The information health workers need is straightforward. Where do people live? How many people live there? Are those homes at risk for malaria? In rural Zambia, for instance, those are hard questions to answer. And for community health workers who need to know how many bed nets are needed or what areas to treat with insecticides, not knowing isn’t an option.
Those of us in the tech world get excited by neural networks and deep learning and cloud computing. And while these are important advances, we can’t forget that some problems still need human insight – at least, for now.
To help malaria elimination efforts, DigitalGlobe is tackling the foundational question of where people live with a unique, two-pronged crowdsourcing approach. Our goal is to produce digital footprints of every building across huge swaths of Southern Africa, Central America and Southeast Asia. Because this is such a vast area, setting a crowd of OpenStreetMap (OSM) volunteers loose on hundreds of thousands of square kilometers of imagery isn’t the most efficient approach. Plus, we want to maximize use of their time and deliver quick results. To do this, we use our crowdsourcing platform, Tomnod, as a way to narrow down our search area. Instead of immediately asking the crowd to trace buildings wherever they find them, we ask them to simply find the buildings – yes or no, is there a building in this image? This helps narrow down the area we present to the OSM crowd for tracing, in some cases by upwards of 90%. By focusing on imagery with buildings, the OSM campaigns can digitize those footprints quickly, thus mapping thousands of vulnerable communities in weeks.
And while the technical aspect of this work seems straightforward, convening the right set of partners is an altogether different challenge. The Bill & Melinda Gates Foundation funded this effort with a grant of $1.3 million dollars. DigitalGlobe, Mapbox and Humanitarian OpenStreetMap Team (HOT) are leading the mapping. And PATH and Clinton Health Access Initiative (CHAI) will put these data to use.
“Maps and geospatial data play a critical role in understanding and meeting the needs of populations most affected by malaria,” says Jeff Bernson, director of results management, measurement, and learning at PATH. “Mapping structures with this level of granularity will only further support the Zambian Ministry of Health as they mobilize their resources in their fight to eliminate malaria.”
The cost of malaria to African economies is huge, estimated at US$12 billion a year in direct costs. It consumes 25 percent of household income in high-burden areas due to lost productivity and health care–related expenses. Eliminating malaria would have potentially huge ramifications on these local economies, would free up limited health resources, and would reduce school absenteeism by nearly half.
Help us put these communities on the map. The most important partner in this malaria coalition is you. Without the crowd, we couldn’t enable some of the fantastic work that’s happening right now in Zambia, Guatemala, Laos and elsewhere. Think about that family that lives in the home you’re tracing. The family that needs a bed net to protect themselves from mosquitoes. That family whose children want to go to school and learn every day. That family whose livelihood depends upon being healthy enough to work their fields. These are the families we’re supporting, even from space, even from thousands of miles away.