This article was originally published by Guardian Newspapers Trinidad & Tobago on 23 August 2020.
For the past 4,000 years, humanity has been looking at various methods for supplying potable water (freshwater resources) and taking away its waste (typically wastewater). Today, intricate networks of pipes, pumping stations, valves, connectors and water treatment plants convey clean running water to approximately 71 per cent of the global population.
However, managing it sustainably is a complex and often labour-intensive task. With the global population on the rise, there’s a greater demand on potable water supplies than ever before and additional pressures from communities across the world to boost efficiency. As water infrastructure ages, the world’s utility companies face the never-ending challenge of identifying failing equipment and repairing it before leaks cause major damage. Simultaneously, issues such as ground movement, changes in land use, development and even vegetation growth due to poor maintenance practices all threaten the security and sustainability of water supplies.
The population across Trinidad & Tobago (T&T) often ask significant questions about the Water and Sewerage Authority’s (WASA) capabilities and competencies in managing its ageing infrastructure and maintenance, which is traditionally a ‘boots-on-the-ground’ affair, involving slow and inefficient manual checks of the water distribution system including pipelines, valves, connectors and pumping stations.
In WASA’s never-ending quest to eliminate leaks and get the most from the Government of T&T (GORTT), investment, the use of satellite imagery and machine learning is needed to optimise the spending on critical infrastructure. The GORTT needs to be far more proactive, and requires a more intelligent, data-driven solution when analysing and giving an overview of the country’s water networks and the current and future problems within the sector.
Traditionally, water utility companies such as WASA have used ground-based surveying to identify and repair issues before they become serious. It is not exactly a high-tech approach in terms of leakage detection. These are simple, “boots-on-the-ground” solutions. In other countries, there are typically a combination of various engineers and technicians trying to listen for leaks with acoustic loggers and performing surveys on larger diameter pipes, where both engineers and technicians manually check for any visible places where leakage could have occurred. But, while these techniques can help find nascent problems, they can’t predict where on the water network they’re most likely to occur. Utility companies also have a limited ability to anticipate environmental stressors that could negatively impact water quality.
The use of satellites and geospatial analysis can help to pinpoint leaks with satellite imagery and big data analytics. This offers a radically different way of looking at the problems of leakage and maintenance. Applications of a machine-learning platform, to combine and analyse high-quality satellite images and other sensor data, are unique analytics solutions where geospatial data analytics comes into play. Such systems can be utilised to vastly improve the water engineer’s understanding of catchment areas and the water quality within them, along with supplying greater levels of intelligence on pipeline degradation and potential points of failure. Artificial Intelligence (AI)-led analysis, for example, can consider precise land movement indicators picked up from satellite images, weighed against factors such as the age of the pipelines, diameter and flow pressures. Vegetation growth and maintenance surrounding the pipeline networks can be identified and carefully factored in, assisting in predicting and avoiding the damage that root incursions causes to drains and sewers. Considering these factors and other inputs, satellites and geospatial analysis can show water and wastewater utilities where they should deploy acoustic loggers and other surveying techniques, helping them anticipate failures and tackle them before they happen.
The use of historic incidents to train models can lead to a greater understanding of the conditions that caused previous failures to occur. This enables the engineering and water modelling applications to continually generate data on the likelihood of future failure. Proactive and targeted leak detection is only one part of the big data solution for water utilities such as WASA. Geospatial analysis can guide other maintenance and investment decisions too, helping inform inspection and renewal programmes.
With movements in the dams/reservoirs (including evapotranspiration rates) and embankments often providing crucial early warning of major failures, water data can play a key role in highlighting the need for urgent inspection, lowering the risk of significant disruption, which continues to occur far too frequently across T&T.
Hydrosystems modelling with satellite data can also inform strategic and business-critical decisions relating to water management. For example, land-use analysis across a water catchment area can highlight risks from soil erosion and even help predict the likely load entering the system. Geospatial analysis can identify crops with up to 95 per cent accuracy, providing unprecedented seasonal information on factors such as run-off that may affect water quality.
The view from satellites in space even has quite surprising benefits at ground level and below it. By weighing local topography and development, water engineers can also model the consequences of failure. This can, for example, help water utility companies such as WASA prioritise at-risk water pipelines if they run near to other important infrastructure such as power lines, roads, highways and nearby buildings.
Data-intensive and a heavily reliant machine-learning techniques with a geospatial analytics platform is an example of a solution which can become viable today due to powerful, secure and scalable computing techniques. Such applications can prove instrumental in delivering a consistent, cost-effective and less wasteful water service to T&T.