PIR Sensor-Based Intruder Alarms

It must be an odd building to have a flat roof covered with lead; roofing felt stuck down with bitumen would be a better engineering solution and would solve the theft problem, making the PIR security system redundant. Has the building been properly designed?

I think it is rather a beautiful building, actually. When it was built, in the 13th century, it was probably at the cutting edge of design. As it is now Grade 1 listed, changing the roof to felt and bitumen is not an option.

http://mugginton.info/church/

Right. So the roof isn't flat. OK. And it's Grade 1 listed now. OK.

1. Has the local Crime Prevention Officer been round for advice?
2. How about anti-climb paint on the guttering downpipes?
3. What does the building's insurance company's representative have to say?

1. Yes. 2. No. Ladders are an equally effective way of getting up there. 3. Our insurance company recommended this firm.

The PIR (Passive Infra-Red) Sensor is a pyroelectric device that detects motion by measuring changes in the infrared levels emitted by surrounding objects. This motion can be detected by checking for a high signal on a single I/O pin.

Theory of Operation

Pyroelectric devices, such as the PIR sensor, have elements made of a crystalline material that generates an electric charge when exposed to infrared radiation. The changes in the amount of infrared striking the element change the voltages generated, which are measured by an on-board amplifier. The device contains a special filter called a Fresnel lens, which focuses the infrared signals onto the element. As the ambient infrared signals change rapidly, the on-board amplifier trips the output to indicate motion.

The PIR Sensor requires a 'warm-up' time in order to function properly. This is due to the settling time involved in 'learning' its environment. This could be anywhere from 10-60 seconds. During this time there should be as little motion as possible in the sensors field of view.

The PIR Sensor has a range of approximately 20 feet. This can vary with environmental conditions. The sensor is designed to adjust to slowly changing conditions that would happen normally as the day progresses and the environmental conditions change, but responds by toggling its output when sudden changes occur, such as when there is motion.

The above write up is for your info. only. May be the system is inadequate for your requirement.

They have demonstrated that if a person moves very very slowly in the range of a PIR device,it won't detect. The secret is rate of motion(speed).

Thank you. In fact we have spent over £4000 on this unsatisfactory system, so I do not propose to take over and let our installer off the hook. We paid good money for a system to do a specified job.

Can this not be made secure physically? Even with the alarm going, you need somebody to immediately investigate & accost the intruder. He/they may even be armed. You will have to provide 24/7 manpower.

Well, a PIR detector does not really detect motion, it detects abrupt changes in temperature. To test your theory about this "...relatively narrow band of infrared radiation emitted by humans...", sit in a chair within the detection field of one of them and let it settle down; this demonstrates an initial flaw in your theory, with a human in the detection zone the detector changes back to normal state. Then, moving as little as possible ignite an entire book of matches. This should trigger the device, revealing another flaw; it detects heat and changes state when there is no warm blooded animal moving within the detection zone.

Are you looking for resoultion to the trouble, or support for your arguement?

"...each sensor may trigger several times a day and it is the job of the software to sort out the alarms..." What, with something like a Kalman filter? This sound like a good way to ridiculously complicate things unnecessarily.

If a PIR detector doe not detect motion, then a speed gun does not detect speed, and a Wheatstone bridge doe not measure resistance. Your suggestion that a box of matches can cause a signal from a PIR output is interesting: are you suggesting that PIR sensors are not tuned to the radiation emitted by humans? Have you actually done the experiment without moving?

Of course I am seeking a resolution to the problem of false alarms. I worry that what the installer has told us (as in "...each sensor may trigger several times a day and it is the job of the software to sort out the alarms..."), is complete rubbish. This remains to be tested against your expert opinion.

phph001, you said "If a PIR detector doe not detect motion, then a speed gun does not detect speed..." PIR Detector = Passive Infrared Detector. A windblown tumbleweed is certainly in motion, but when the 'weed passes thru the detection zone of a PIR detector, the device will not change state... no infrared radiation is being emitted to detect. However the speed of that same tumbleweed will register on a speed gun. Gotcha!

You asked "Have you actually done the experiment without moving?" Yes I have. A caveat: as you stated the equipment type is unknown to you, thus unknown to us. There may be PIR detectors I am unfamiliar with (of course there are), but all I have ever seen for this sort of application are the type described previously.

I agree with your comment "I worry that what the installer has told us... is complete rubbish." Is your contractor offering this solution at no charge to you? If yes, I would suggest you let him try. I seriously doubt it will fully resolve your issue, but it certainly won't hurt. As I said, it seems a complicated and (probably) irregularly performing solution.

