How Bird Flocks Resemble Liquid Helium

sciencehabit (1205606) writes "A flock of starlings flies as one, a spectacular display in which each bird flits about as if in a well-choreographed dance. Everyone seems to know exactly when and where to turn. Now, for the first time, researchers have measured how that knowledge moves through the flock—a behavior that mirrors certain quantum phenomena of liquid helium. Some of the more interesting findings: Tracking data showed that the message for a flock to turn started from a handful of birds and swept through the flock at a constant speed between 20 and 40 meters per second. That means that for a group of 400 birds, it takes just a little more than a half-second for the whole flock to turn."

Comments

My message came first

[Anonymous Coward]

My message came faster than the rest of the flocks

There have been attempts before

[Artifakt]

One factor not mentioned in the summary, is that bad computer models for flocking can still generate what looks like realistic flocking behavior. The herd dinos in Jurrassic Park are an example of this - the animation formula assumed each dino was instantaniously aware of all the rest, without allowing time for their nervous systems to work, but the flocking motions still looked right to most people, including professionals. People should remember too, humans probably have some pretty good mechanisms built into their brains for analyzing flocking, so that our ancestors, going at least as far back as the ape-like ones, could successfully hunt birds in flocks, and we collectively and historically certainly have had a lot of practice at that. We, as a species, ought to have some skill at detecting what constitutes real flocking behavior, but if we do, it doesn't always make a bad formula look jarring or wrong. So when somebody claims they have a real formula for what's going on when birds and such flock, the next question is "Can this claim even be proven or disproven?"

Re:There have been attempts before

[TapeCutter]

Can this claim even be proven or disproven?

Silly question on a nerd site, you don't "prove" anything with science, and Jurassic park was a movie, not a scientific model.

Back then the short cut they took probably saved them weeks in rendering time, and as you say, came out looking realistic. A scientific simulation would be comparing real data points to the output, it would be able to identify the "handful of leaders" that initiate each manoeuvre of a real flock, it would definitely not be a bunch of lab coats looking at the pretty pictures and nodding.

Disclaimer: I like Crichton's stories too, but he tends to write in "false document" style and every story has the same "science gone mad" plot.

Re:There have been attempts before

[newcastlejon]

Back then the short cut they took probably saved them weeks in rendering time, and as you say, came out looking realistic.

Why is that? There's no reason that I can think of why one couldn't just decide how the creatures would flock using simple stick figures then add the rest of the models later.

In any case, we're in no position to judge how accurately a film recreated the behaviours of creatures that haven't been found in the wild for millions of years. Certainly we can infer a lot based on what we can observe in their distant descendants but it's still one of those things that takes some dramatic license (just like Lego genetics and the noise that a roaring T-Rex makes).

Re:There have been attempts before

[OzPeter]

Can this claim even be proven or disproven?

Silly question on a nerd site, you don't "prove" anything with science, and Jurassic park was a movie, not a scientific model.

Years and years ago I saw some academic research that modeled bird flocking with a simple "Try and keep a constant distance from my neighbors" algorithm. The video (vector graphics with the birds rendered as simple triangles) of the animations produced a very lifelike behavior of a flock of birds flying around and through groups of fixed objects. I'd say if anything that the animators of Jurassic park were probably aware of such techniques.

Re:There have been attempts before

[RaceProUK]

Wouldn't it be funny if it turned out the T-Rex actually barked like a Yorkshire terrier?

Re:There have been attempts before

[justthinkit]

a simple "Try and keep a constant distance from my neighbors" algorithm

Probably the same algorithm birds use when they fly into a tree -- "try to keep away from branches". They just do it ten or one hundred times faster than we can, so it be black magic to us. A tight loop, run with highly priority, and featuring a few key bits of inline code.

Re:There have been attempts before

[Anonymous Coward]

From my casual reading of the article and my faint recollection of diff eq and control systems, it sounds to me like they just added one more derivative term to the differential equation of each bird's motion.

old model: "If a bird angles right, the ones next to it will turn to stay aligned."
new model: "The team proposes that instead of copying the direction in which a neighbor flies, a bird copies how sharply a neighbor turns."

So I'm sitting here imagining the original model as conceptually some big nasty n^2 matrix for all the xyz position/velocity/rotation terms, and these guys came along and added more rows and columns to account for angular acceleration. Then they plugged in the measured boundary conditions, and they ended up with an expression that looks just like some other problem.

