OK, I wanted to cast my net out for some high-level, general ideas on how to approach a problem. Any ideas or thoughts will be much appreciated.

Let's say we have two discrete objects that we will call, oh I don't know, how about "cats." Now let's further say that these two cats are: (1) living in the same house, (2) not kindly disposed to one another, and (3) capable of harming each other when engaged in physical contact.

What I would like to do is create a mechanism that serves to prevent them from getting within melee range. This will ensure that each cat is not terrorized by the other one, as well as preventing me from having a vet bill. Here are the conditions for the solution I must end up with:

1. I have in mind two paired objects which interact in some way so as to create an inverse relationship between distance and discomfort. Put another way, the closer the two cats get to one another, the more discomfort (i.e. negative motivation) they feel.

2. The paired objects need to be portable, obviously. A specially modified cat collar comes to mind.

3. The word "discomfort" was chosen very carefully. I do not believe in the existing shock collars. I simply will not hurt the cats. I understand there are enlightened people on both sides of this debate, and am not looking to argue with fellow nerdies whether shocking pets for behavior modification is OK or not OK. In this project, it is not OK. So, what does that leave us? I was thinking perhaps of a noisemaker of some kind. Perhaps a squeal that starts low at a certain distance (30 feet, perhaps), and then rises in frequency or volume as the distance goes down?

Some ideas that crossed my mind as I brainstormed this, in no particular order:

A. How do "invisible fences" for dogs work? I understand that one buries a wire at the perimeter of the property, and connects it to a base unit, but that is all I know. Can the equivalent of this base unit be made portable? Further, if so, how can this be made into a bidirectional process? In other words, in the original setup, anytime the collar and fence have a proximal interaction, the collar is the "punished" while the wire is the "punisher." If I were to adopt this technology, how could I ensure that both parties are "punished?"

B. For info about proximity to be communicated between two devices, the traffic could be in the form of: i. electromagnetic - maybe radio waves, ii. optical - perhaps infrared light, iii. sound - preferably ultrasonic, so that my partner and I (and the cats themselves) don't hear the base "carrier" tone at all times. [hmm, now that I think about it, I would need to investigate what range of hearing the cats possess...ultrasonic might well be heard by them, making this option unacceptable...]

C. Could one collar be some sort of RF wireless transmitter tuned to an unused frequency, while the other collar is a receiver set to the same frequency, such that the effect of bringing the two together creates the loud squeal of feedback?

So, what do you all think? Please chime in! Any thoughts? Ideas? Suggestions?

P.S. My partner suggested I adopt Dr Evil's "sharks with frickin' laserbeams on their heads" for the cats, but apart from being inhumane, that would tend to destroy furniture, start fires, and wreak havoc in general. :-)

Ok this is will be fun.

The invisible fence is a antenna once the collar device is close enough to the antena to pick up the signal (< 5 feet) the electronics in the collar triggers the shock. This is a fixed apparatus so is not applicable.

Using a RF (radio frequency) monitor would/could work. If a signal is received it could be designed to strengthen with proximity.

So you could produce a equivalent alarm.

Find a signal (frequency) that the cats do not like. There are Ultra Sonic deer and rabbit deterrers possible there is a frequency that cats do not like you could adjust the "loudness" to the strength of the proximity signal.

You could do this outside of the hearing range of a person.

I theory this sounds good but the problem would be weight and size.

I am working on (in my head) a tiny RF circuit apparatus that would fit on a cat collar but I really have no idea what I am doing so the likely hood of ever coming up with something is unlikely but who knows.

The idea I have would be a commercial product and something like what you need could use the same or similar components leading to further commercialization prospects.

I hope others will contribute as I really like this idea.

Ralph

This is just my two cents and some about this project. Feel free to consider or disreguard, but as a dad to two fuzzy monsters, this what was came up when I was thinking about this project. I have been thinking about this project abit and honestly don't think doing this as a completely self contained collar is such a good idea. You would have several difficulties and a couple safety issues, as mentioned below. I am honestly thinking that if you could do small "sensor" collars and a wall mounted base station or two it would solve several of the issues. Some of the issues first, and then my suggestions.

