World Mosquito Programme
The World Mosquito Programme (WMP) uses Microsoft Azure to model population areas and determine the best release sites for disease-fighting mosquitoes to stop the spread of deadly dengue fever.Learn about WMP
390 million people are infected with dengue fever every year. And with a 30-fold increase globally in the last 50 years, the problem isn’t going away. The World Mosquito Programme has a highly effective weapon in the fight against this deadly disease: Wolbachia. Wolbachia is a naturally occurring bacteria that, when introduced into a mosquito population, stops the spread of dengue. However, deciding on a release area for altered mosquitoes is a resource-intensive process that requires specialist GIS skills. Without proper models, researchers run the risk of missing human settlement areas and also releasing mosquitoes into protected areas.
The World Mosquito Programme uses Azure and machine learning to automate a key part of the process. First, they gather high-resolution satellite imagery. Images are then sent through a pipeline built on Microsoft Azure to train models, build footprints and create grids. The grids and maps are processed, and the pipeline outputs a new map with filters and masks included. These new maps allow WMP experts to easily identify areas of human settlement, view building areas and determine spatial relationships. A process that used to take a GIS specialist two to three weeks to complete is performed by the model in just one day. Wolbachia can now be introduced to populations more effectively, and at lower cost.
Machine learning helps communities fight disease globally
WMP’s machine learning model taps into data on human population densities, land use, industrial sites, weather and other variables. Satellite imagery can map out large urban areas with strategic and granular accuracy: the goal is to pinpoint multiple impactful release points within blocks of as little as 100 square metres. Geospatial AI enables researchers to deliver data analysis and visualisation in hours rather than weeks, accelerating their ability to help communities.
How mosquito spread modelling works
WMP gathers satellite imagery, building locations and population data. Data is sent to Azure Pipelines, where models identify human settlements, building footprints and estimated population density. Model outputs are combined in a simple app that WMP experts can use to help determine the best sites for releasing disease-fighting mosquitoes. Mosquitoes are released, and more data is collected.