car in the garage?

[dBeau]

What's the best way to determine whether a parking space is occupied or not? In particular, I'd like to know if there is a car in a stall in the garage. The sensor needs to interface with my Elk, so a dry contact output would be ideal. The driveway sensors I've seen all seem to imply they detect motion (not helpful). Will they detect presence instead? I am not interested in solutions that require the car to be tagged in any way (ie no RFID like solutions).

Thanks!

 

[signal15]

Get a beam sensor. Automated Outlet has them. If you have a support between your stalls like I do, you can have one beam going from the wall to the center support, and another on the other side. If not, put the unit about 24" off the ground at one side of the garage door, and the reflector kitty corner in the corner of the garage, so it cuts the car diagonally in two.

 

[roussell]

You could also use a Maxbotix ultrasonic sensor placed at the ceiling pointing down toward the middle of a parked car. If the sensor reports the distance to the floor, no car. You can also record the sensor distance to the top of all of your vehicles so you know which one is parked in which space. I suppose theoretically you could place it at the opening of the garage door and have it capture the profile of the vehicle as it's passing underneath, then you could not only determine the vehicle, but if it's coming or going based on the recorded profile. You could probably also determine if the vehicle was backed in or pulled normally. Probably couldn't do all of that with the Elk, but maybe the one in the middle of the bay would work.

Terry

 

[kd5crs]

If you always park in the same place, you can use the Pressure-sensitive-mat-hooked-up-to-a-DS10A trick. Then you just secure the mat to the floor and park on it.

The mats I use come from Harbour Frieght. Never tried this application though, but it should work.

I think in my garage a photobeam would be vulnerable to being kicked, because there is no center support to mount one end on.

EDIT: Just saw you're using an Elk, so you won't need the DS10A. Just wire the mat up as a NO sensor and you are good to go.

 

[dBeau]

I think both the pressure sensitive mat and the ultra sonic sensors are great ideas. I looked for the mat at Harbor Freight... seems like it would take some work to get it ready to be parked on. With easy access to the rafters, an ultrasonic sensor would be perfect, out of the way, easy to install. Is there a product out there that is ready made for the application (dry contact outputs with adjustable distance sensitivity)? I'm sure an Arduino would be up to the task, but that would take a lot of effort.

So that leaves me with the beam sensor. In principle I like the idea and I think I can make it work. The "PR-1B" at automated outlet is a "direct reflective" sensor with a 1m range. So long as I can mount it near enough to the car it sounds perfect since I dont need to worry about also mounting a reciever or reflector. Does anyone have any stories to share about this device?

 

[ano]

With easy access to the rafters, an ultrasonic sensor would be perfect, out of the way, easy to install. Is there a product out there that is ready made for the application (dry contact outputs with adjustable distance sensitivity)? I'm sure an Arduino would be up to the task, but that would take a lot of effort.

Yeah, too bad those ultrasonic sensors don't have a way to set a distance and output if the distance is greater or less than that set point. As it stands now, the outputs they offer would be pretty hard to use.

 

[signal15]

What do the ultrasonic sensors use to report the distance? If it's an analog voltage, I believe there is a way to use one of the 16 onboard inputs in the ELK M1G to monitor the voltage and do things with automation rules based on that voltage. I seem to remember reading this, but have not tried it.

 

[tmbrown97]

very curious here - if you see E's post on the master list of technologies, one contains a link to the ultrasound sensors - very affordable. It would need some sort of range setting (thinking my suburban with the ridged roof and ski racks) - but would be an awesome solution for car detection.

 

[roussell]

The Maxbotix sensors report the distance in inches to the target - up to 255 inches. It reports it via a proportional analog voltage or a serial data stream. In either case, you would use the Elk to provide the range logic. They're a few examples in the archives - such as Brave Sir Robin's pretty nice salt level sensor in his water softener built using one of them.

I've used them to monitor the water level in a Koi pond and as a package sensor for our front stoop. I don't use an elk for the logic, but I would imagine that it would work fine.

Terry

 

[BraveSirRobbin]

I'm going from memory here, but you might want to scratch your head and think a bit before using the analog output of the Maxbotix sensor with the Elk. The Elk has a pull-up resistor that could influence the reading, especially if the source doesn't have enough current drive to overcome it.

Also, I believe the change in voltage per unit distance is also very small. That plus the Elk's analog to digital converter is only eight bits. So that means the Elk can only reigister a resolution of 12 volts/256. You would have to compare this number to the volts/unit distance the Maxbotix puts out and see if you would have enough resolution for your application.

 

[roussell]

Good points BSR. I use the serial output and haven't played with the analog portions of the sensor at all.

Terry

 

[Lagerhead]

. . .think a bit before using the analog output of the Maxbotix sensor with the Elk. The Elk has a pull-up resistor that could influence the reading, especially if the source doesn't have enough current drive to overcome it.

Also, I believe the change in voltage per unit distance is also very small. That plus the Elk's analog to digital converter is only eight bits. So that means the Elk can only reigister a resolution of 12 volts/256. . .

Exactly right. An M1 zone can measure resistance rather readily, and if you are willing to cut the pull-up resistor , you can measure voltage. Anything else will have you pulling your hair out. And then you have the limited resolution. Even Elk, with the M1ZTS, does not use analog zones for temperature sensing. In short, keep your expectations low for the M1 "analog zone" definition.

 

[signal15]

Here is some info on the analog output for the Ultrasonic sensor:

http://www.maxbotix.com/MaxSonar-EZ1__FAQ...._voltage_output

Looks like you need to multiply the voltage by 100 to get distance in inches. Which means the ELK would have to be capable of monitoring 0-2.5V in .01V increments. ELKRP displays only to .1V accuracy. Does anyone know if internally the ELK measures more accurate than this?

 

[Lagerhead]

ELKRP displays only to .1V accuracy. Does anyone know if internally the ELK measures more accurate than this?

The theory is, roughly, 0V-13.9V is "full scale" and this range corresponds to 0-255 "analog number," or about 0.5V per bit.

But it doesn't matter because whether you use rules or the ASCII command protocol, you can only get 0.1V resolution.

 

[PaulD]

I don't have an Elk so I may be missing something here. Why is everyone so concerned about high levels of accuracy? If you are only trying to determine if your vehicle is in the garage or not, there should be a measurable difference in voltage between the floor of an empty garage and a car roof parked in the garage. Install a sensor, measure the voltage with a car in the garage and that becomes your trigger level.