10
\$\begingroup\$

I grew up incorrectly believing that motion detectors (as part of burglar alarm systems) were designed to detect movement, when in actual fact it is the movement of a heated (like body heat) object which causes them to trip. I also understand there are some which are so sensitive they can even detect the body heat from fingers typing on a keyboard!

Question #1: Are there motion detectors which just detect motion, without the moving object having to radiate heat?

I ask this question because the above sensors can be blinded by a small glass picture frame placed over them, with a long pole. It would be much more difficult to bypass if they detected movement.

I assume we should exclude sonar sensors since they are easily bypassed with a thin layer of fabric, such as wearing a bed sheet like a ghost, according to Myth Busters.

Question #2: Aside from microwave motion detectors potentially being problematic when they can detect movement through solid objects such as wooden walls, furniture, and glass which could lead to false alarms - why are they not used more frequently?

Question #3: Why do some PIRs in alarm systems have three LED lights yellow | red | green under the plastic sensor? I ask as perhaps they detect three different types of movement, so this might be related to the question and not off-topic.

\$\endgroup\$
6
  • \$\begingroup\$ 1: yes, the security industry has many types of sensors to pick from. Always use the right sensor for the right location, or you'll get people opening doors from the wrong side with a vaping pen. \$\endgroup\$
    – Mast
    Commented Jun 13, 2021 at 10:38
  • 1
    \$\begingroup\$ I'm pretty sure glass is not a reliable way to attenuate the IR. I once had a false alarm from a PIR in a room when a deer wandering in the yard came right up to the double pane glass door at the far end of the room. \$\endgroup\$ Commented Jun 13, 2021 at 15:08
  • \$\begingroup\$ @AdrianMcCarthy Then it was not a PIR, it was likely an indoor microwave detector which can pass right through glass. \$\endgroup\$
    – questioner
    Commented Jun 13, 2021 at 22:10
  • \$\begingroup\$ @AdrianMcCarthy The motion sensor is based on passive infrared detection, therefore it works through clear glass. ... Most modern-day glass used in housing is 2 or 3 pane insulated glass and the necessary infrared heat detection will not pass through it. \$\endgroup\$
    – questioner
    Commented Jun 13, 2021 at 22:15
  • \$\begingroup\$ @questioner: It was absolutely a PIR detector (in fact, it still is). Yes, modern window glass does reflect much of the energy in the IR range back outside, but it's not an absolute filter. (And picture frame glass, as posited in the question, isn't much of a filter at all.) When we complained to the alarm company about the false alarms, they didn't believe us either, until we demonstrated it with a human outside the glass door on a cold morning. \$\endgroup\$ Commented Jun 15, 2021 at 4:24

6 Answers 6

8
\$\begingroup\$

PIR sensors have quite a few advantages:

  • they’re very cheap
  • they use very little power (since they’re passive) – very important for battery powered/wireless sensors. We’re talking microwatts here.
  • they don’t transmit anything (again, because they’re passive), which means they’re easier to get certified
  • they yield few false positives
  • they don’t need calibration
  • they can have a pretty wide field of view
  • they need a single device
  • they are solid state (no moving parts) and have very good reliability
  • they have no privacy or security issues (unlike camera-based systems for instance)
  • they don't see through walls or glass (useful to avoid detection of stuff "outside")

I’d be curious to see a PIR sensor that can actually detect fingers moving on a keyboard at any decent distance (1 or 2 meters), though. Given how a PIR detector works this seems highly unlikely.

Some of the cons:

  • They're usually not very sensitive at distances of a few meters
  • They only detect movement in the plane orthogonal to the axis (it's a bit more complex than that, but it's a good approximation)
  • Most only detect movement along a single axis (depends whether they use 2 or 4 surfaces and the shape of the Fresnel lens)
  • They only detect what they can see, so any obstacles will prevent them from seeing anything
  • They only detect heat sources, usually humans or animals
  • You can't control the exact area they cover

There are many, many other ways of detecting presence or motion, though, like PIR sensors, each has its own pros and cons, and many may not be suitable depending on what exactly you want to detect.

You could use:

  • Doppler (on various frequencies)
  • distance sensors (ToF, ultrasonic, LIDAR…) and measure change
  • vibration sensors (MEMS or piezoelectric)
  • break-beam sensors
  • force resistive sensors
  • thermal imaging
  • shape recognition
  • a combination of any of the above…

There are probably many more, and depending on you exact use case, some or many may or may not be applicable.

