In a fascinating read, Science Mag describes new research on the seemingly perfectly coordinated turns of large flocks of birds.
Long a topic of speculation and study, these turns were even attributed to telepathy by ornithologists in the 1930s. In the 1990s the proposed theory, that each bird simply matches the direction of an adjacent bird's turn, didn't adequately explain in detail how the flock turns as one, and didn't explain why there were few if any stragglers. Even studies as late as 2011 suggested "each starling was connected to every other" by some unknown mechanism.
In the new study, the team, led by physicists, used high-speed cameras to film starlings, which form spectacular synchronized flocks. There is no "connection", simply a self preservation instinct and rapid information signaling.
Using tracking software on the recorded video, the team could pinpoint when and where individuals decide to turn, information that enabled them to follow how the decision sweeps through the flock.
The tracking data showed that the message 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. The birds mimic not only the direction but also the sharpness of the turn of neighbors in three dimensions. Information about direction changes propagates across the flock with a linear dispersion law and negligible attenuation.
The fact that the information telling each bird to turn moves at a constant speed contradicts the 1990's model, and explains in detail how the signal flows through the flock to precisely coordinate turns which are signaled by which ever birds happen to be on the leading edge or the edge nearest a threat. The signal's speed and near absence of attenuation, mislead earlier investigators into imagining some "connection".
Being physicists, the researchers couldn't help but notice that the formula modeling this process was mathematically identical to the equations that describe superfluid helium. When helium is cooled close to absolute zero, it becomes a liquid with no viscosity at all, exhibiting a cohesion that's mathematically similar to a starling flock.
(Score: 2, Interesting) by sudo rm -rf on Monday July 28 2014, @12:03PM
(I'm sure others will comment on the headline, so i won't.)
Let's assume the space between two birds next to each other is 0.5m. The message travels at 20m/s, so each bird reacts within 0.025 seconds. That would be amazingly fast! Given some minimum response time needed (can't find any numbers now, but it must be around 80-100 ms) those feathered friends must be watching not their direct neighbour, but the one 3 birds away.
(Score: 1, Funny) by Anonymous Coward on Monday July 28 2014, @12:31PM
Obvious opportunity omission!
(Score: 3, Interesting) by VLM on Monday July 28 2014, @01:19PM
Something else to contemplate, is assume a cubical flock (this is starting to sound awful) then the cube is 400 ** (1/3) on a side or about 7 birds on each edge of the cube (I bet you were imagining tens of thousands of birds darkening the sky, but a borg cube, 7 birds on a side, isn't quite so impressive)
All you need to do to get peculiar behavior is have every bird calculate independently a function based on the outside threat and how close they feel to the edge of the borg cube and respond accordingly. They'll get plenty of practice to refine this calculation in the real world. I'm not seeing a coordination being required other than simple avoidance.
You could probably run some interesting experience by firing a "supersonic" hawk or whatever into a flock. "supersonic" not as in sound in air, but as in moving faster than the supposed communication speed, which a diving predator bird could probably achieve. The shock wave patterns of behavior would be interesting to analyze.
This might also turn into a latency vs bandwidth problem. Although I'm not a sports guy, it takes a "long time" to decide where to throw a ball that you catch. However, if you run the math on the timing accuracy required to achieve baseball player level accuracy with a thrown ball it works out to a whole body coordination problem where the fingertips have to be coordinated with the toes to an accuracy of microseconds but that doesn't mean messages are being passed between them... just means precision is required. If the entire flock solves for an external threat and then each bird adds a delta to their response based on their flock position, then later when they all act at the same time I'd no problem at all to have "coordinated" within milliseconds responses by individual actors. Aren't two kids playing catch with a ball, coordinating their entire muscle systems both individually and with each other to at least single ms levels, even though their brain and nerve latencies are much longer? This is probably why it doesn't take extensive practice to learn how to open a beer can, but it does take extensive practice to play catch.
(Score: 2) by isostatic on Monday July 28 2014, @01:31PM
friends must be watching not their direct neighbour, but the one 3 birds away.
Exactly the same as driving, you're looking at the cars ahead, not just the car in front of you. Only way to drive safely on a modern highway with gaps sub-one-second.
(Score: 0) by Anonymous Coward on Monday July 28 2014, @02:49PM
(Score: 2) by CoolHand on Tuesday July 29 2014, @03:36PM
Only way to drive safely on a modern highway with gaps sub-one-second.