But how do you know that it was not tumbleweed causing the false alarms? Gotcha! :)

If a 'weed is causing the alarms, he should be arrested for trespass AND speeding!

At least no one is going to de-roof the church while they're up there, so let them take their time.

Sorry this doesn't answer the Q...
The problem is that heritage England (or whatever the body is called) insists on lead. EPDM would be vastly better and look V similar (the detailing can can even be laid to look like lead), it would not attract metal theives. It will also outlast roofing felt and bitumen which is about as much use as a tissue paper font.
The answer is to put pressure on the authorities to change their nonsensical ruling.
Churches were built and modified over hunders of years, an inconspicuous change in roofing material should be aproved. It would also reduce the loading on the roof.
Del
_{(IMO Metal theives who damage the infrastructure or endanger life should be prosecuted for treason, that might deter them)}

Thank you. I was not aware of EPDM as a possible material, and am not at all sure that English Heritage would countenance this, but it is worth investigating. How long will it last? I do know, however, that a neighbouring listed church has used stainless steel. How they managed the flashings I am not sure.

This is the company I got mine from, they have plenty of info and installation guides etc on their site and the home page say proven to last in excess of 50 years.
I've done the roofs of 2 extensions on my property and plan to do the porch at the front next year.
Del

Thank you.

Thank you. The wording does suggest that not all sensors are set to register a narrow band of radiation frequencies.

Do you think you could stabilise this. If the outputs from the PIRs had been analogue I think you might have had a chance but the outputs are usually digital. If both PIRs were to detect then with both inputs to the diff amp energised you would see nothing i.e. you would miss the intruder. The chance of a high wind activating both detectors at the same time is I believe unlikely. I'm not convinced about the heat shimmer from the roof. The PIRs react to sudden changes not slow changes. Common mode initiation is reduced by the 2 PIRs in series.

Selling the lead and putting a rubberised roof on is the obvious answer.

Read the Wiki link that I posted, I am sure they know EXACTLY what they are talking about.....it makes good sense.

General comments and observations:

Sensors voting, n-out-of-x sensors reporting: Do pigeons really travel individually in Great Britian? Ravens or sparrows? An individual intruder on the roof... how many sensors will he trip at once?

Body temperature of raven: about 38°C. Body temperature of pigeon: about 42°C. Body temperature of grey squirrel: about 38°C. Body temperature of human: about 37°C.

Thermal radiation emitted by human, wearing an insulated winter coat and insulated trousers, outdoors in -8°C weather: Dunno, but you get the idea.

Did the supplier also supply the software that's supposed to sort the false alarms?

Is the supplier not wanting to resolve the issue or what?

Was he contracted to supply a working solution or just install hardware in accordance with somebody else's design?

PIRs are intended for indoors???

I suppose this gets rid of the outside spots / floods and wallights that people put onto the outside of their houses as intruder deterents.

The PIRs can be desensitised or as I've suggested run in pairs.

A beam detector is a good way forward.

Best way is put a voltage onto the lead roof and then you can usefully collect the culprits at grade thus removing the danger of climbing up to them

Thank you. I shall have to challenge you, though, on your statement that PIR detectors are intended to be used indoors. My internet researches reveal this firm

http://xtralis.com/product_view.cfm?product_id=35

with a range of outdoor PIR sensors. I do note that this particular range does not specify a model with 180 degree coverage either. I am also intrigued by the idea that a PIR detector can sort objects by weight. Surely the detection is by intensity of radiation, and a small object close to will radiate the same as a large object further away.
My problem is that the firm seems unwilling to specify the make and model of the sensors, let alone what the software is supposed to do, and I am not too keen on getting a ladder out for that purpose alone.

Surely these details were spelt out in the supply contract.....

How long has this system been in "service"?

What leverage do you have over them?

You are right to ask. Sadly, the engineer who set up the order died shortly afterwards, and I have not found a full specification in his papers. Essentially we ordered a system to do a job, like buying a car for transport without asking whether it's got alloy wheels or not.
The system has been in service for less than 6 months. We have accepted that the proposal to improve matters by installing new software. I have no idea how the new software differs from the old software.
As for leverage, we have no more than English law allows. I want to get a chartered engineer in to review this, but my church council is wary of spending the money. The church architect is not skilled in this area.

That is sad.

Not good form to blame then....

What was on the purchase order?

Has the new software been installed and commissioned yet?