Safety Issues: - Device Weight: I don't know the safe amount of weight you could put on a cat's collar before it could cause issues, but I imagine it is very small. It would not work to protect them from each other but have the device give them neck or back issues. So if you have something big like 18lb Maine Coones they "should" be able to take a little more weight on a collar than a 4lb Siamese would, but each cat is different and I am not a vet. Definately an area to think about.
- Snagging: Also any kind of device would need to be in a very streamlined case to prevent snagging on things as the cat moves/runs around, over, under, and most importantly rubs up against things. Again, it wouldn't be good to have your cats seperated but stuck on something and possibly choking. Some people have their cats on break-away collars, for this reason alone.
- Power Safety: whatever battery powering this "collar" would need to be small, last long enough, and be completely contained so in case it did leak it would not harm your cats. While no battery is really "safe" (they all have nasty stuff in them, one way or the other), some are nastier to pets than others. Once again, I am not certain off all the fine details but it looks like most of the "electronic fence collars" I can find online use alkaline batteries(AAA, mini-6V's, or N's). - Ultrasonic Safety: Ultrasonic sound is a decent way to have a cat deterant, but it can be VERY harmfull to cats if used improperly. We use two ultrasonic devices in our yard to keep the neighbor's cats from terrorizing ours through the windows and spraying our cars. The big issue about ultrasonics is because you can't hear it, it is difficult to tell when/if it is on and how loud it is. There are BIG warnings on both of our ultrasonic devices warning that because they emit 3 second sweep 21-25kHz at 115db they CAN NOT be used indoors. I have read a few scarey stories of people buying them using them indoors, claiming they work great for a couple weeks then not at all, never realizing they made their cats deaf to ultrasonics. So if you decide to use ultrasonics, you will need to incorporate some visual method for detecting that the ultrasonics are ON, as well as determining a "safe" volume threshold that the device will NEVER go over.
- RF: I mention this as it is a big concern to some people(I have friends who refuse to have cell phones because of the cancer concerns.) If you use an RF signal for proximity detection, you will be exposing the cats to some (hopefully very samll) dose of RF emissions. Once again, I am not a vet and don't know if this is an issue for you, but wanted to put it out there as something to consider for this project.

Design Issues: - Proximity Detection: This could be a tough one. Unlike dogs, cats tend to go under tables, furniture, and other unaccessible places when you least need them to. This kind of rules out most of the line of sight things such as IR and will impact some of the lower power RF and ultrasonics. Also depending on whether your cats are long hair or short hair, it could affect how well the collar could get/send the signal. - Power: Running an ultrasonic punish circuit will chew up a lot of power. The send/recieve will use a decent bit of power too, but can't run out mid day or else its worthless. This is why I was thinking of 1 chip plus coin cell to a sensor collar and a wall powered base unit with punish circuit.

Suggestions: -#1. To some it sounds cliche, but I would talk to your vet first. Depending on the vet, they might be more than happy to give you suggestions and information for this project. If not, well, it looks like you need to do some more research. If they can give you information or good places to find out about the "safe max ultrasonic volume", max safe weight for the collars, additional info on battery safety(ie lithium worse than alkaline for safety), any RF concerns and questions; you can close some of the big safety issues neatly. They should know your cats and hopefully be able to give you more specific info than other sources.

-#2. I suggested the sensor collars and the wall/table base unit concept. If the collars only have a small send/recieve sensor and battery, it should cut the weight and power use down considerably. Also having the "punish" circuit mounted seperately in a base unit would give you more room and size for additional circuitry (ie "punish/no punish" indicators, "collision/cats too close" monitoring/tracking ) and wall power for the "punish" unit.

-#3. I am thinking ultrasonic or standard loud noise and light as a "punish" would be the better option for the output of this. If you can put up with it, the loud noise and light is the easiest, simplest, and safest method. But its REALLY annoying when you are around. As mentioned before, the ultrasonic is an effective method of cat deterrant, but needs some carefull thought. If I were to build such a thing, I would want a visual indicator of the "punish" noise and a way to tune/turn down/mute the ultrasonic volume. I would also want a specific volume detector to "tune" and test to make sure the ultrasonic volume is below the threshold for safe indoor use. I have noticed that when one of our ultrasonic cat deterrant's batteries gets low it drops frequency to where I can't really hear it but can sense it somewhat. Another plus for the wall mounted, wall powered "punish" base unit would be that the frequency output should be more stable than battery powered.

-#4. The proximity sensor is the tough part of the technology, as we all pretty much agree. I am thinking that IR and ultrasonic proximity sensing are most likely not going to work between the nature of cats and the extra issue of fur. I am thinking that RF might be the way to go. I am not an expert on how RF signals work, so my idea might not work at all, but my thought goes like this. If each cat's collar had a RF send/reciever, and each cat "pinged" at a different frequency. The other collar would be set up to listen for the other cat's frequency ping and if it heard a ping would send a signal to the base station. So say, once a second, Cat A's collar would send a 50ms ping at ?125kHz?. Cat B's collar would send a 50ms ping at ?145kHz? once a second. If Cat B's collar detected cat A's ping, it would send a signal to be base station. The circuitry would be set up so that the louder the ping, the longer/stronger the pulse to the base station. and if both collars are strongly signaling the base station, the "punish" would be louder/stronger. The pings would have to be low power, a for cat safety and b for reasonable proximity detection, after all it would be no good to punish the cats if they are 20 feet a part. I am assuming there would need to be some fine tuning on the power output to adjust the range detection. -#4b. Chip planning: The locals on here could help you with this more than I, but I wonder if you could use the atmega16 set up two interrupts using PWM/A-D inputs set for the RF frequencies you want. Set up tuned RLC/RC/LC with small antenna, so pings show up as pulses and the chip sends pings every second, listens for pings, etc... Dunno if it would work, but a thought.

-#5. Low Battery Detection: This is an important one, but could go several ways. You could add monitor chip/circuit to monitor power and signal the base station when power is low. You could also just make sure you change the batteries regularly and leave some chance to it. There are a few more options out there, but those are two to start.

