressure and flow data, where available, and use those to detect unusual situations that could be a rising main burst and send an alert to the customer.
Are software-as-a-service (SaaS) applications part of this portfolio?
George Heywood: Yes, BurstDetect is a SaaS – or cloud-based – analytics solution. That means the analytical software runs on servers in the cloud and can be accessed via any device with an internet connection.
This software is usually licensed on a monthly or annual basis rather than requiring a one-off investment. There are several advantages over more traditional software – firstly, there is no large initial capital cost, either for licensing or for purchasing hardware. Secondly, it’s easier for cloud-based software to be set-up and maintained.
Another advantage to solutions that use machine-learning and artificial intelligence techniques is the algorithms ‘learn’ what normal looks like for a particular set of assets - you don't have to do an asset-specific configuration.
Perhaps the most powerful advantage is that, as the size of the dataset grows, the quality of the results that are produced can also grow. These machine-learning algorithms feed off data, so the more data captured, the larger the datasets that get assembled, which means better results will come out of the solutions as time goes on.
To find out more about BurstDetect, check out our recent webinar "Preventing Pollution from Rising Mains through Machine Learning".