Occupancy Sensor

rfletcher

I need to add an occupancy sensor to a job I'm working on. Since I've never speced one before I thought I'd ask y'all what sensor you'd recommend (or if there are any I should stay away from).

Thanks in advance for any suggestions.

-Ryan

TurnipTruck

rfletcher wrote: »

I need to add an occupancy sensor to a job I'm working on. Since I've never speced one before I thought I'd ask y'all what sensor you'd recommend (or if there are any I should stay away from).

I have done this two ways.

1. I have used feedback from motion sensors from integrated sensors.

2. When there is no alarm info available, I have installed generic alarm system motion sensors to I/O ports of AMX. Motion sets a flag in code indicating occupancy. DSC makes very popular motions sensors that I have used with good success.

Auser

I recommend sensors with both PIR and microwave detection, generally with contact closure outputs wired back to IO's on the controller. Sensors with microwave detection are more expensive, but less likely to overlook occupants who aren't moving about much such as people watching TV and uni students sleeping in lectures.

Make sure that the sensor you choose has a "latch" time which can be adjusted down to a few seconds - you want it to trigger as often as possible when there are occupants in some circumstances.

We've used sensors from Rokonet with success (who appear to have become Risco: see http://www.riscogroup.com/Products.aspx?subcatid=110&id=12).

DHawthorne

Auser wrote: »

I recommend sensors with both PIR and microwave detection, generally with contact closure outputs wired back to IO's on the controller. Sensors with microwave detection are more expensive, but less likely to overlook occupants who aren't moving about much such as people watching TV and uni students sleeping in lectures.

Make sure that the sensor you choose has a "latch" time which can be adjusted down to a few seconds - you want it to trigger as often as possible when there are occupants in some circumstances.

We've used sensors from Rokonet with success (who appear to have become Risco: see http://www.riscogroup.com/Products.aspx?subcatid=110&id=12).

I must concur. Pure IR detectors will even pick up gusts of air from heating systems and air conditioners. I may also add, if pets are involved, you want the kind that will discriminate between human and pet-sized detections, or at least can be masked to not pick up the family cat.

ericmedley

Another thing I've found is to not be too draconian about the programming. At first I tried to be pretty active in turning off lights and whatnot within a short amount of time of 'thinking' the roomo was empty. When I moved the time out to about 30 minutes of thinking it was empty, I quit getting 'Hey, I'm setting in the dark' calls.

John Gonzales

ericmedley wrote: »

Another thing I've found is to not be too draconian about the programming. At first I tried to be pretty active in turning off lights and whatnot within a short amount of time of 'thinking' the roomo was empty. When I moved the time out to about 30 minutes of thinking it was empty, I quit getting 'Hey, I'm setting in the dark' calls.

LOL, that's pretty funny . Definitely good advice too.

I'll agree with Auser's Rokonet suggestion. They're small and work pretty well and trouble free. I also like the Honeywell DualTec series a lot too.

Like TurnipTruck, I prefer interfacing through the alarm if it's possible. You can do a lot by trapping for things like arming status to know if everyone's left the building, door movements from the magnetic contacts, etc...

--John

rfletcher

Thanks for all the advice guys, its been really helpful.

-Ryan

jcerecke

Got a question about programming one of these into the system. I need to turn off some lights either an hour after movement is detected, or an hour after a button has been pressed on the touchpanel. For the touch panel would I just create:
BUTTON_EVENT [dvTP,0]

Or could I let the Touchpanel itself handle the 1hour timeout and make it send a string or something? TP is NXD-500i.

Obviously I'd create an event for the I/O PIR trigger too.

Cheers

DHawthorne

Your timeout is going to need to depend on how the area is used. If it's a hallway or some other transient area, it should be no longer than it would take for a slow person to pas through. If it's an area where people may be sitting quietly for a while, it has to be much longer ... but even that depends. An office might not need more than an hour with people getting up and moving about, but a reading room in a private home might have to stay on for several hours (in which case you may need something outside the room to detect them leaving, just so lights won't stay on too long). It's really going to be something you need to play with and adjust. One thing I have done in the past is to add a door contact so that lights come on when the door is opened, rather than requiring them to step inside the room and into range of the detector before they have any light. And yet another factor is sunlight ... no point turning on lights where there are a lot of windows at high noon. And don't forget the pets.

Whenever I have done this, I have stuck to public areas and hallways, leaving more private areas for manual control. Spaces people live and work in they prefer to have under their own control. Passing through, well, not so much.

I've always used the Napco motion sensors with PIR and microwave. They make a lot of various models for different environments. An alternative to motion sensors is some type of RF proximity device; I understand that is how they do it at the Gates estate. The downside is that people have to carry something around with them, which would be fine for an office, not so great for a home.

Bottom line: there is no hard and fast answer. It is going to vary greatly with the environment.

PhreaK

The other thing to remember is that if you've got any mics and a DSP in the job, depending on the model, that'll also give you an acoustic sensor through gating / signal level reporting. Combine that with motion sensing, door switches, user interface interaction, RF tracking etc and you can start to get pretty good at guessing where people are (or aren't) within an environment.