Room Occupancy Counters -- Need Help!

Sounds like you have a pretty kick-ass setup, Jon.

...could even sell the kit for the lazy people and program the chips for them...
Something tells me that there is nothing about this project that would appeal to lazy people.

If I want to have the lights follow every step I take, I'll just wear a miners hat with a light built-in.
A new WAF all-time low. ;)

...All beds in my house have sensors on them that not only tell when someone gets in them, but how many people are in them as well...
...and if there are three or more people in the bed, the red lights, strobe lights and bed vibrator come on. LOL
 
Seriously - that's taking automation to a whole new limit... I'd have to win the lottery and become at least partially disabled to spend that much time - though it's awesome to see (I'm one of those lazy people!).

And in a half joking, half serious curiosity - with sensors on every bed, what happens when someone - um - becomes "physically active"? Do the lights flash on/off and confuse the security system? Or did you sacrifice such activities to make more time for automating?
unsure.gif
 
I use motion sensors in the bedroom to detect when someone gets up at night, and turn on the lights (10% or so) in the bedroom and bathroom automatically.

I definitely wouldn't mind trying this, it has a lot of potential.
 
Great work you're doing Jon. Sometime when you get time, it would be cool if you could provide more information on your servo motors closing your shutters. This comes up from time to time on the forums, but I haven't seen anyone previously figure out how to do it.

Deane

Have you seen the new shades from Lutron that are motorized? Not a DIY solution, but pretty neat.

Venetian Blinds with Intelligent Tilt Alignment™

http://www.lutron.com/Products/ShadingSystems/SivoiaQSWireless/Pages/Overview.aspx
 
All beds in my house have sensors on them that not only tell when someone gets in them, but how many people are in them as well. Using the household count I can see if everyone in the house has gone to bed, and immediately turn off lights if they are. If only one out of the two people are in bed, then turn off the lights just in the room where they are sleeping and when someone walks into the bathroom adjoining that room, dim the lights on. When that person finally gets in a bed, then shutdown the house. If someone gets up int he middle of the night, turn on guidance lights. If all people are up, dim the lights on gradually if they were sleeping for more ten 30 minutes.

As one that has always strived for an accurate bed sensor, I'd love to hear your ideas. I tried many solutions, but the only one that was accurate and repeatable is a weight detector device which I've built under the center leg of the bed. Its pretty accurate at detecting person or no person, but certainly not how many people. I agree, its an important sensor to get right.
 
Why isn't there a cheap way to count the number of people entering and exiting a room using IP based cameras and video detection software? The detection software could run on a PC and send the data out a serial port for integration with other systems.
 
<tinfoil_ hat>I thought about that, using Vitamin D, but personally, I don't want any cameras in the house</tinfoil_ hat> ;)
 
Seriously - that's taking automation to a whole new limit... I'd have to win the lottery and become at least partially disabled to spend that much time - though it's awesome to see (I'm one of those lazy people!).

And in a half joking, half serious curiosity - with sensors on every bed, what happens when someone - um - becomes "physically active"? Do the lights flash on/off and confuse the security system? Or did you sacrifice such activities to make more time for automating?
unsure.gif

Lol well, to answer your question, unless you have a habit of constantly getting off the bed or standing next to the bed when you are "physically active", the downward pressure on the bed won't change fast enough for the sensor to report it. My sensor takes the average force, once every second, and averages that across the last 10 seconds. So any sudden change in downward force won't cause any effect unless it is constant for at least 10 seconds. In other words if the bed is rocking, the average from all 10 seconds will never be enough to report a substantial change. For anyone interested, I have a writeup here on how I did it: http://www.mysmarthomeblog.com/page--15.html
 
Why isn't there a cheap way to count the number of people entering and exiting a room using IP based cameras and video detection software? The detection software could run on a PC and send the data out a serial port for integration with other systems.

My biggest problem I have found with this is a) the cameras are so big and hard to mount/wire in some locations on the wall, and B) Lighting plays such a crucial role with the accuracy. I have 5 camera setup around my property and the biggest thing I have to worry about is light screwing with the motion. For example, headlights. I use ZoneMinder (Free, open source) and it works great, but I can't rely on it enough to do anything besides capture video and report back when an alarm happens. It's a cool concept though.
 
I am heading on vacation tomorrow for 10 days to the Caribbean, but when I get back I'm going to work on getting these door sensors/counters installed. Hopefully I can take some pictures/video of my door frames and get you guys to help me figure out the different pieces together that make it up. The paint makes it all look like one big piece even though I can see some cracks here and there throughout the house. I just want to know before I start trying to pry boards off Lol
 
Done! I just finished the first door frame after weeks of testing and contemplating on the best way to do this. It works perfectly. Lights are instant on when entering, and instant off when leaving :) I have about 14 more door frames to do. Once I get a few more I'll start implementing a bunch of different features like music following, whole house counting, guest mode, etc. I'll do a writeup on my website in the coming weeks as well. For now, here are a few pictures. They were just taken, so it's dark. I need to take some in the daylight. Thanks everyone for all your encouragement!

door1(1).JPG


door2.JPG


door3.JPG
 
I think the only issue I see for this is, in my house, it's a bit more OPEN, so I would only be able to implement this on upper floor rooms. My bottom floor has too much traffic in too open of a space. I'll have to rely on things like pressure pads on seats in order to tell if people are in the room.

Great idea! Great implementation! Would you be putting a kit together at some point? I'd love to put something like this in my upper floors. Especially the bathrooms!

--Dan
 
Back
Top