Scientists form Essex University try out robot fish...
formatting link
Hey...this is inspiring... :)
How about robot fish aquarium as a tourist attraction? Would you visit a tourist attraction aquarium with 100's of real fish or 100's of robot fish?
How about robot fish for home use? Benefits No feeding No dead fish No filtering
Just charge up the fish..fill up the tank with Javex (whatever) and presto...one maintenance free fish aquarium..
I got an idea for recharging the fish.... When a fish dies....it drops to the bottom...(like a true dead fish :) ) The bottom has coils that put out a power field to recharge the fish.. Once the fish electronics sense complete charging, the fish takes off and resumes it's swimming code.
But dead robot fish on the bottom don't look good..So how else to charge the fish? D from BC
Bleach has a funny color. How about just adding a whole bunch of salt?
The fish could remain upright and the range that the power from the coils covers could allow the fish to get recharged just by hanging around in the bottom of the tank.
If you use a "super cap" as the storage, the charging could be only partial. The fish could be designed so that they always try to swim upwards slightly. The amount of "up bubble" and swim speed PWM could be a couple of some simple random functions. A fish that randomly hangs about at the bottom for a while, would get a bigger charge. Its thrust would be higher so it would go up to the top of the tank and swim up there for a while. Others would make only short trips up from the bottom.
It's been awhile since I looked at bleach... :) Salt water sounds good... IIRC fungus, bacteria,bugs, snails and algae shouldn't survive in salted tap water. Fish corrosion risk??
Nice idea, so..... The fish could just "chill out" with low activity near the bottom... It's not as entertaining but it's better than dead still on the bottom. (More funny if the programming put the fish upsidedown when charging..) Once charged, it can dart around until it has to "rest" again and cruise the bottom for charging... ( bottom feeders :) )
I think I understand your other charge up idea.. How it randomly charges determines how it swims.. Ex: A random deep dive (very close to the coils) will make a more active fish. D from BC
But let's say the fish have RFID technology too.. Fish #1 can be ordered to dock..then fish 2 and so on.. But there might be something fishy about that idea too. D from BC
The light source would be from the aquarium lighting so the fish are active all the time. But let's say that's not enough energy from small the solar cells. Maybe a fix is to use intense UV light. I might dig for some data sheets. But solar cells on fish would add weight, probably look ugly and there's not much area on top of a small robot fish.. D from BC
This is my second try on saying this. It looks like google groups lost the other post. If you see another that makes the same points, you will know that google found the lost one. I will add a few new ideas in here.
[....]
I was thinking that the randomness was in the PWM with no feedback. This means that a more charged battery automatically makes the fish faster and therefor on the average up.
If the center of thrust is below the center of drag, the thrust will tilt the fish up and make it swim up. This increases the degree to which the charge on the battery makes the fish swim upwards.
The whole swimming motion could be just one motor with a crank or cam. It could also be a pair of electromagnets. The body would be made of some flexible material so that just driving one point would create the curved shape of the fish motion.
The coupling from the motor to the externals would be springy. This with the resistance of the water would reduce the peak to peak motion of the tail. To turn, the motor's power is pulsed off briefly as the tail hits one extreme.
At the extremes the tail also bends so that it creates a down force, tipping the front up. Up down control is by pausing the power briefly at the extremes or the center point to either increase or decrease this effect.
The light on the top of the tank can be pulsed so that the following works. I will explain it as though it is steady light but the sunlight may prevent this from working without the modulation.
The fish can have 3 photodiodes. The two eyes each cover a 120 degree area that nearly overlaps. The third photodiode is hidden at the back of the mouth and only sees a narrow forward view.
An improvement to this idea is to use 3 pairs of photodiodes with little masks over them so that a point light source switches between them as it moves across the field of view. This would give a better signal for the below.
When the fish changes relative position WRT some object, this will cause the light falling on the eyes to vary. The more rapid the variation, the closer the object must be so the tail control will be commanded to swim away from whichever eye has the more rapid variations. I think a simple highpass filter and rectifier would serve as the measurement and a comparison would be the logic. There is no need for a linear control. There would be sort of a noise based PWMing happening on this control.
Variations in the photocell signal in the mouth are used to raise the gain of the above process. Earlier I had suggested that it just made "right rudder" into "full right rudder". I think I can improve on this idea with this observation:
The changes in the mouth's photocell will happen at the same time as whichever eye points more at the nearby object in front of the fish. Some nearby object behind the fish may cause the light on the other eye to vary at some other time. If the gain for the rectifier was raised when the change happened, this would bias things towards the fish avoiding objects in front. If the tails of the fish are made with the right sort of reflective and color properties, you can make the fish strongly avoid running into each other.
Now if we make the various feedbacks have a high and very nonlinear gain with some memory in them, the result will be chaos. This would make the logic a lot simpler because the chaos would serve in place of the randomizer.
Unless they are so small that they look like scales. ;-)
RF and a small signal diode array to rectify the RF to charge the batteries. You could even let it jam their cell phones, if the RF could be contained inside the building.
--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.
Michael A. Terrell
Central Florida
Wow...Now there's a sample of how much thought is required for making a robot fish.. I was only thinking about an entertaining fish patrolling pattern in the tank and achieving inductive charging...I haven't even considered fin movement yet..
If I start now...I'll have swimming robot fish in say....4 years with my spare time... Or..I hire a bunch of engineers..maybe 1 year or less depending on how much ping pong they play.. Heck..it's a fun job...anybody hiring for making robot fish??
It'll be great! Robot fish for dentist offices and restaurants. Kids will love it...Get the whole Finding Nemo collection! Fake koi in Japanese fish ponds.. etc.. D from BC
Salt water can't that bad on alternating magnetic fields.. Say at f=20khz.
When the robot fish code detects low charge, the fish will change routines and head toward the bottom within range of the induction coils under the tank..Fish to coil distance maybe 1/2". It'll do low energy swimming, soak up some mag energy and then take off into action again once charged.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.