Slashdot Mirror
Mechanical Butterflies?
MImeKillEr writes "According to an article on BBC News, two researchers from Oxford took highspeed photographs of an Admiral butterfly in a specially-designed windtunnel to study how butterflies fly. The resulting research brings insight into small-scale flight dynamics. Although the article doesn't give an ETA on this, they expect to be able to build an aircraft with a 10cm wingspan that will be either autonomous or radio controlled. This will allow them to be used in rescue missions, cave exploration and possibly even on Mars."
The cool thing about digital photography is that all you really have to do is a flyby. You can stop, zoom, and process the captured video images afterwards.
Also, all this thing has to do is broadcast a live video transmission. Recording it can be done remotely, so you don't have to worry about either recovering the device, or taking up weight with memory or recording media. Simply fly in as far as the power source will allow then either recover the device later or hope the images you have are worth the cost of losing it.
"Everything you know is wrong. (And stupid.)"
Moderation Totals: Wrong=2, Stupid=3, Total=5.
Ok, to clarifly; I think this problem, or almost any other AI problem, _can_ be solved by this aproach, just not in a efficient way (like NP hard kind of efficient ) Intelligent behaviour arises directly out of the relationships of parts, not out of any real ability of the parts.
You say your watch is about as smart as a butterfly. Let's see it fly to Mexico and get laid.
In the end, I think you might just be able to get a Casio watch to pilot a butterfly, but it's going to take a lot more insight than some high speed photographs.
"A language that doesn't affect the way you think about programming, is not worth knowing" - Alan Perlis
Scale is important. Turbulence happens more readily at large scales. Viscous drag is more significant at smaller scales. Gravity is more significant at larger scales. A very small insect is effectively rowing through the air, using most if its effort to propel itself along. An aircraft spends most of its effort creating lift - and drag, because the two always go together - to keep itself up. So, we're not going to have 747's with butterfly-shaped wings flitting from building to building. Which is a shame....
from Dune.
Again, reality imitates (science) fiction. Nice!
Sigged!
We can produce land vehicles than can travel across continents without refuelling. No animal can do this.
We can produce aeroplanes that will fly around the world without refuelling. No bird can do this.
I see no fundamental reason why we can't produce a mechanical butterfly that can operate for days without refuelling as real butterflies can achieve this. If you are really small then the energy required to keep you aloft is really small also. I've absolutely no idea how much energy a butterfly requires to keep it in the air for a day but my guess would be that it is considerably less than that contained in one drop of petrol.
wot no sig
A few people have suggested the military applications for these robotic butterflies, particularly in the area of espionage. The problem is this: butterflies make terrible spies, because everyone notices them!
How many times have you heard someone say, "look at the pretty horsefly on the windowsill"? Eh, never right? But we notice butterflies, because they are beautiful and elegant. In fact, of all of the insects around, I'd say butterflies are the ones most likely to be noticed, picked up, and examined because they are colorful and generally harmless. Well, that's probably the last thing you want, someone picking up your robot spy. "Hey, this butterfly has a resistor soldered to it's back..."
So the idea of making a robotic butterfly spy is probably not workable. Maybe a robotic cockroach spy..? Never mind, they'd just get stomped on sight. That might just be the real problem, trying to find an insect that wouldn't provoke interest, either positive or negative.