Poor water management can increase maintenance costs and lead to unhappy customers, but software can help to keep problems at bay.
The UK has a water management problem. In the twelve months from April 2017 to March 2018, the amount of water lost to leakages – 3,183 million litres or 1,273 Olympic swimming pools – rose for the second year in a row. What’s more, nine of the private companies that supply water in England and Wales failed to meet their targets. This resulted in the regulator Ofwat slapping Thames Water with a £120 million fine to compensate customers for the company’s poor management.
The situation is exacerbated by the fact that finding leaks is complex because most are not visible on the surface, and reporting them can be time-consuming. Then fixing them causes disruption, as it usually means water supplies have to be turned off and, when in a residential area, roads and pavements dug up.
There’s also the fact that the tools used to finding leaks are low-tech and, arguably, old-fashioned. The main way engineers detect leaks is by listening for them using metal or wooden sticks. When a leak occurs it causes the pipe and surrounding environment to vibrate, which can be picked up using a stick fixed with an ear-trumpet.
Realising that this is an outdated way of doing things, researchers and technology companies have been exploring how to bring water leak detection into the 21st century. For example, the WADI project have been exploring how unmanned aerial vehicles mounted with remote sensing cameras can be used to carry out rapid inspection of pipes. Images can be taken in the visible, shortwave infrared and thermal infrared spectral bands. The images can then be used to look for any changes in environmental conditions, such as soil moisture and ground temperature.
Other technology that has been implemented includes smart, wireless sensors. In July 2017, Australian water company SA Water installed a network of more than 400 sensors spread across Adelaide city centre. The network enabled the company to identify and repair potential problems, preventing around 15 water leaks in the first 12 months after being installed.
“The advantage of a network of sensors is that you’ve got all of this rich data being generated,” says Dr Mike Williams, chief technology officer at Inflowmatix, a Southampton-based firm that provides water flow and pipe health analytics to water utilities worldwide. “But, of course, you need to be able to turn this data into actionable insight and this requires software.”
Unlike relying on listening sticks and then fixing a leak or burst pipe after it has happened – some would consider this a sticking plaster – a combination of smart sensors and software allows for preventative measures to be taken, and quickly at that.
Inflowmatix recently carried out high-frequency pressure monitoring at a Severn Trent Water pumping station, which had been showing significant variations in water pressure. As a result of improvement works to the pumps, there has been a 70 per cent reduction in the burst rate and savings equivalent to £60,000 per annum.
As well as cutting maintenance and repair costs and boosting operational efficiency, using software helps to address problems swiftly, thus reducing downtime and avoiding disruption, also benefits customers. It enables water companies to provide them with a safer, resilient, and more efficient service.
At the end of the day, customers want a steady flow of water. If this flow is constantly being disrupted by engineering works and poor water management, then customers will be inclined to switch providers.
For people who live in blocks of flats, making that switch can be difficult. It’s generally a property management company or a residents’ management committee that makes utility decisions, rather than individual tenants, freeholders and leaseholders.
To keep several dozen residents happy all at once, property management companies and residents’ management committees need to be able to report issues with their buildings in real-time.
This is where software can help to escalate the reporting of problems, says James McGivern, chief technology officer at Shepherd. The start-up has built an analytics platform that can collate and analyse data from multiple sources, including building management systems and IoT devices, such as pressure and temperature sensors.
When a residential building is fitted with smart technology that is connected to a software platform, the volume of rich data being generated can be analysed using machine learning and artificial intelligence.
“If the software detects a possible anomaly, such as a breach of the water level over a period of time, then an alert is raised based on the severity of the expected problem,” explains Mr McGivern.
He continues: “These alerts can be sent out via email or SMS, alerting more people, more urgently. If the [property management company] is dealing with the problem, then the alert can be closed. It’s also possible to indicate [to the software] that it’s a false positive and shouldn’t have been identified as an anomaly. The software system can then learn to adapt to behaviours.”
The more software learns, the fewer false positives that are raised and the better the system becomes at identifying the critical issues and escalating problems so that they’re dealt with quickly and efficiently. This keeps the water flowing and customers happy.
In-car assistants: The Voice
To enable comments sign up for a Disqus account and enter your Disqus shortname in the Articulate node settings.