We've posted these bird stabilization videos before (chickens are amazing!) but they never cease to impress. As good as we are with digital electronics, nature always does better with analog circuits.
(via Kottke)
You need to be a member of diydrones to add comments!
Comments
Funnily this was a discussion I was having with my wife just a couple of days ago. We have a chicken we keep as a pet and you can move her around, rotate her, yaw her, generally waggle her around and her head just remains absolutely rock steady. Plus the eggs are wonderful.
If I knew how to use neural network software to build algorithms, I'd attach an IMU and camera to the chicken and collect data. Train the network to match the chicken in performance and then export the algorithm in C and put it on a chip. Cracker Jacks still hasnt sent my PhD, so it will be awhile.
Somebody has already started work on using a chicken as a Steadicam:
https://www.youtube.com/watch?v=UytSNlHw8J8&feature=player_embed...
But it still needs some work. Obviously, it's hard to control the aim point. And the chicken's PID loop needs to be retuned to accomodate the weight of the camera. You can see some jitter, especially after fast movements.
In neurons that transmit action potentials firing an action potential is 1:0, but the signal is often in the frequency of firing, and not all neurons transmit action potentials - many just apply inhibitory or excitatory stimuli to other neurons. Transmission from neuron to neuron is through release of chemicals across the synapse (usually) and that is far from digital.
When the owl's eyes are covered the lag and damping change. When you cover our steppe eagle's eyes and turn him round he falls off his perch. The visual input is key.
However, what they are doing is very odd - they stabilise the sensors, but stabilise them independent of the center of gravity, and allow the body and the aerodynamic surfaces to move radically relative to the sensors. Very odd indeed.....
What's interesting is that the tendency to hold the eyes level to the horizon is common to pretty much all animals, whether owl, human, or insect.
Also, you have to keep in mind that not only is the head held in one orientation, but the eyes themselves are another layer of stabilization.
As for analog vs. digital- the firing of a neuron is a 1/0 activity, but the timing of the neuron firing isn't. It's pretty complicated, and it can be said that "the brain" has both analog and digital processing in it. But the digital processing is nothing like the algorithmic processing you see in an Arduino (or a supercomputer).
I'm no neuroscientist, just a microbiologist.
But at some level things break down to a digital level. You either get an electrical pulse or you don't.
An analog system with perfect fidelity and a digital system with infinite resolution would be indistinguishable, so the debate pretty much comes down to semantics.
As a point of fact though, the neurological system is digital at the inter-cell communication level, which makes it a digital system. This is the same as in an electrical circuit. Only the communication in any system is ever actually digital.
> fire based on spacial and temporal arrangement of incoming impulses on a cell membrane...
Doesn't make it analog just because it has some analog inputs. The neuron it's self can be looked at like a processor. The effects you mentioned are part of the processing. In the end the neuron either fires or it doesn't that is digital by definition.
no your right chris the mind works in analog always had, in fact like you said the sell don't work in on off stats the closest digital analog (lol pun) would be a balanced trinary (-1,0,+1) but that's still more limited.
Jake: You may be right; I don't know enough neuroscience to be sure. I know the pulses in calcium channels are not pure digital on/off step functions, so I called them analog, but perhaps I'm getting the semantics wrong.
-
1
-
2
of 2 Next