Ok, I have rambled a lot. Hopefully there was some usefull thoughts and suggestions for you in there.

WOW! What a very well reasoned, and thoughtful reply--thanks very much! You have included so many good points, and with such detail, that it deserves a thorough read on my part before responding. I will say now though, that I totally agree with you about the safety of my babies first and foremost. As to the details of what you said, I need to read it and absorb it when I'm not in the office (like now!) and reply after a lot of thought. Thanks again!

Ok since we are just imagineering this, as I was reading 6ofhalfdozen's great post I thought (danger watch out) how about a self energizing power supply?

I have a flashlight that is powered by motion, tipping it back and forth charges a small storage device. It has a clear case so I can see a 1" wide coil of fine wire and there is a slug in the center that might be magnetic.

So imagine no need to worry about low batteries. The mechanism could be made quite small it seems, which would limit power output but most cats (unless the are sleeping) are constantly in motion.

I saw this at Stony Brook University (LI, NY) as a research project but bought the flashlight commercially so the technology is out there. There would still need to be some capacitors for power storage but they would be sealed units so probably not as potentially toxic as a lithium battery.

Like I said this will be fun to think about and 6ofhalfdozen once again a excellant post.

Ralph

What about water? Cat's hate water (at least 3 out of 4 mine do). Maybe some small motor + water tank? Well, you'd have to refill the tank, and it'd probably be better to make it in some jacket rather then collar (because of weight).

Also maybe some stinky stuff? then it can be kept on collar.

Here is a thought... Ultrasonic harmonics. Each has a collar generating a unique ultrasonic frequency. Nobody, not even the cats would be able to hear it but when the two get closer, the harmonic tones between the two frequencies becomes louder and louder. They might get the idea that whenever they are close, the noise follows them. I don't know if it would actually work but hey, it's a good thought experiment.

@Hexo I like the idea of each cat having a water pistol mounted to a harness! :-) as the two get within fighting range (aka the Danger Zone), then a water fight ensues. Of course, that would probably get them more hostile! Still, your idea is interesting...makes me imagine two Borg Cats

@Mongo Ultrasonic is something I have considered and am open to...I need to run that past my vet first, as I am thinking cats' hearing might be sensitive to that range, and I don't want to do anything potentially injurious to them.

Good ole fashioned RF might be an elegant working solution. Perhaps one is a transmitter of a simple audio tone, and the other is a receiver? That would end up producing a feedback squeal when the two came within close proximity.

@Hexo I like the idea of each cat having a water pistol mounted to a harness! :-) as the two get within fighting range (aka the Danger Zone), then a water fight ensues. Of course, that would probably get them more hostile! Still, your idea is interesting...makes me imagine two Borg Cats

@Mongo Ultrasonic is something I have considered and am open to...I need to run that past my vet first, as I am thinking cats' hearing might be sensitive to that range, and I don't want to do anything potentially injurious to them.

Good ole fashioned RF might be an elegant working solution. Perhaps one is a transmitter of a simple audio tone, and the other is a receiver? That would end up producing a feedback squeal when the two came within close proximity.

Ack, sorry for the double posting!

My other half is impatiently amused with my ideas...suggesting a simple solution: two rooms, two litter boxes, two water dishes, and keep them separated when we're away. Effective, true, but where's the fun in that? :-)

I like mongo's ultra sonic theory if each collar had a transmitter and a speaker the interference of the two signals could produce a squark. Fascinating concept.

Ralph

I know there are several devices on the market (bark control collars) that are made for various sizes of animals. You may be able to employ one of these to house whatever circuitry you come up with. You may even be able to obtain just the housing/collar w/o whatever circuitry they were originally fitted with if you contact the manufacturer. As far as detecting one another, how about RF, but have a Tx & Rx in each collar. I don't know enough about RF to know if this would be size prohibitive or not. I have remote controlled airplanes and I do know that the tiny size of a full range 6 channel receiver is absolutely amazing, and for something like this, you are not dealing with near the power or functionality. The two sets would operate on two seperate frequencies and each collar would have an opposing frequency Rx & Tx in it (if that makes any sense, I know what I mean :-) ) . Have the RF at very low power . Perhaps even use a potentiometer to adjust output power,thus adjusting the range at which the two detect one another. As far as punishment, how about about strobing bright LEDs, something like a camera flash emitting from the frontside of each collar. The nearer the two cats get to one another the more annoying the light would become without having to figure out how to create some sort of varying level of punishment based on how close the two units were to one another . This type of punishment shouldn't be too harmful since the cat is in complete control of it's own ability to escape the punishment. Further, the punishment is being dealt out by it's opponent, therefore reinforcing the idea of staying away from one another. The big downfall to this is that strobing lights at night while you are trying to sleep could be quite bothersome. The flipside to that is that cats are fairly intelligent creatures and if you figure out something annoying enough, they will learn to avoid one another like the plague. This sounds like a great idea and a fun project, I wish you the best of success and look forward to seeing your progress.