\$\endgroup\$
3
  • \$\begingroup\$ I am just confused as to why I never see these in commercial alarms. \$\endgroup\$
    – questioner
    Commented Jun 12, 2021 at 22:40
  • \$\begingroup\$ @questioner Most if not all lack one or more of the pros of PIRs. More expensive, more difficult to set up (requires calibration or requires multiple devices), can't run on battery (needs wiring), yields false positives, narrow field of view (sometimes just a line)... There are plenty of commercial alarms that use other systems, including door/window opening detection systems, vibration detection, and more... \$\endgroup\$
    – jcaron
    Commented Jun 12, 2021 at 22:46
  • \$\begingroup\$ They are called presence detectors and motion detectors: ivoryegg.co.uk/product_guides/… \$\endgroup\$
    – questioner
    Commented Jun 13, 2021 at 22:11
4
\$\begingroup\$

VHF or UHF signals in a quiet channel using a splitter on one of the 2 ports can measure reflected signals off walls and detect reflections up to 10 or more wavelengths by using a Schottky diode and amp to detect changes in level using a constant power amp of ~100mW or so. This is the same as using RSSI on a Return Loss bridge.

I found by accident this worked to 10 yds in a lab when tuning a dipole back in the 70’s and later found commercial solutions doing the same. The idea is to differentiate changes in mV in the reflected “standing wave” off walls and anyone disturbing those paths changes the amplitude up or down so an AC full wave rectifier with a slow high pass filter and a comparator is all that’s needed.

Rev A

IR motion sensors are extremely accurate using pulsed IR to get an AGC average reflection back with a PD and signal conditioner. Then any small change relative to the AGC’s signal, triggers the light on. They are found in big box stores for outdoor lights and are very cheap to turn on floodlights. Just hack the sensor to an Optocoupler for an alarm signal. Then add in battery backup if your grid is unreliable.

  • another cool method is a series of small mirrors to reflect an LED IR Laser beam for interruption but more costly.
\$\endgroup\$
8
  • \$\begingroup\$ So if this really detects movement (even if it's just a book falling off of a shelf) why do burglar alarm companies not use these?! \$\endgroup\$
    – questioner
    Commented Jun 12, 2021 at 19:00
  • \$\begingroup\$ Could you link me to a commercial company which does this, because every company so far seem to do heat sensing PIRs. \$\endgroup\$
    – questioner
    Commented Jun 12, 2021 at 19:10
  • 1
    \$\begingroup\$ Why is it so hard for people to search with new clues? google.com/search?q=rf%20motion%20sensors \$\endgroup\$ Commented Jun 12, 2021 at 20:03
  • 1
    \$\begingroup\$ @TonyStewartEE75 In order to develop Google-fu, one ironically has to go train the mountains where there is limited internet (but not no internet). In other words, experienced the internet when it was not as streamlined as it is now. \$\endgroup\$
    – DKNguyen
    Commented Jun 12, 2021 at 20:21
  • 1
    \$\begingroup\$ @tonystewartee75 most if not all of those search results are actually for PIR sensors, the RF part is just how they send the information (I.e. wireless rather than wired). \$\endgroup\$
    – jcaron
    Commented Jun 12, 2021 at 23:25
3
\$\begingroup\$
  1. Motion sensing is pretty broad.

    Doppler radar can detect motion. Anything that detects changes in light will also detect motion (which is pretty much just the same concept as a PIR sensor but using different components to detect changes in light instead of changes in radiated heat).

I assume we should exclude sonar sensors since they are easily bypassed with a thin layer of fabric, such as wearing a bed sheet like a ghost, according to Myth Busters.

As opposed to a thin emergency blanket for PIR sensors?

  1. That's a form of radar. Radar tends to be expensive technology and also tends to require licensing because it's actively emitting radio waves. It would require use of dedicated radio bands and if not, then it would interfere with the band that all your other wireless stuff operates on.

