Occupancy Sensors for Automated Lighting Control

[rdog]

I am planning a whole house automation system for my new home that is now under construction and I want to use sensors throughout the house to control my lighting on/off (mostly off). I am most interested in being able to turn off lights that were left on in my kids bedroom, master closet, laundry room, bathrooms, office…and less so in the kitchen/great room area of the house.
Based on the reading I have done it sounds like an occupancy sensor (not a motion sensor) is what I will need to accomplish this. I am having a difficult time finding some answers related to occupancy sensors, and I hope someone might be able to help me out.

After reading the forums, I have a short list of sensors I am considering:
* Visonic Disc ET
* Visonic SRN-2000
* SensorSwitch CM-9-R

So here are my questions:
1- Does anyone have experience with any of the sensors I listed? Would you recommend one over the other? Or is there a different sensor you would recommend?

2- Do ceiling fans affect an occupancy sensor? If I turn a fan on, will it cause the sensor to sense motion until it is turned off? If fans do affect an occupancy sensor, what is an alternative solution to turn off lights in a room with a ceiling fan?

3- What should I take into consideration when choosing a location to mount the occupancy sensor? Windows? HVAC vents? Anything else?

4- Any difference in function/performance of a ceiling mounted occupancy sensor vs a wall mounted?

5- Other than occupancy sensors, what else have people done to automate turning lights off?

 

[wuench]

I am working on this too, but so far I ordered a CM-9-R and had to cancel it because AO was having trouble getting one in. My plan is to use this in a bathroom.

For closets, furnace room, etc, I am just using a recessed door sensor instead. Door open/light on, door closed/light off. That is working pretty well. To minimize the delay, I have these wired directly to the Elk's main inputs and have the fast loop response set on those zones. The delay is a little less than 1 second, using ISY/Insteon lighting. So that's probably the best that can be achieved with Elk/ISY/Insteon.

In the past I used to use x10 wireless IR and x10 switches in the bathroom. There the delay was 2+ seconds. It was a little slow, but still workable. For guests, it still provided the wow factor and I got used to it. So I think the 1 second delay for the above will be acceptable, especially if I can get the occupancy sensor to extend out the door a little to pick up someone on their way in.

 

[programmergeek]

Im my experance occupancy sensors dont' really work well in long period occupied areas. For example if you stop moving they no longer knwo you are there. At least the ones I have tried. They work good in halls bathrooms where people are in and out and moving all teh time not sitting ins achair motionless.

 

[Ranger Digital]

If a ceiling fan is in the "motion" path of a sensor, ABSOLUTELY it will affect it. Its motion.

 

[wuench]

Im my experance occupancy sensors dont' really work well in long period occupied areas. For example if you stop moving they no longer knwo you are there. At least the ones I have tried. They work good in halls bathrooms where people are in and out and moving all teh time not sitting ins achair motionless.

For those type of areas an occupancy sensor with a microphone might be better solution. That is what many office environments use, because the IR can't reach into all the cubicles, etc. So better for the bedrooms, etc. Not good in bathrooms where you have a fan, or maybe rooms with other noise sources like TV, etc..

But some of it is also adjusting your expectations. You have to have the expectation that if you are still (and not making any noise) for a period of time the lights will go off and you will need to move around if you want them to come back on.

 

[mustangcoupe]

But some of it is also adjusting your expectations. You have to have the expectation that if you are still (and not making any noise) for a period of time the lights will go off and you will need to move around if you want them to come back on.

I think this is only true for sensors that can not be intergrated into a HA environment.

I use my motion sensors from my ELK, UPB switches and ELVE to control my lightning in a few places, motion sets a timer, and keeps that timer set until it sees no motion and the time runs out then it turns off the light. People have talked about using other sensors weight/floor, ultrasonic sensors etc... when you start using multiple technologies with a smart set or rules/conditions you get around the need for adjusting your expectations. one easy thing I have recently done is in my kitchen set the light to 99% when turned on via motion, and if I am going to be in there for an extended period of time I press the switch, to set it to 100% which the software interprets as keep on, until I turn it off.

 

[ano]

First, I really have to agree with others in that motion sensors really aren't a good solution in most cases. They won't go on fast enough in most cases, and they will both turn off when you don't want them to, and not turn turn off when you want them to, so, you have been warned...

Having said that...

2- Do ceiling fans affect an occupancy sensor? If I turn a fan on, will it cause the sensor to sense motion until it is turned off? If fans do affect an occupancy sensor, what is an alternative solution to turn off lights in a room with a ceiling fan?

I use many types of occupancy sensors (IR, microwave, etc.) and the answer is no, in 99.99% of the times, fans won't bother them. Very occasionally, fans blowing on other things, like curtains, may trigger them occasionally.

3- What should I take into consideration when choosing a location to mount the occupancy sensor? Windows? HVAC vents? Anything else?

You have to know the kind of sensor, and read the instructions. IR sensors typically trigger when you go across their field, micro sensors trigger when you go toward or away from them. The best sensors use multiple technology.

4- Any difference in function/performance of a ceiling mounted occupancy sensor vs a wall mounted?

Definitely, but which is best depends upon your goal.

5- Other than occupancy sensors, what else have people done to automate turning lights off?

