Pet door project

project_pets
project_animals

(John) #1

Starting a DIY automated pet door. Have a 1st gen ST hub, and have just a few devices on it - a lightbulb, a Schlage lock, and a Chamberlain garage door. Those are all prebuilt systems, this would be first DIY.

I already have an in-wall flap dog door, and it has a metal sliding cover on the inside. We have been using this with just he metal door pinned up and the dog can go in and out to his heart’s content. However, sometimes at 4 am he wants to go out and tell the deer or fox to get off his lawn.

If we close the metal slider, then that means in the morning I have to get up and go downstairs to open it so he can go out.

Additionally, winter is coming, and I think when it gets cold that flap is going to be leakier than we would like.

I think that this will be relatively straightforward, mechanically. I got a linear actuator and inbound is a monoprice curtain controller (I have found custom device code that will allow that to work with ST). So that is the open and close control, and all by itself would be easy enough to just open and close on a schedule.

HOWEVER - I started thinking through some scenarios. Suppose I have it set to close at 10 pm, the dog goes out at 9:59 and comes back to find it closed. Then he can’t get in on his own. A motion sensor outside would do the trick, but a raccoon could also trigger that, so I want it to be RFID or something like that. And, I would also want to have the RFID on the inside so it wouldn’t stay open all the time, only when the dog came over to open it.

I am thinking, then, I need an inside RFID sensor that can only open the door within certain hours, and an outside RFID sensor that will open the door anytime.

I have seen some RFID implementations for ST and it looks like arduino is the way to go there.

So, any thoughts or direction on this? Have searched the community and googled this but haven’t found anything directly like this.


#2

I am quadriparetic, use a wheelchair with limited hand function. I have a service dog.

Most dogs can very easily learn to use buttons that are mounted at nose height. That solves most of the issues. The dog uses the button when they want to.

(You can put a cover over it when you don’t want the dog to be able to use it. Just make sure the cover is really noticeable or they get frustrated. It’s good to make it a different shape than the switch itself, so if the switch is a rectangle make the cover a big circle. Most dogs distinguish geometric shapes very well. )

WIth SmartThings, we use the battery powered smartenit 3 toggle switch. We just program each toggle to do exactly the same thing so it doesn’t make that much difference where the dog hits it. It’s hard for him to be precise about the individual toggles.

There’s a new panic button coming out from Iris, and it might be even easier form factor for a dog to use. But is not available yet.

The Easy Button–Really easy for dogs

Some people also take a Staples easy button and convert that. Those happen to be a really easy form factor for a dog, we’ve used those in training my dog to press the buttons on automatic doors. But you do have to convert them so they can work with smart things.

a new small button option?

I also have some of the new flic buttons on order, but I don’t know if my dog will be able to use them yet. These are about quarter size. They don’t work directly with smartthings, but they have an IFTTT channel so they should have indirect integration.

Anyway, I would you some kind of press button rather than something that the dog wears. Dogs are generally very hard on clipon equipment. It gets wet, they manage to get it off the collar, it breaks, or if you have two dogs, one of them chews on the one on the other dog’s collar. So my preference is to train the dog to use a switch or button instead.

FWIW. :sunglasses:


(Bruce) #3

You are inspiring me to create this as my next DIY. How about a Smartthings presense sensor that is added to the dogs collar? Unless the racoon is able to wrangle the collar from your dog well then you should be able to accomplish this.

Please post specific details for those of us considering implementing something similar. My use would be to roll over in the morning and press a button on a tablet or phone and raise the dog door shield so that I did not have to get out of bed and walk to the dog door to physically remove the dog door shield.

Things I would be interested to learn :smile:

  1. What Linear Actuator
  2. What is the throw on the linear actuator
  3. Battery powered so there is no need to run wiring and bulk up the appearance on the inside of the door
  4. Link for curtain controller
  5. Could this just be accomplsihed with a relay for on that uses a limit switch to determine when it is at full height and stop the motion of the linear actuator and then use some method to reverse the direction

Did some research and determined that the Z Wave curtain controller needs 12 Volt DC at 25 milliamps. This is low enough current that can be supplied by an external / auxillary battery case with 2, 3, or 4 12 volt DC Batteries wired in parallel ( for increased capacity / longer life) .

I am going to 3 D print an enclosure to house the multiple cells.


#4

The smartthings arrival sensor will not work for this purpose. It’s not directional enough. Basically it can’t tell the difference between being in the yard and being in the house.

Besides, it’s honestly not weatherproof enough for most dogs to wear. I know they show a dog with it in one of the marketing materials, and I’m sure there are some dogs it might work for, but it’s just not going to hold up for your typical dog who runs around.

In order for it to work for a dog door so the door doesn’t just open every time the dog passes, you need to limit the area detection to about 4 feet. That’s pretty small.

You can do it with an RFID device, but there’s no way to connect those directly to smartThings. And if you look at the reviews for the ones that are commercially available, you’ll see a lot of complaints about them opening at the wrong times. Just hard to get the range right.

But again it will work for some people so no harm in trying. :sunglasses:

There’s a guy who’s working on a project for a cat, and I did suggest a collar device with an ibeacon for that, but cats don’t usually roll in the mud. :smile_cat:


(Scott G) #5

I’m not sure if it’s still available, but Lowes sold a pet door with some of these features. We got it to work like a normal door lock, but not with the collar device. Check out the below thread (post 27 has code).


#6

BTW, from an engineering standpoint, I really like the Solo doors. They work very differently from the other commercial designs. The tag which the dog wears is literally just a magnet that causes a switch to close inside the door that then activates the motor.

The motor only works to pull the door up. It comes down on its own on a dragline, which makes it both slow and safe.

Just really nicely made. (In the USA)

If you just want a door that only opens when the dog goes near it, I would look at this.