And to this I say: I'm no longer able to do the math in question, but it does not surprise me at all. A couple of decades ago my PDE teacher made us solve the heat equation a dozen ways from sunday, because it was mathematically identical to dozens of other problems. That, and because it was one of the only problems that they actually know how to solve analytically. :D

Re:There have been attempts before

[JustOK]

How would a T-Rex know what a Yorkshire terrier sounded like?

Re:There have been attempts before

[Kielistic]

The model you are talking about is called boids. It is a relatively simple AI model that demonstrates emergent behaviour.

Re:There have been attempts before

[gurps_npc]

Ever hear of the "jinx" tradition? It is when you say something at the exact same time you have to do "x". X is irrelevant (sorry math teacher). The point is that if two or more creatures are intent on doing the exact same thing - say eating grass while looking out for predators, they are very likely to spot something strange at the exact same time, both reacting at the exact same time. It is not telepathy, because the creatures are not reacting to each other.

The same thing often happens in flocks. All the creatures sense and react in the exact same way (herd animals are not noted for their individuality), at the exact same time. Even if the herd is large, they see, or hear/

Now, if they smell the creature, or if the herd is so large that some creatures can not see or hear the stimulus, then the herd starts to react to itself. But quite a lot of the time, an entire herd will become aware of the stimulus at the exact same time and react at the exact same time.

Re:There have been attempts before

[GTRacer]

I almost, *almost*, spit out a mouthful of my breakfast. Kudos! You've brightened my otherwise off-to-a-horrible-start Monday!

Re:There have been attempts before

[Anonymous Coward]

Can this claim even be proven or disproven

Perhaps a bird neurologists could prove an existence of a neural short circuit, or that the birds observe every boundary member of the flock they can see simultaneously.

Re:There have been attempts before

[140Mandak262Jamuna]

Bertram Wooster, is that you? Slowing stirring from the bed sipping tea and eating breakfast at the ungodly hour of 10:45 AM? How is Jeeves, old chap, Can I borrow him, I, no, no not I, my friend is batting on a sticky wicket, say. Toot. toots.

Re:There have been attempts before

[Artifakt]

Any hypothesis that doesn''t allow being disproven isn't science. period. That's hardly silly to point out. I may have been too polite by phrasing it in basic English - maybe I should have jumped right on a bunch of working scientists with the bold claim they had departed fully from the basic scientific method, before actually taking the time to read the original paper in detail and recrunching all their numbers, if that would make you feel better. Better yet, why don't you take "Let's You and Him Fight" elsewhere? I'm raising the question of whether the researchers took something into account, not accusing them of not understanding falsifiability as a fundamental of science, and if you want to turn a legitimate question into an accusation that insults both them, and me by the implication I would make it without doing a lot more work than could be done in the few hour since this article was posted, why don't you make that extraordinary claim, and sign your real name to it. A letter to the journal that published the original paer is appropriate there, not discussion in a non-vetted online "news" source. So I didn't spell out that I thought there were implications for falsifiabilty like I was lecturing the thinking impaired, particularly when I would much rather hear just what the paper's creators think are possible tests rather than assume they just didn't think about it.

            This also isn't a question of either whether Jurrassic Park got something scientifically right or whether Michael Crichton was a good author. That was just an example many readers would recognize. I could have used examples they wouldn't have even seen before, but I picked one they might know.

            Tell me, when somebody says there's hugh potential trouble in the nation's underfunded infrastructure, and mentions, as just one example, how many truck drivers are putting in excess hours and falsifying logs, does that make the whole article, in your mind, about trucker's bad penmanship? The real questions (now pay attention this time) are firstly "Do humans have a blind spot in the way they percieve flocking, even though there's 'logical' arguments why they should not, and we aren't bothering to look for evidence of a blind spot because those arguments make it so easy to ignore?", and secondly "Is an experimental model of flocking only going to be scientific if the researchers first make sure they have accounted for that blind spot?" My argument is that both questions need to be answered yes. Since that's my opinion, I'd also argue that a good mathematical model that ignores this, vrs. a bad mathematical model that just knowingly fakes flocking well enough, becomes like a better Planetary Epicycle model vrs. a worse one or even a deliberately false one. It doesn't matter much if the planets don't move in epicycles at all.