For it really to be successful, you need to use several technologies operating together. For example, in my family room I detect if the TV or stereo is on. More reliable than movement, as you don't move a whole lot during a movie or TV show. In the bedroom I use a bed sensor to see if a person is on the bed. In 99% of the cases, sensors like this are far more effective then a motion type sensor.

 

[Ranger Digital]

If a fan blade is in the path of a motion sensor, it WILL trigger the sensor. A sensor senses, a fan blade moves, thats motion. It gets sensed.

You may be referring to a motion sensor mounted on the wall and a ceiling fan on ceiling. I have never had a problem with that. Why? The motion of the fan is not in the path of the sensor.

 

[ano]

If a fan blade is in the path of a motion sensor, it WILL trigger the sensor. There is no if ands or buts. A sensor senses, a fan blade moves, thats motion. It gets sensed.

You may be referring to a motion sensor mounted on the wall and a ceiling fan on ceiling. I have never had a problem with that. Why? The motion of the fan is not in the path of the sensor.

Not quite true. There are several varieties of "motion" detector, so lets look at a few.

The type initially listed in this discussion is the most common type, they use a technology called passive infrared. Most use an etched film sensor with "fingers" of sensitivity" These sensors respond to a rapid change of received IR when compared to the background, which can be looked at as a change of radiated temperature. So they detect because they are not only detecting your body heat, but its movement across fingers. For a detection to occur, the moving object needs to be warmer or cooler than the background, and it needs to be moving locations. Humans moving usually meet those requirements, but spinning fan blades do not, especially because they are typically the same temp as the background.

I live in Phoenix, have an ceiling fan in every room, and many run all the time. I also have motion detectors in all room including the ceiling kind, and I've never had a ceiling fan trigger a motion detector. There are cases where a fan blowing curtains or moving balloons could trigger these, but even then, the sun or a sunny window also is involved causing the movement of reflected light(heat) to the sensor.

A second type of motion sensor, not as common, is microwave. These sensors broadcast a very low power stream of microwaves and use doppler shift to tell if an object is moving toward or away from the sensor. In this case, because a fan moves at a very steady pace, even when starting or stopping, I've never seen this trip a motion detector. In my house I use sensors that combine microwave AND passive IR so if fans triggered these, I would know it.

So, unless you mount these sensors incorrectly directly ABOVE a ceiling fan as I'm sure the instructions tell you not to, I don't think you'll ever have a problem with fans with these two technologies, and even if you did you still wouldn't have a problem. Remember motion detectors are designed to detect humans, and designed to reject other noise like non-human objects, pets, etc.

 

[Ranger Digital]

I was talking about putting it ON the fan blade, lol. Just kiddding. you are 100% correct Ano. Good points.

I had in my head those Leviton sensors that know if you turn a page of a book.

 

[broconne]

I haven't really gotten a chance to deploy them yet.. But I wired for two sensors, a motion and an occupancy. I plan to use the motion to detect when someone enters a room and the occupancy to detect when they leave. I am hoping that will cut down on false positives.

 

[ano]

I haven't really gotten a chance to deploy them yet.. But I wired for two sensors, a motion and an occupancy. I plan to use the motion to detect when someone enters a room and the occupancy to detect when they leave. I am hoping that will cut down on false positives.

That's another good point. Typically "Motion Detectors" are devices used by alarms to detect when someone breaks into your house. "Occupancy detectors" are used to control the lights. Occupancy detectors aren't so much worried about false alarms, so typically they are much more sensitive. Motion detectors don't want false alarms, because the police wouldn't be too happy. They are a bit hard to find, but some devices do both. I have some older sensors with an input, in motion mode, it takes both a microwave and passive IR alarm to trip, in occupancy mode, either will cause a trip.

Now, back to the original problem, using these things to control the lights. There are a few problems here, first, you want the light to go on as soon as you enter the room. So 1) the sensor must trip upon entering, and 2) your controller needs to be fast, and 3) your lighting technology needs to be fast. I can tell you, with an HAI panel, connected to CQC, connected to USB, it takes a few seconds for all this to happen, which seems longer than it is. Motion detectors are SLOWER than occupancy detectors, so I'm not sure I would use a motion detector as the first sensor. My HAI box is instant, and I now connect this to CQC using the TCP/IP driver, so no "polling" delay, but UPB does take a second or two to work.

So my first advice will be to try this all out, to make sure these delays are not TOO long for you. HomeSeer connecting to HAI must POLL, so this adds an extra 1 or 2 seconds. This is also true with the CQC UDP driver. Even with about the fastest technology, I usually have to ANTICIPATE entering a room, by using sensors outside a room to turn on the lights for that room. In other words, when you walk down a hall leading to the bedroom, turn on the bedroom light. By the time you enter the bedroom, its too late.

So you got the light on, now what? The next challenge is keeping the light on with people in the room. When people watch TV or read, they really move very little, so what often occurs is the light shuts off. How long do you make the timeout? 15 minutes? 45 minutes? An hour? Its here where you need extra input, like a bed sensor, TV power sensor, etc. Without that significant other gets really mad.

And last, you need to turn the light off when everyone leaves, but remember, they just moved to leave, so you 45 minute, hour timer has just been reset. Even worse, is when the light was turned off on the way out, but do to the delay talked about it turns back on when the person leaves. VERY annoying. These are all problems you must solve for a good solution, and it gets much more complex then it first appears. Let us know how it works out.