Images courtesy of Microsoft

Global health threats are on the rise, due in part to the widening reach of the world’s deadliest creature…the mosquito. Mosquitoes spread deadly diseases such as dengue, Zika, and malaria just to name a few, and Project Premonition aims to give dengue outbreak prevention a jump start that we could have only dreamed of just a few years ago. Microsoft takes its own technology know-how to the fight against dengue and other global health threats.

Wouldn’t it be great if you had a heads-up that dengue is heading your way? Or that the dengue you’re currently fighting is changing tack?

That is exactly what Microsoft’s Project Promotion is aiming to give communities around the world. The Microsoft Research team shared their ambitions in a blog post on Microsoft’s site last year: “Project Premonition could eventually allow health officials to get a jump start on preventing outbreaks of a disease like dengue fever or avian flu before it occurs, whether or not it is a disease spread by mosquitoes.”

The three key elements to Project Premonition all leverage the computer giant’s latest technology advances:

1. Smart insect traps use machine learning to distinguish disease vectors from other flying insects.
2. Smart drones use machine vision and learning to identify mosquito hot spots where smart traps could be placed.
3. Genetic sequencing uses advanced analytics to take a closer look at any vectors caught in the traps.

The smart trap

Let’s start by taking a closer look at Microsoft’s traps. Unlike most traps that simply catch any insects passing through, Microsoft’s smart trap intelligently checks out what it has caught before deciding whether to keep hold of it or set it free.

The Microsoft research team reports in its post that the trap “is designed to only collect the type of mosquito an entomologist wants to track, instead of a hodgepodge of mosquitoes, flies, moths and other critters that scientists then need to manually sort through.”

How does it do that? By analyzing the wing movements of the little creature. If the insect looks like a ‘mosquito of interest’, the trap shuts its door.

That intelligence relies on some very advanced technologies: machine learning in particular. Machine learning is not only critical for the initial training each trap needs, it also helps ensure each trap continues to learn from its mistakes once deployed in the field.

Mosquito traps were deployed in Houston, Texas, last summer to gather the insects and the data needed for the training. Microsoft explains how, to recognize which mosquito it’s supposed to catch, the trap “needs what experts call training data: Lots and lots of examples of mosquitoes and other bugs flying into the traps. […] That data can then be used to build an algorithm that uses machine learning to help the trap learn to correctly identify the mosquito it should be capturing.”

Building a clearer picture

But the traps do more than simply catch potential vectors. They also record when the insect was trapped and the local weather conditions at that time, wirelessly downloading that information to the Microsoft cloud for scientists to then explore.

“That could help scientists understand the specifics of how a virus is spreading, such as what type of mosquito is infecting people and whether those mosquitoes are more likely to feed at night or when temperatures reach a certain point,” according to Microsoft.

The smart drone

In addition to the smart traps, Microsoft is also developing technology to make drones smarter. With these advances, Microsoft hopes drones will one day be able to autonomously (on their own without human input) place the smart traps in mosquito hotspots.

Again, these smart machines will rely on some of the latest technological advances: computer vision technology and machine learning.

Together these technologies could allow the drones to seek out remote areas where mosquitos are gathering. The team also hopes to train the drones to place the smart traps in suitable locations in these areas and pick them up again.

“Using drones instead of people [hiking for several hours] could be another way to save time and money,” Microsoft says.

The Asian tiger mosquito has adapted to colder climates, raising the risk of dengue spreading into new regions of the world.

Cutting-edge molecular biology

The third part of the solution is all about examining the mosquitos caught by the traps. Not just at the microscopic level, but right down to their genes.

A view on how the mosquitos, or even the pathogens, are evolving would be invaluable on so many levels. We’ve recently looked at how Aedes albopictus, also known as the Asian tiger mosquito has adapted to colder climates, raising the risk of dengue spreading into new regions of the world. Any warning on further changes in the vector’s DNA could be a crucial warning of its potential to survive in new environments.

On another level, visibility on any evolution in the dengue virus itself is going to be critical to programs developing therapeutic antivirals for treating the disease or vaccines for its prevention.

So, what elements of its technology know-how is Microsoft using for this analysis?

“The latest advances in both computer science and molecular biology are making it more practical for researchers to cull through the ‘soup’ of mosquitoes and the blood they fed on to figure out what viruses and microbes might be lurking there,” Microsoft reveals.

The ‘soup’ is being used to identify the species of mosquitoes collected, the animals they have bitten and the pathogens they may have encountered. New algorithms are needed to compare its contents with reference genomic (DNA) and protein databases to help scientist identify viruses and whether they are evolving. Microsoft recently shared news on a separate project aimed at accelerating genomic sequencing.

Global health threats at the crossroads of technology and society

It’s great to hear about how one of the technology giants is taking its own technology advances to the fight against dengue and other global health threats.

“Microsoft is driven to find innovative solutions to pressing societal problems that intersect with technology. With this project, we believe there is the opportunity to revolutionize disease monitoring in a safe and efficient manner. Our hope is that by combining custom-built traps with drones and the power of the cloud, we’ll be able to collect and analyse a plethora of previously unavailable data on a number of vector-borne diseases,” Ethan Jackson, a researcher in The Research in Software Engineering (RiSE) Group at Microsoft Research focusing on formal methods for safe cyber-physical systems (CPS) who is leading Project Premonition.

We’ll be keeping a close eye on the progress being made by this truly innovative project. Stay tuned…

—