YOU control the gap in front of you, keep it at two seconds, and if you're being tailgated, increase it. If cars start cutting you off and decreasing the gap, just take the foot off the gas a bit for a half second and get your gap back. It doesn't really cost you much time, and is much safer than being a fractional second behind the car in front of you. You must just get over the feeling that you're "losing the race", but once you're able to accomplish that, it makes for much less of a stressful drive.
Anyone who is capable of getting themselves made President should on no account be allowed to do the job-Douglas Adams
(Score: 2) by isostatic on Wednesday July 30 2014, @09:52AM
No, you drop back, then a car fills the space. Then you drop back, and a car fills the space. Before you know it you're stuck doing 55mph and lorries are overtaking.
(Score: 2) by CoolHand on Wednesday July 30 2014, @11:49AM
That's the thing, you don't ever "slow down." You're speed remains as fast as before, the only time you lose is the smallest fraction of a second to create the space each time someone cuts in. The number of cars actually cutting in isn't huge - if it's really packed, not that many will be able to get from behind you, and if coming from the other lane, you're likely they'll jump back in that one as soon as time allows if you're lane hits a minor slow down. This same discussion was on the "old site", and someone brought this argument up, and I was NOT a believer, but after trying it a few times, it really does work. Admittedly sometimes I still lose my way, and if I'm in a really big hurry, I'll still get aggressive and ride bumper to bumper at 65mph, but when I can bring myself NOT to do it (and not get mad at those people "cutting in"), it really does become a much more stress-free time, and it actually isn't noticeably slower in total travel time. I commute an hour each way every day, mostly on freeways, so I've had a lot of time to practice. You should try it a few times.. :) (again, the trick is to not get mad at people cutting in front of you, just maintain your gap - and speed)
Anyone who is capable of getting themselves made President should on no account be allowed to do the job-Douglas Adams
(Score: 3, Informative) by Zinho on Monday July 28 2014, @03:47PM
Let's throw some numbers on that, then:
I found one reference [jstor.org] stating that song birds when singing "duets" with each other will react on average within 120 ms (with only a few ms deviation), so your ~100 ms guess is pretty close. That's about half of human reaction time [humanbenchmark.com] (median of 215 milliseconds).
On a related note, one ornithologist suggests [loe.org] that the best way to listen to bird song is slowed down significantly; otherwise we miss much of the fine detail of the warbles. For us to even understand what's going on in the song requires a slow-down of 1/2 to 1/8th.
I suspect that many other species experience the passage of time differently from us, as evidenced by these intricate birdsongs and the coordinated flight. It's been shown that octopuses don't react to images of predators or prey on SDTV CRT monitors [biologists.org], but do for LCD HDTV; 50 frames per second and updating the full screen at once seems to work. I interpret this to mean that the low framerate and design for human Persistence-of-Vision inherent in a CRT lead to the Octopus not perceiving it as an image, but instead just a slowly moving bright line. If the birds perceive the world passing in slow-motion relative to our experience then it is much less surprising that they are capable of superhuman reaction times. This may be attributed to the birds' higher body temperatures [ktna.org] which improve the chemical reaction rates in the birds' nervous systems.
I like your theory, too, that the birds are watching farther-away neighbors for cues. Speed of light is always going to be faster than individual birds' reaction rates, even if they are faster than ours by an order of magnitude.
"Space Exploration is not endless circles in low earth orbit." -Buzz Aldrin
(Score: 2) by PizzaRollPlinkett on Monday July 28 2014, @01:18PM
Quit being cute and just write a headline that humans can parse.
(E-mail me if you want a pizza roll!)
(Score: 2) by Thexalon on Monday July 28 2014, @01:45PM
As always, alliteration and assonance are asinine!
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 1) by MikeRo on Monday July 28 2014, @03:23PM
Hurts Hacher's Head? Hexcellent.
(Score: 0) by Anonymous Coward on Monday July 28 2014, @04:26PM
If I parse that as every atom is mathematically similar to a starling flock due to similarities in their cohesion, it makes no sense. Do they mean "adhesion" and not "cohesion"? That would seem a more appropriate use of the term when comparing starling flocks to atoms (!!). Or do they mean every atom is mathematically similar to a starling WITHIN a starling flock, which would make more sense to me?