Satellite monitoring for the predictive maintenance of sewage and water networks

I was flattered to see my interview featured in Mladen Stojic’s keynote at HxGN LIVE 2019. During his speech, Stojic, the President of Hexagon Geospatial division, demonstrated the power of connections and how they integrate our geospatial and operational worlds to trigger life-changing chain reactions. This is exactly the case with our customer, HERA Group, that was featured in Stojic’s keynote.

HERA is among Italy’s largest multi-utilities, working mainly in the environment (waste management), water (aqueduct, sewerage and purification) and energy (electricity, gas distribution and sales, energy services) sectors. In recent years, HERA had been searching for a method to better understand ground surface movements caused by structural defects in its collectors that could affect the area above the primary network and adjacent areas.

Today, we are monitoring thousands of kilometres in their network to detect changes in the Earth’s surface that could impact infrastructure. The results achieved by our monitoring service for HERA are outstanding, as we improved inspection efficiency by 40%. I am eager to see what the results of the next step will be, as we are now building an “optimized” monitoring system to enable predictive maintenance of the water and sewer networks.

Actionable knowledge from satellite monitoring

We started with HERA two years ago, trying to respond to our customer’s problem – a problem common to many utilities – which is to reduce the costs related to the maintenance of water and sewage networks. Traditionally these companies intervene either following a collapse (when the damage has already occurred), or they do sample inspections. We proposed an approach to use radar interferometry to identify ground motion that worked well. Thanks to the continuous data acquisitions made by the Sentinel satellites of Copernicus, we have highlighted areas that present instabilities that can be either the cause or the consequence of problems underground. We have thus identified numerous problems, monitoring over 6000 km of pipelines in an area of ​​3,500 square kilometres.

Last year, we decided to take it a step further. Using Hexagon technology we built Rheticus® Network Alert, a Smart M.App web application that combines the results of time series InSAR analysis with the sewage or water network graph. Thanks to sophisticated data fusion algorithms and geocoding, Network Alert returns an always-updated list of those features of the network that may present problems. This way we have simplified the use of complex information such as that derived from satellite radar, providing actionable knowledge to HERA. We’re helping them address field inspections in order of priority, instead of responding to a problem or following the traditional planning.

 Predictive analytics for preventive maintenance

This year we are even moving beyond integrated data visualization. To further improve the reliability of the information provided each month by Rheticus® Network Alert, we combine in M.App Enterprise the analytics derived from the SAR analysis with many other data sources: citizens’ reports, traffic flows, characteristics of the network and other ancillary data. This way, the identification of the sections at risk becomes even more precise, and that enables the preventive maintenance of the sewage network. With this approach, we helped HERA Group identify as many as 13 critical situations out of a total of 23 inspections, and again, we have increased monitoring effectiveness by 40%. This is a very important result compared to the traditional approach, which is to carry out inspections by planning them on a purely statistical basis or intervening only after the problem has occurred.

We are now continuing to improve the integration and analysis of this big data using predictive techniques based on artificial intelligence. These techniques rely on the recognition, from the combination of different data sources, of recurring patterns that allow us to predict where a problem in the network is most likely to occur. This is the way we build predictive analytics, to help our customers who manage strategic assets solve problems before they occur.

Massimo Zotti

Head of the Strategic Business Unit "Government & Security" of Planetek Italia, Zotti is responsible for the marketing of application solutions and services for public entities and Defence at the national and international level. Massimo Zotti is a passionate promoter of issues related to Geo-ICT and Earth observation. In more than 20 years of professional experience, he gained extensive expertise in geospatial technologies, remote sensing, client-server and web-oriented architectures, geospatial standards for data interoperability and open policies, notably in Earth observation.