Then you’d have have to look at time locks and stuff if you wanted to add something more to it and integrate it with smart things. Just wanted to mention it because the design is worth studying.

Do be aware, that you can get some weird cases if you walk a vacuum cleaner past it or even start a washing machine near it. Read the reviews. Apparently it’s a very bad choice to put in a garage or a laundry room. But nice in a regular back door or porch door.


(John) #9

Well, apparently some people (or a robot) decided that linking to amazon to show the pieces was bad. Sorry if that was bad form. So, anyway, it is an 18 in linear controller and a 12v remote controller and power supply.


(John) #10

I looked at the ST sensor and as noted it is not very granular - it is are you home, not are you standing by the door.


#11

We link to parts on Amazon all the time, It’s probably just because you’re a new forum member, not sure you’re allowed to post any links for awhile. It’s just to help cut down spammers. You get more and more posting privileges as you’ve been a member for longer.

Anyway, I do understand the budget issues.

For some of us, it’s not just a first world convenience issue though. I’m quadriparetic. It takes me about 45 minutes to get out of bed into the wheelchair. So we automate a lot of stuff around the house. :sunglasses:


(John) #12

I guess I just had too many links out of the gate

I have seen the petsafe door as well as some ST threads on it.

First of all, the dog door in this house is legacy from the previous owner, so it is already installed. That makes the automation piece a retrofit. And, I would like to not spend a lot on this - really, this is a first world problem, having to get up to let your dog out, so I can’t justify TOO much money on it.

I just finished a proof of concept with it as far as being remote control. I have not received the z wave controller for it yet so I have another remote piece I ordered thinking I would hook that up with some sort of Arduino controller before I found the curtain controller.

And I am talking about convenience for me, JD I can imagine that this stuff can be real game changer for you.


(John) #13

Current POC. Takes about 65 seconds to fully open or close.


#14

Sure, everybody’s use cases are different. :sunglasses:

Take a look at the cat project I link to a couple close up. That’s using an IBeacon that is designed as a wearable for people to use it conferences, and it would definitely work on a collar. Just a question of whether you think your dog will knock it off. But I beacons let you get a really nice range control.

I have one set up to trigger when I reach the base of the wheelchair ramp at my front door, and then not trigger when I’m in the house.

That’s also what I called the “bus stop” problem which is where the bus stops half a block from my house and it takes the bus driver a few minutes to get me unloaded. I don’t want my arrival routine to trigger until I’m actually with inside of my front door. Neither phone nor the smart things arrival sensor gave me enough granularity for that, but the IBeacon does.


(Bruce) #15

I like the linear actuator approach except for the way it will all “look” when bolted or screwed to the inside of the door. Plus it would be great to not have to run wires across the door opening on the side with the hinges. If I want to remove all the casing I could run the wire behind the casing and then create a positive and negative plate on the casing side that was inset to not interfere with the door closing and then create those same plates on the door side.

Then when the door was open there would be no contact of the plates and no power supplied to the linear actuator. Then when the door was closed the sets of metal plates would sandwich together and create the needed voltage/ current to allow the linear actuator to operate.

This is a pretty in depth way to go. Concealment would lend more mass acceptance I think though.

I am considering it for a main door that we use all the time. I guess I could have them bend a long piece of sheet metal in a horseshoe or U shape and fit that over the linear actuator. That would conceal it to an extent.


#16

Might take a look at some of the pictures for automatic door openers for people in wheelchairs. There is pretty good aesthetic acceptance when the actuator hardware is horizontal along the top edge of the door, either on the door or just above it. As long as it is vertically at the very top of the door, it tends to fade out of the picture. It’s hardware that’s right in the middle of the door that tends to seem jarring.


(John) #17

On a door would be a little harder to get power to, Though this particular actuator has about a 2 foot lead that is not very thick, you could get that over to the door frame and have some sort of plastic cable protector on it at the door jamb. And your controller is on the wall with an enclosure or something. All depends on what is where I guess.

We luck out because this is an in wall pet door, and that is the laundry/mud room so it doesn’t necessarily have to be the ultimate aesthetically. Right now there is some amount of Gorilla tape involved. If it were in the kitchen I don’t think mama would be as pleased with the development project.

Hard part will really be the RFID controller. Conceptually I get Arduino but have never worked with one. My programming skills are best with cut and paste so I am hoping I can make that work. The ST arduino shield is $35, so a $20 arduino and then another 20 or so for the RFID part, and then making it x2 (inside and outside) would keep this within acceptable budget.


#18

Many of the Arduino experts in the community hang out in the following topic. You might check it out. :sunglasses:


(Bruce) #19

Yeah, you lucked out with an in wall dog door! To complicate matters for me . The dog door is on the most used door of the house which enters into the kitchen area. BAM! I got double screwed.

I am probably going to have to route cabling inside the door and over to the door jamb area and then figure out how to run power in the most elegant fashion.

I am still looking for extremely low profile long throw linear actuators. They are “unicorns” I tell ya.


(John) #20

That is the linear actuator - the long part there is a about an inch thick. maybe an inch and a quarter. The piece with the motor on it though makes it thicker. Not terrible but maybe not as low profile as you are looking for.

What would be ideal I guess is if they could put the motor inline with the actuator shaft at the bottom.


#21

@johnr has done a lot with actuators, he might have some other suggestions.

I do suggest thinking carefully about the closing option. If you don’t go with gravity on a dragline like the solo, there is a risk of closing the door on the dog. Or a kid who is trying to climb through.

Closing is always the biggest area of safety engineering for an automatic door.


(John) #22

I have been reading that chicken coop thing for a while but seems to be overkill for what I want to do. But it was my inspiration for this. He is selling that on Amazon btw for $695 I think.