Looking for Cheap Proximity Sensor (For Cat-Flap)

[...]

Most prox sensors I'm familiar with are non-discriminatory: They would open for the neighborhood dog just as well as the cat, which probably isn't what you want (if so, just use a normal cat flap).

RFID tags could do more than you need, though I'm not sure you (or I) would consider them cheap. TI sells a mid-range reader module good to about 175 mm for less than $200

, though you may want the eval kit, which is a bit more expensive ($645, , but includes some transponders (tags) and other helpful stuff.

Longer-range readers are available if you have a lot more money to spend.

The tags themselves can be very cheap, but tend to be sold in volume. I'm not sure where you could buy just a handful (other than the eval kit).

HTH,

-=Dave

--
Change is inevitable, progress is not.

Retinal scanner or kitty facial recognition would be my reccomendations. You can mount either the laser scanner or camera on the side of the cat door, then run the scanning algorithm on the PC in your house.

Another option would be a wireless ID unit embedded within the cat. RFID tags sound nice, but what happens if the cat loses its collar and its RFID tag on a cold, winter night. It'll be stuck outside and freeze to death! No, you definately want to mount the electronics inside the cat. Its the only way to be sure.

I can see it now. Entrepreneur buys thousands of identical tags, and markets system. Becomes popular, and three households on the same block buy in. Households immediately become common to all three sets of pets. Cats, dogs, ferrets, maybe the odd skunk. All householders add additional collars and invisible fences. Pets, weighted down by innumerable special purpose collars, die early of physical exhaustion. Lawyers enter the field, armed with innumerable writs, liens, depositions, etc. Owners move out. Remaining pets become feral as collars wear off. City ordinances are passed.

--
Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net)
   Available for consulting/temporary embedded and systems.
     USE worldnet address!

Get a terrier. Guaranteed to detect a cat 100-200mm from the door and doesn't need batteries.

In the Netherlands, you can have your pet injected with an RFID tag. Wouldn't be too hard to read that one out.

Meindert

Many dogs and cats already have unique grain-of-rice sized implanted passive RFID tags. (My cat does.) Used with national registries, they can help reunite lost pets with their owners, and possibly also provide medical information.. As I recall there are at least three different companies doing this in the US, and no doubt their tags are all mutually incompatible.

Jim McGinnis

[%X]

Sounds like an ideal application for an RFID tag. Small, inductively powered from the receiver and, if you have more than one cat (or other pets) would know which one was entering/leaving (useful status info).

--
********************************************************************
Paul E. Bennett ....................
Forth based HIDECS Consultancy .....
Mob: +44 (0)7811-639972 .........NOW AVAILABLE:- HIDECS COURSE......
Tel: +44 (0)1235-811095 .... see http://www.feabhas.com for details.
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************

[...re: RFID...]

But you can't. Each tag (ISO 15693) has a unique 64-bit identifier. Even if the numbers do eventually repeat, 18.45e18 is a Really Big Number(tm). Finding two tags with the same ID would be... unlikely.

Many tags also have user-writable areas of up to 64k bits. You could put your name and address in there.

In fact, companies like AVID will inject RFID tags under your pet's skin (Now that's a *real* embedded system...), though these are generally just ID numbers with the real ID information kept in a cental database.

Regards,

-=Dave

--
Change is inevitable, progress is not.

I believe that the average terrier would have almost no false negatives when used as a cat detector. But are false positives a problem? aka barking for no apparant reason?

Kelly

Repeat are you serious?

If they could manufacture tags at a rate of 1,000,000 a second, and ran their machines 24 hours a day, 7 days a week, it would take over 584,000 years to start repeating identifiers.

Finding two tags with the same ID would indeed be unlikely. And won't really be a problem unless 584,542 years from now you still have the same, *really* old cat. ;-)

-Zonn

-------------------------------------------------------- Zonn Moore Zektor, LLC

Remove the ".AOL" from the email address to reply.

Not really. But some people (Benneton? for example) are talking about putting one of these transponders in every item they sell, so they're talking volume production.

Well, that assumes 1 manufacturer using all 64 bits. I don't know, but I'm guessing that each manufacturer gets a block of numbers to use.

If I don't know this for a fact, I sure don't know the size of the manufaturer ID, but let's assume 16 bits. That leaves 48 bits for

281.5e12 unique IDs. Using your 1e6/minute, and assuming your math was correct, that still gives each mfr almost 9000 years. Should be adequate... At least until some newer technology replaces it.

Regards,

-=Dave

--
Change is inevitable, progress is not.

I wasn't really serious either, I just happened to have a calculator in my hand when I read the post. I knew it was tongue in cheek by the "Really Big Number(tm)" trade mark. ;-)

But as long as I have that calculator, let's talk volume...

If 100 tags weigh an ounce, and only 48 bits are used per manufacturer, then

2^48 tags would weigh 87,960,930 tons. Now, that's volume production! :-)

-Zonn

-------------------------------------------------------- Zonn Moore Zektor, LLC

Remove the ".AOL" from the email address to reply.

Hi Anton,

You might think the shape-recognition idea was frivolous, but I have seen at least one fielded project using simple shape recognition of the animal's silhouette to correctly allow the designer's cat to enter [within certain time periods], while rejecting the cat if he has something in his mouth or if it's some other animal. The principal reason for building the device was to prevent the cat from bringing in mice, sparrows, etc.

The web site for the project shows lots of sample photos, rather sinister looking skunks and so on (all in silhouette) captured by an IR-camera.

Unfortunately I can't find it in a couple of moments' search with Google, but I KNOW it exists. Maybe it was mentioned here last year sometime. It seemed very cool.

In message , Anton Erasmus writes

Was it a Staywell device?

Anton, the coded keys you talk of use RFID 'chips' Philips, Texas Instruments and Microchip are amongst the larger manufacturers of these chips. The simplest way to obtain a reader *might* be to find a junk yard with a written off vehicle and you will need to work out which RFID was used . The reader coil is usually around the barrel of the ignition switch and the drive electronics is fairly easily found by following the cables from the coil. Most of the RFID chips are readable at 20 CM and are small enough to incorporate into a pet collar.

--
Clint

My cat actually have an RFID Tag already. (I got her at the SPCA, and they put one in as standard practice). The reader they use has to be very close (Next to the skin) to read the tag. What sort of circuitry would be needed to read the existing RFID tag over a distance of 200mm ?

Regards Anton Erasmus

Now if you can suggest a cheap way to interface this terrier to an AVR, then it might work :-)

Anton

Yes , I have seen a short snippet of this on "Ripleys". I might look into this idea more seriously when one start to get the camera modules used in Cell Phones and PDAs with some sort of standard interface that can be easily interfaced to something like these new LPC21XX ARMs from Philips. So far I have found one Korean manufacturer that is bringing out a module (2004) for

No it is some other UK manufactured Cat-Flap. The Staywell was 3 times the price. With shipping and everything I ended up paying about UKP70 for the thing.

It seems that the RFID is the way to go. The biggest problem with this seems to be that it is a market dominated by very big volume customers with the resulting difficulties in getting hobbiest quantities and support.

Regards Anton Erasmus

What about a small RF transmitter on the cat and a directional antenna near the door?

Cheers

Klaus