Moth to a Flame

[Classical Motion]

Have you seen to new report at CNN Science? It's about why insects flutter around a light bulb at night. And it's quite a novel idea, especially for people who contemplate motion in our questionable existence.

It explains that insects use lights of the sky for navigation. Flying insects have no contact with the ground. For up and down reference they use sky light for that reference. And the position of those reference sky lights, changes little during navigation. This is what makes sky lights such a successful reference. Especially at night.

A synthetic ground light however will drastically change position with any insect change in position. The insect keeps thinking the light is above them and keeps changing their flight position for that reference. The insect winds up, orbiting the light source.

The insect has no reference to navigate away from it. Their normal long distance reference has been replaced by the short distance reference. An inverted navigation reference.

Trapped in orbit.

 

[billslugg]

Insects can tell up from down. They are programmed to think the light is the sky, thus the light is "up". A point source is confusing them. This is why you can't get a flying bug out of a room until you turn the lights off and crack a window at the top. It is still true to say insects are attracted to light.

 

[Classical Motion]

We have two eyes that give us one spot to look at, at one time. Thus when we see motion, we track it with that one spot. We sense and interpret motion from sequential view dynamic. This takes time and thought(calculation). For orientation.

What if we have 10(or 500 hundred for fine control) fixed singular eyes? I ran across this years ago with autonomous drone navigation software. They were researching insect flight and the small "brain power" that was used to do this. They were looking for simple, fast, software algorithms to do this. Small brain navigation. For flying thru a thick forest or a dense asteroid belt.

And apparently, not only did the fixed, multi-direction eyes give depth perception, It gave views of all objects all the time. And the change in distance to all objects at the same time. But the neat part was that only the proportional changes of the objects are needed for save passage through a thick forest. This decreases processing power for other processor functions. Navigation might become a polled, background process. Automatic. And simple. How much processing can an insect do?

These engineers were very excited about this. It's been years now and not a peep. And I haven't heard anything about compound camera lens. Or sensors.

Insect navigation might have a simple hardware solution. We use many hardware solutions now for specific functions to avoid the repeated software set-ups and changes, and time needed for it. Many eyes and a hardware positional proportion-er might solve it.

Just spit balling.

 

[Classical Motion]

Might not 1 or 2 constantly rotating eyes do the same thing. Radar.

 

[billslugg]

Each separate "eye" is not an eye, it is a single pixel. A fly's eye has pixels spread around two hemispheres. Many of them overlap thus giving good depth perception. With so few pixels, little brainpower is required. It is basically no different than a human, just at a much smaller scale. The fly has a big advantage in the very short distance the nerve impulses must travel as compared to a human.

 

[Classical Motion]

If you take a field of pixels and lay them in a concave bowl, they will provide a field of vision at one location in one direction. One out of 720 directions.

If the pixels are put on the surface of a sphere, we get all 720 directions, with a depth perception also. Both a divergence and a convergence can give depth. Or distance.

I do think it opens new avenues for simple navigation strategies.

Moths are still attracted to light, but only believing it to be a direction, not a destination. That was their point. And mine.

Good point.

When the moth encounters a light source stronger than the sky light, the moth thinks that direction is up. When the moth encounters many objects in the way, he wants to go up to clear them. So he moves toward the light to gain height. But he encounters more objects on the way to the light. So, like from skylight he climbs higher, seeking clear space to fly in. But going to the light encounters more objects. Now, he is so close to the light, that he own movements changes the direction of up. He orbits around the light trying to get up to stand still.

What a predicament. Spinning in circles. Lost and won't admit it. It's probably the male insects.

 

[billslugg]

In any triangulation scheme, there must be a baseline between two receptors imaging the same object. A single fly's eye hemisphere cannot do this. Each facet of that eye sees a narrow cone of light that does not overlap the one next to it. Each facet produces one "bit" of information, not a complete image. Two fly's eye's hemispheres are needed for depth perception.

 

[Classical Motion]

With what I have read, that can easily be dis-proven by taping/painting over one eye of an insect. Will it bumble around? Yes for a while, but shortly fly just as well as others.

And recent studies have shown that one eyed people have depth perception also. But it takes practice. Cause we don't normally use it.

A triangular scheme doesn't seem to be as necessary as thought. Put the triangle on the other end. The observed size of a known or familiar object can denote distance. The rate of size change can denote velocity.

 

[billslugg]

Yes, you are correct in asserting there are other ways of judging distance, which is why I said "any triangulation scheme".

Another way to establish a baseline is to move sideways. An eye with a single pixel and a narrow angle of view can judge distance just as well simply by moving sideways to establish the baseline. This is why chickens move their heads fore and aft while walking.

Having two eyes allows one to judge distance even when not moving, like when a fly is on the counter and flies away just as your hand gets near. The fly might also be looking at rate of size change. Any darker (or lighter) area seen by a single compound eye, that gets big real fast is a danger warning, for sure.

Depth perception by focusing can give a crude estimate but our binocular vision is far superior. Flies cannot focus, they have many narrow fields of view thus an infinite depth of focus. They don't need to focus. They see just as sharp nearby or far away, in the angular sense.