h a l f b a k e r yNumber one on the no-fly list
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Taking off from lostdog's roaming goldfish bowl, as a way to let fish roam the house, we will first junk the airtight seal, and secondly, we will junk using the surface of the glass for providing motion.
We can now have a more or less regular aquarium mounted on caterpillar treads. A special lid,
similar to a sippy cup will be needed to prevent spills, but this is a minor detail since most aquariums come with a hood of some kind anyway.
Filtration will be provided via a special undergravel filter plate, with slits around the side, but not in the middle. A variable speed impeller will be mounted on a turret beneath this plate. (I'll explain the purpose of this complicated "filtration" later.)
Motion control will be provided by way of multiple PIR (Passive infra-red) sensors focused on different quadrants of the aquarium. The PIR sensors could be mounted to the floor of the tank, or to the sippy cup lid. When the fish moves into a new quadrant, the PIR in that sector of the tank detectes them, and the caterpillar treads will be commanded to move in that direction, and the impeller will be rotated to blow water towards that direction.
The directional impeller will force water to flow out of the plate towards the side of the tank that the fish was in, and it would then rush back the other way, giving the fish a sense of motion that would be more or less in sync with the actual motion he was making.
Getting the aquarium moving in a new direction with tank treads might be rather dificult, so for design purposes, wheels, or a track ball would probably be more ideal, but then to call it a fish "tank," I would have to install at least a 50 cal. gun, and I just don't trust my fish with that kind of firepower.
[link]
|
| |
//I just don't trust my fish with that kind of firepower// -tee hee! "What's this big red button do?" BLAM! "Ooh. I won't be doing that again in a hurry..." <30 seconds later> "What's this big red..." bun, btw. |
|
| |
This assumes that the temperature difference between the fish and the surrounding medium is enough to trigger the PIRs. But, hey, this is as bun-worthy as I've seen.
Maybe dolphins? |
|
| |
Bun for the title. Why not give little guppy a touch pad with some pre-recorded messages? Now he can roam around and interact. How about "Hey! Get up and feed me!" and "Can you change the channel? Jacques Cousteau is on". |
|
| |
Well, my idea was actually to have the PIR devices inside the water. Admittedly, specialized versions would be needed to avoid electric shock, if they work at all. |
|
| |
While fish are cold blooded, the laws of physics indicates that any moving thing operates at less than 100% efficiency, and the lost efficiency is given off as heat. |
|
| |
Whether the small amount of heat given off by a fish could be detected against the rising heat of the devices' motor is admittedly questionable. I was just trying to suggest some off the shelf parts to do the deed with. |
|
| |
A lot of inventions here reccomend putting some device on the fish to help track it, but I feel that doing so would be enormously difficult, and probably dangerous to the fish. |
|
| |
Infrared beams might work. Sonar is another possibility... HEY, now there we go. Of course, underwater mikes are a bit expensive... |
|
| |
Well, there's a reason this idea is at the Halfbakery. |
|
| |
We laugh in the face of high costs around here. |
|
| |
Laughing in the face of high costs here is fine... but most people cry in the face of high costs if they meet them at their workshop. |
|
| |
the large number of devices for letting fish move on land suggests to me that the workshop is where these ideas need to go at some point. |
|
| |
Great idea - probably an AI vision system would do the trick for guidance. Makes it more halfbakerly too. |
|
| |
I'd love to have a "warning: armed guard-fish" sign on my front door. |
|
| |
I doubt PIR sensing would work underwater. The good news
is that good computer vision is a lot more accessible and
computationally cheap than it was 13 years ago. (It would
have been doable back then, but it probably would have
been expensive, difficult to implement, slow, and
unreliable.) You could put a small camera at each corner of
the tank, and have the controller find the location and
direction of motion of the fish from those, as well as
interpret the fish's emotions (assuming fish have body
language—anybody know?) to add some extra features. |
|
| |