  2. Why do devices have status readouts?

\$\endgroup\$
7
  • \$\begingroup\$ 1) So what if the room is dark...? I thought rather quickly your heat would leak through the thin emergency blanket? On Myth Busters someone's body heat leaked through a wet suit, special mud, and a naked man being covered in a fire extinguisher did not work. 2) Okay. 3) Why do some PIRs have three LEDs which are different colours, when most just have one, what are they showing with those lights? 4) So what did burglar alarms use then, just heat motion sensors which seem easy to trick? I have a camera in my laptop screen which detects any motion, even a cold dead book... \$\endgroup\$
    – questioner
    Commented Jun 12, 2021 at 19:07
  • \$\begingroup\$ 4) ... which seems like a must better security device to prevent the vulnerabilities and dramas like licencing for microwave radars discussed so far. \$\endgroup\$
    – questioner
    Commented Jun 12, 2021 at 19:09
  • \$\begingroup\$ @questioner 1. It's a matter of degree. PIR sensors work by having two horizontal sensing elements and measuring the rate of change between the two. What they are doing is trying to detect horizontal changes in heat. That means that if you move vertically or slowly enough, you can fool the sensor. Cameras are expensive and require processing. PIR sensors are dirt cheap in comparison. 3. What I'm getting at is there's not too much point asking about things like LEDs. It's just up to the designer. Sometimes I put lots of LEDs on the things I design for indicators, sometimes less. \$\endgroup\$
    – DKNguyen
    Commented Jun 12, 2021 at 19:26
  • \$\begingroup\$ 4. Have you ever tried sneaking past a bunch of PIRs when you don't know where they are? Let alone trying to do some nefarious time sensitive activity in their range? Do you expect something breaking into your house other than a warm-blooded human? And would you pay more for it? \$\endgroup\$
    – DKNguyen
    Commented Jun 12, 2021 at 19:44
  • 1
    \$\begingroup\$ @questioner You can give it a try. Get a few of those motion sensitive lights, get someone else to stick them on the walls and then try to make it past a couple of rooms, grab the golden idol and get back out without setting them off. \$\endgroup\$
    – DKNguyen
    Commented Jun 13, 2021 at 0:05
3
\$\begingroup\$

No other answers yet address #3 - the three LEDs. What you are probably looking at is a dual technology motion detector, not a PIR. These contain a PIR, which only when triggered (to save energy) activates a second technology to detect motion (often microwave) for a few seconds. The PIR activates the first LED, and the microwave the second LED. When they both triggered simultaneously for a period of time, the third LED is activated, and the signal tripped.

However, this is trying to do a different thing from the use case you are concerned about. You are concerned someone might block the PIR's usage - this would still work (as then the microwave motion sensor would never be triggered). The reason for this arrangement is to reduce false alarms.

Note also that most home intruders are not sophisticated opponents who are well versed in defeating home security systems. They're often poorly educated desperate sorts, looking to fund their next hit. The main two effects of an intruder alarm are first to encourage them to pick an easier target (e.g. next door), and secondly to make them exit as quick as possible when the loud noise goes off. There are far easier and more reliable well known ways (which I won't post here) of making sure a cheap intruder alarm does not alert anyone than sneaking past the PIRs wearing something. If your adversary is well versed in defeating traditional home security systems, you will need a far better security system (as e.g. banks, high end jewellery shops use).

\$\endgroup\$
3
\$\begingroup\$

Motion sensing can be done with cameras. OpenV and other technologies can detect motion from image section analysis over time instead of heat changes over sensor sections. This is unlikely to be defeated with a simple mirror unless very very slowly moved into place.

\$\endgroup\$
0
2
\$\begingroup\$

@jcaron's answer mentions Ultrasonic motion detectors and I'll expand on this. They are easy and can be cheap but can see around corners and outside the room; they are not limited to line-of-sight like the PIRs, which can be both good and bad.

An ultrasonic transducer emits an ultrasonic wave (sound at a frequency higher than a human ear can hear) and receives reflections from nearby objects. Exactly as in Doppler radar, heterodyne detection of the received field indicates motion. The detected doppler shift is also at low audio frequencies (for walking speeds) since the ultrasonic wavelength of around a centimeter is similar to the wavelengths used in microwave motion detectors. One potential drawback of ultrasonic sensors is that the sensor can be sensitive to motion in areas where coverage is undesired, for instance, due to reflections of sound waves around corners. Such extended coverage may be desirable for lighting control, where the goal is the detection of any occupancy in an area. But for opening an automatic door, for example, a sensor selective to traffic in the path toward the door is superior.

Cypress Semiconductors Ultrasonic Motion Sensor reference design Cypress Semiconductors Ultrasonic Motion Sensor reference design

Cypress Semiconductors Ultrasonic Motion Sensor reference design

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.