          I'd also say it's vitally important to figure out why the human brain seems to have many such blind spots - for just one, watch all the people, on all sides of the debate on the Theory of Evolution, who keep slipping into talking about what "Nature's Goals and Intentions" are. That's either because English (and at least most other languages) has/have a lot of superstitious cruft built in and we need to work at improving that or we will never be able to communicate properly, or it's something more fundamental to the human brain, and if it is the latter, figuring it out is probably going to be the biggest scientific achievement of whatever century it happens.

Re:There have been attempts before

[Anonymous Coward]

I can find no info on the 'jinx' tradition.

Re:There have been attempts before

[gurps_npc]

See here: https://en.wikipedia.org/wiki/... and here: http://www.lawoftheplayground....

Re:There have been attempts before

[Anonymous Coward]

from the noise it made just before becoming a pre-dinner appetizer?

This is cold, man.

[Anonymous Coward]

So cold. Liquid helium cold.

Smart Birds

[Anonymous Coward]

Bird#1 - "Hey lets do our formation exercise as always after 5 seconds we turn right then left after 6 seconds....the humans will think we are telepathic or something"

That's good.

[aliquis]

But can they run Crysis?

Re:That's good.

[Neil Boekend]

Of course not. Birds don't run, they fly.

Re:That's good.

[newcastlejon]

Birds don't run, they fly.

Meep-meep?

Re:That's good.

[Anonymous Coward]

Birds don't run, they fly.

Meep-meep?

Penguin wept.

Opposite of the gravity video detection algorithm

[Anonymous Coward]

Where there is not the additional pixels to explain the change in direction, because flocking.

They loves them some flying

[smittyoneeach]

. . .and it be showin' like a mother flocker!

The explain....

[Anonymous Coward]

That the damn birds do what is basically random movements. flying from one tree to the next, you get 10-50 wierd random "STOP FOLLOWING ME" types or turns and typically get the flock splitting as a bird dies not get it and the others around it simply follow it.

"get the signal" is funny. they freaking simply follow each other, there is no leader so it becomes the chaos bird cloud that we all see twisting and turning in the sky before it goes back to land pretty much where it took off from.

Helium?

[disposable60]

Does that explain why their singing voices are so high-pitched?

Re:Helium?

[JustOK]

no, it's because most of them crack their nuts with their beaks.

Re:Helium?

[gstoddart]

no, it's because most of them crack their nuts with their beaks.

LOL, once again, I am going to have to invoke rule #34.

Somewhere, in a dark and nasty corner of the interwebs is the human analog to this.

Now, excuse my, I have to go apply brain bleach.

So who is going to come out

[rossdee]

with a 'flock of birds' CPU cooler?

Somehow I don't think it would be as effective as liquid He

Don't blame GCC

[NemoinSpace]

swept through the flock at a constant speed between 20 and 40 meters per second.

I'm not even sure gcc would let you do this. But what do I know? I don't blame cigarettes for the death of Eric Garner either.

You mean a "murmuration" of starlings

[sideslash]

FTFY

Collision Avoidance

[sycodon]

I wonder if its just a case of one bird getting a itch in their ass and decide to turn and the guy next them turns just to get out of the way, etc. etc.

Liquid Helium?

[Chocolate Teapot]

That's very cool.

Re:Liquid Helium?

[Matheus]

I c w u d t .

what happened to the Voronoi polyhedrons?

[140Mandak262Jamuna]

Last time the bird flock was mentioned it was about how a flock flies past an obstacle without colliding. Something about each bird maintaining its position by maximizing the distance to its nearest neighbors. The math would work out such that each bird would form a vertex in a Delaunay tessellation and the "space" associated with each bird would form a Voronoi polyhedron.

I was kind of scared. You know, you spend all your life learning computational geometry and suddenly a flock of shearwaters or starlings show up and solve the same problem you have been solving for decades. You are given the pink slip and be replaced by a flock of bird brains. Man! that would suck.

But I am glad now, the birds are after bigger prize. No stupid engineering and mesh generation for them. They want pure science and may be they are after the Nobel prize. Glad they have moved on to simulating the liquid helium. Good for them. I think next thing will be they have solved the Cauchy-Riemman integral and they have a deterministic solution to Shroedinger's equation. They are going to finish off with a solution to Navier-Stokes equation with k-epsilon turbulence modeling.