Drones Create Real-Time Maps of How Diseases Spread

As tools for good and not evil, civilian drones are often most useful in environments that are otherwise hard to navigate. We've seen UAVs pitched for used humanitarian and environmental initiatives to help map and monitor remote areas. Now researchers are going one step further, and using that data to map the spread of disease.

A team led by Kimberly Fornace at the London School of Hygiene and Tropical Medicine published a paper in the journal Trends in Parasitology that explores how drones could be used in epidemiology. Their findings are based on a case study that used drones to map risk factors for malaria in Malaysia and the Philippines.

Fornace explained in a phone interview that collecting environmental data, whether by satellite image, remote sensors, or drone cameras, is particularly useful with diseases like malaria, because variables like rainfall and vegetation can effect mosquito habitats. Mosquitos, of course, are the prime malaria carriers.

In the study, the researchers also looked at macaque habitats, because the malarial parasite they tracked, Plasmodium knowlesi, was zoonotic—it can be transmitted between the monkeys and humans.

Tracking how mosquitos, monkeys, and humans are all moving in relation to each other therefore gives an insight into how the infection spreads.

"We started using a drone, actually, because the remote sensing data and the satellite data that we have access to is quite limited to the tropics," she told me. "The area where we work is changing really rapidly. It's changing kind of monthly; major changes."

Google Earth images were hopelessly out of date, showing forests where there were now clearings, and cloud coverage obfuscated satellite photos. So, they took a commercial senseFly eBee drone—which we've seen used in the region before—and collected their own data over 158 flights. The data from the onboard camera was able to put together detailed aerial images and 3D models.

The scientists could update the models whenever necessary, giving them what essentially amounted to a real-time appreciation of changing disease vectors. In the paper, the authors recall re-mapping an area of forest as it was cleared to make way for a rubber plantation. "This ability to map changes as they occur is critical for understanding how land-use change affects the distribution of human populations and disease vectors," they wrote.

Fornace told me that the key to their work as epidemiologists was to use the drone data in combination with ongoing studies. They recruit malaria patients in hospitals to find out more about their movements, and also track monkeys by GPS collars. "On top of all of that, we're also catching mosquitoes in the different environments we've been mapping by drone," she said. "It's kind of adding layers to all of this to get a better understanding of what's going on in that setting."

While that study was specific to malaria in one region, Fornace suggested drones could be useful to monitor the spread of other diseases. Mapping water bodies, for instance, could help study any disease carried by insects. Even non-infectious diseases, such as those caused by air pollution, could benefit from different drone-borne sensors.

Still, hobby drones have their limitations. Twenty percent of the researchers' flights didn't garner usable data, usually due to adverse weather conditions or because of the drone enthusiast's biggest nemesis: battery failure. The researchers also complained of extensive and often unclear regulations, which required them to apply for permissions that didn't always seem relevant to their work. Hobbyists can no doubt relate there, too.

The authors concluded that, "The use of UAVs is most appropriate when detailed maps of relatively small geographical areas are needed in areas where high-resolution satellite data are not readily available."

Fornace said satellite data is still better to cover larger areas. For now, her team is getting ready to work with their second drone, which will use a thermal camera to try to pick up populations of macaques hidden in the forest based on their heat signatures. It beats trying to collar every monkey in the area.