You asked whether it was possible for heat to give false triggers. I figure yes but there has to be a way to reduce these.

Ask the supplier if he believes in the Hereafter or not!!!!!

Suggest you don't go for 180 deg pattern. Wider the spread the more chance of false alarm. Its proved that burglars have a good look around before they start work. They will soon trip a detector.

PIRs are intended for both in and out door applications. PIRs are susceptible to false alarm through windy conditions.

A PIR cannot really detect weight its just a generalised statement for a small lump, a mid sized lump or a really large lump. Its looking for radiated heat not mass.

This is not someting I am well versed in, but Don't the sensors use a doppler shift to try and eliminate false alarms?

Also, this is a theory to check. The world has been "blessed" (its an act of GOD not green house gas afterall) with some incredibly hot days. Could the sensors be detecting the heat radiation (waves) coming off the roof. Lead, once hot, tends to be highly temperature reflective more so than some other "normal" roofing materials.

Just a couple of left of center thoughts.

A Doppler shift is used to detect motion by sensors which rely on ultrasound emitters and detectors, not by PIR sensors, which react to emitted infrared radiation.
I think we are all in agreement that a lead roof can generate rising air. However, the roof itself, which is radiating heat, is not moving. I am asking for expert opinion on whether the heat from the rising air (which I understand to have a wide range of radiation frequencies) is capable of triggering these sensors.

I do not think the rising air should be emitting IR at high levels. However, the waves in the air caused by changes in the refractive index of the different temperature portions of the air will cause apparent motions in the images of the different temperature objects behind the air portions. This may seem like motion of the air, but will certainly register as motion of the IR image.

Read Post #37, it explains how this can be fully adjusted for.....

What about a 10,000 volt DC charge to the roof lead with the negative connected to the gutters and an alarm set to trigger whenever there is an electron flow.

The police can pickup the dazed would-be thief laying on the ground after he/she grabbed the first piece of roof.

Unfortunately in this era of ridiculous political correctness it is a bigger crime to protect yourself than it is to predate on others.

Check out this security system "Smartwater" it puts an indelible coating on anything so that it can be traced forensically.The perpetrator will also have non erasable residue from contact.

http://www.youtube.com/watch?v=sKl9v9Tihis

Thank you. We had Smartwater on the roof already, as required by our insurance company, before the lead was stolen. We also have a warning notice on the gate. All the thief has to do is roll up the lead with the Smartwater dots on the inside. Moreover, it is not entirely indelible, and has to be renewed every few years.

Considering that the OP states that it has been claimed that the "false alarms have . . . been due. . . to heat waves coming off the roof", it seems appropriate to at least seriously consider this explanation as it may be coming from the installers who may have actually thought seriously about the problem. No, #37 does not deal with that problem, but only the slow large surface concerns. The "shimmer" and apparent motion of objects viewed through the unstable, thermally turbulent hot air pockets can cause readings to change in less than a second and over a small area. Moving the detectors so that the reading paths are more perpendicular to the plane of the roof and thus see the roof from a greater angle might help, but this presumably should be the obvious testable modification to be looked at by the installers, if they were the ones offering that explanation. As is usual in these blogs, more detailed information would make speculation less fun but more productive of solutions.

I read more into the Wiki article than you appear to.

I read that two PIR sensors with a slightly different view of the same area will cancel out the same temperature differences and only show individual "movement" of a burglar for example, that only one sensor sees for at least part of the time and not the other....There was no mention (why should there be) as to how fast or slow the (warm) air needs to be as long as they are basically similar.

Of one thing I am sure, that reputable companies can make an alarm system that reacts to intruders and not to temperature problems, or it would be common knowledge on the web....it is not! Most links on the subject refer to this blog........

I found another link which has some good ideas and tips, one of which is combining a microwave sensor and a PIR in the same unit, both must trigger for an alarm to be made. Sounds good to me....

http://www.diy-alarms.co.uk/catalog/which_pir.php

I assume that your theft problem is occurring at night.

Set your system to discontinue reporting during daylight hours....or at all.

If I were facing your problem, I would focus on the human nature of theft. These goons are stealing your roof because it is easier and quicker than working for the money. Make it less easy and comfortable for them.

Besides setting your system to night time only, run a water line to the roof. Install some solenoids and waterspray misting nozzles in inconspicuous locations. Zone these solenoids to your detectors, and set one minute spray periods. Add a strobe light up there so they know they've been had.

It's typically cool at night there right?

What goon is going to want to steal roofing with a strobe light flashing while getting soaked ?