Facebook’s progressive AI project will achieve a result, previously thought impossible by most researchers in data storytelling. The Social Media giant will deploy its advanced AI capabilities to map the population density of the entire world. The project took its roots back in 2016 when Facebook managed to create in-depth maps of 22 countries, — the latest one being the map of Africa.
Creating maps like these is not an easy job for human beings. Although, modern times have compelled science to invent extra-ordinary satellites that provide very high-resolution imagery, converting these images into meaningful information is a very tedious and time-consuming task. Consider this very project as an example — the procured images will need to be manually tagged and then cross-referenced with the census data available which can be tricky to get an accurate number.
What Facebook is doing is automating this task of data visualization using AI and machine learning. Facebook’s data analysts and engineers leveraged geospatial data from an open-source mapping project (Open Street Map). Then, it applied Computer Vision to the imagery delivered by the satellites. Then, they just removed open chunks of land from the satellite images. With faster systems as compared to 2016, Facebook mapped Africa’s population density by using 11.5 billion images and validated their work with Columbia University’s Center for International Earth Science Information Network (CIESIN).
Facebook says that it will be releasing the maps for free and for all. This discovery will be a big boon for humanitarian purposes such as natural calamities or man-made hazards. It will help relief teams get to the affected area faster irrespective of their resource volumes. The American Red Cross has in the past, used Facebook’s data for such initiatives.
Although, the coming years will gradually reveal how Facebook monetizes this initiative — or if it doesn’t.