WADI Project heading to the next step

Data sets from the recent airborne trials in the WADI project have now been processed, resulting in indicators that are believed to reveal the presence of underground water leaks. It marks a significant step in WADI project’s goal of developing an airborne water leak detection system.

Funded as part of the European Horizon 2020 programme, the WADI Project is coordinated by youris.com and EEIG, and is being developed in cooperation with 11 partners in six European countries – Belgium, France, Portugal, Spain, Poland and Italy. The aim of the project is to prove the feasibility of airborne water leak detection for rural areas.

With as much as 50% of water resources being lost to leakage in some parts of Europe, the WADI project hopes to contribute to a more water and energy efficient society.

Flight trials were conducted over 4 months between April and July 2017, with ONERA’s aerial platform BUSARD fitted out with two hyperspectral cameras and an uncooled infrared camera coupled with ground truth measurements. The trials saw the aerial platform fly over several areas of the water network infrastructure of Société du Canal de Provence (SCP). The areas chosen for the tests all had a high potential of soil moisture due to artificial water leaks.

The results of the test are a detailed image database of different weather conditions, types of soil and vegetation, and rates of water leaks.

Now, WADI has announced that the data sets have been pre-processed (geometrics and radiometric correction, registration), allowing the optimised wavelengths for leakage detection to be determined. The outcome is a series of maps of different indicators believed to reveal the presence of water.

It’s been determined that the easiest way to create a water index with the triangle/trapezoid method is to combine thermal infrared signal (8-12μm) with OSAVI indice (one wavelength on each side of the red edge, between 0.65μm and 0.675μm and between 0.79μm and 0.87μm). Other methods were not chosen because of their low performance in detecting water leaks.

Over the coming weeks, the WADI project will select the most promising cameras for use in the project.