ANTHONY LEONARDO, SYSTEMS NEUROSCIENTIST, HOWARD HUGHES MEDICAL INSTITUTE - JANELIA FARM RESEARCH CAMPUS:

So this room we're in is this sort of indoor flight arena we have built to study how dragonflies forage and catch prey. Basically we've tried to do is craft and environment that is sufficiently complex that dragonflies will actually be happy living in here. They'll think it's the real world. They'll forage and catch things the way they do outside but simple enough that we actually measure what they are doing with very high reproducibility without being subjected to this variability of temperature and lighting and air and climate outside.

These obviously are beautiful insects. And if you watch them for a few minutes, it's sort of hypnotic to see the incredible agility with which they behave. But like fruit flies and other insects, their brain has many fewer neurons than a mammal so there's fewer things to measure. On the other hand, they are very large for an insect and so it is possible to load them up with sensors and actually measure them while they are behaving. And that is very difficult to do with a small insect.

So why are trying record from neurons from these dragonflies while they are flying? We like to look at the individual nerve cells. These are the nerve cells, or neurons as we call them, and think about how they steer the animals. So a simple example would be playing tennis where your simultaneously doing two things: you're looking at something moving through the world, say a ball for example, and you're predicting where it's going and you're trying to steer your body to intercept the ball at some future location.

JASON OSBORNE, INSTRUMENT & SYSTEMS DESIGNER, HOWARD HUGHES MEDICAL INSTITUTE - JANELIA FARM RESEARCH CAMPUS:

You have these really small components for this backpack, this tiny little device, and you have to hold these things. You have to manipulate these little tiny circuit boards and antennae and circuitry and what have you. And you're actually pushing the limits of engineering to a degree and actually the materials themselves.

Try to think of a backpacking manufacturer that's making backpacks for mountaineer climbers. A climber is going up hill, he's climbing Mount Everest and he or she does not want to feel the weight of this backpack. And the ergonomics have to be just right. It's the same thing with these animals. With these backpacks systems, we're actually splitting it down to the milligrams. They have to behave in a way that you can actually get really scientific results. If the payload is too heavy or it's awkward for the animal, you obviously aren't going to get the results, recordings or data that you are looking for.

ANTHONY LEONARDO, SYSTEMS NEUROSCIENTIST, HOWARD HUGHES MEDICAL INSTITUTE - JANELIA FARM RESEARCH CAMPUS:

What are the applications of this? I mean, if you want to repair an injured brain, the best way to do that is to understand how it actually works, right? And ultimately I think research like this that really explains what these circuits do is going to be the future on how you want to use that technology: how would you control the neurons to actually do it. Let them do something in a goal-directed manner and move their own legs. And so the best way thing about these types of projects is just fundamental research into what the brain is doing to solve complex behavioral problems.