Mechanics of Insect and Bird Flight

Bird and insect flight is still not fully understood. So scientists are now using video backpacks for eagles and virtual flight simulators for flies to work out exactly what'...
25 March 2007

Interview with 

Graham Taylor from the University of Oxford

Chris - Now it's time to talk to Graham Taylor, another Oxford University guru, on the science of flight. You actually take a slower and more sedate approach to flight than Jenny does, not travelling at Mach 6 obviously. So how do you study the science of flight?

Graham - Nothing close to Mach 6 but perhaps not sedate if you look at it close up. If you've ever looked at the way dragonflies are dancing over a pond or perhaps how a hoverfly hangs in a woodland in a shaft of light, then you must have wondered how they're able to control themselves so well. That's what I'm interested in understanding; both how they're so manoeuvrable and so stable at the same time.

Chris - Sounds easy when you listen to it but actually how do you manipulate and watch and study a tiny insect to see how it does these tiny things. Their wings are beating at 800 times a second some of them, aren't they?

Graham - Yeah that's right. And you've put your finger on the really difficult thing there, which is that they're tiny. In fact most insects are actually much smaller than those which you're familiar seeing, and that poses real problems with how you can go about studying them experimentally. So what we do is to pin them down. Now the problem with pinning an insect down and then trying to understand how it flies is you need to convince it that it's flying. And so what we've done is to build a virtual reality flight simulator, a bit like the thing you might have sat in at a science exhibition. We put the insect inside of that and it allows us to simulate exactly how it would be experiencing things as it's flying through the air.

Chris - So you're showing it pictures as though it were flying and then you're seeing how it changes in response to things you show it.

Graham - Yeah that's right. But that in itself poses more problems because an insect is, unlike us, able to see all the way around itself. So for a start you have to immerse it in a sphere that has video projection the whole way around it, so that poses problems. And on top of that, flies are also able to see extremely quickly. So whereas you and I are sitting in this studio here don't see the fluorescent lights as a problem, for a fly that was sitting on the wall it would see the lights flickering on and off. So we have to project patterns extremely fast as well. So we put the fly in the middle of this large sphere, project around the outside of it, and then it feels to the fly as though it really is flying through a visual environment.

Chris - And then how are you monitoring what the fly is doing exactly to work out how it's flying?

Graham - The fly itself is mounted on a little balance that's a bit like a complicated weighing scale, but it measures the forces and turning moments or torques, a bit like you'd apply to a tap to turn that on. So it measures all of those, which tells you how the insect would have been flying if it hadn't been superglued on top of this balance in the first place.

Chris - And do you then try and work out how its brain is responding or are you literally taking a step back and working out how the whole fly responds to visual stimuli? Because obviously understanding the nervous system correlates of how it controls flight must be quite important, because I know people are interested in working out if insects can do this, can we therefore make a better computer programme to control our planes and our artificial flying machines better.

Graham - That's the thing that we're most interested in getting at, so we look at both levels really; the overall black box level where you treat the insect as a black box that you don't know what's in. On top of that and at the same time as recording the forces the insect produces, we're also able to monitor what the nerves are doing. So by making recordings from its nerves, we're able to tell what signals are being sent to its brain and how it's using those to control itself. So yes, we're trying to get into that black box.

Chris - Well that's insects, but what about the bigger animals? Do they apply the same principles that an insect does? Does an eagle soaring around using exactly the same mechanisms as a gnat over a pond?

Graham - Well it's almost certainly not using the same mechanisms, but interestingly we actually know rather less about how birds fly than insects. It's not quite clear why this should be. They're bigger, which makes them a little bit easier to study but of course it does mean that you can't put them into a virtual reality flight simulator. So with the birds we do something quite different. What we've done is to make a backpack which we put on an eagle. He's called Cossack because he's a Steppe Eagle and comes from that part of the world. He flies around over the cliffs in Denmark and is actually coming over to Wales shortly. As he's soaring over the hill tops, what we're able to do with this backpack is to put a couple of miniature video cameras on and monitor where he's looking, what his wings are doing and what his tail is doing. These are radio signalled back to the little base station where we record this, so we get in-flight video at the same time as having the sort of instrumentation you'd get on an aircraft. So we know how fast he's accelerating, how quickly he's turning.

Chris - And when you actually do this, does it give you clues about how birds achieve something that's actually quite an enigma I think; how do they manage to fly without having a tail fin? All of the aeroplanes we've ever built have to have one or else they don't fly properly, but birds don't have one.

Graham - That's right and that for me is the sort of holy grail of this line of research: to understand how birds are able to make do without having a vertical tail fin. Now there are some aircraft which manage to do this.

Chris - Harriers and things?

Graham - They still have a vertical tail fin, a harrier, but there are one or two that manage it. And there what you have to do is have a very complicated control system where the computer controls the plane. So with something like the stealth fighter, the pilot's almost fooled into thinking that he's flying that aircraft, because really it's the computer that's controlling all of the detail.

Chris - So what are the birds doing?

Graham - Well we're trying to find out whether they're doing something similar; whether they're also making use of active controls as an alternative to having this vertical tail fin. Another thing that might be going on is that there might be something clever in the way the tail is shaped, so the twisted triangular tail has some unusual properties. Now if we could mimic those in an aircraft then you'd really be onto something quite interesting.

Chris - I can't believe that people haven't already. If you wanted to build a flying machine, surely the logical place to start is to borrow from biology and steal what nature has evolved over millions of years.

Graham - Well in many ways that is the logical place to start and it was the logical place. So early on in the history of flight people did look to birds and try to copy some of those mechanisms. If you look at the Wright flyer for example, unlike modern aircraft which have flaps on its wings and uses those to control itself, the Wright flyer had its wings twist and deform in a way that mimicked the way the Wrights saw buzzards over their house doing. So yes, that's where people started but in more recent years, for very good reasons actually, people have moved away from copying in a sort of slave-ish fashion how birds fly.

Chris - So the bottom line summary is that at the moment we still don't really know, but we're getting some quite good insights into the tricks they use.

Graham - That's right.

Chris - Are we any closer to actually building an artificial bird?

Graham - There are quite a few orni-thopters, which are flapping aircraft, little remote controlled ones that you can buy. You can get these things on the internet quite cheaply. So there one or two of these things around but as of yet they're not very operational in the way that people who want to use them would want them to be. 


Add a comment