SmartThings Motion Detector Range? I want to limit range so it can be used as a switch

I’m quadriparetic, use a power wheelchair, and like things to be as handsfree as possible. `(Hoping for eventual voice control, of course, but meanwhile here we are.)

I have the opposite goal of many people: I want the motion sensor to check in a very small range so I can use it as a switch without having it turn on every time someone walks by.

Since I’m in a wheelchair I can’t just move it very high or I can’t reach it when I want to.

Also, I can’t easily test a bunch of different configurations, so I’m looking for specific details on ways to create a limited range space.

Right now I have the SmartThings motion sensor on a bookshelf facing up. That works pretty well, but we’re still getting too many triggers when someone walks on that side of the room.

Any thoughts or suggestions for ways to limit the trigger area but still leave it available for completely handsfree use? That is, I don’t want to put it inside a closed box or behind a sliding panel or even blocked by a bookend because I don’t want to have to do physical manipulation to get to it or I might as well just get a rocker switch. `

I could try putting it facing up inside a box without a lid, I suppose (so I wouldn’t have to lift a box lid), but I don’t know if that would work or what the most effective box material would be. So any information on what limits the range of this device would help.


Great question JD…

Every PIR (passive infrared) motion detector has different sensitivities – both range and angle.

Most PIRs work by using a multi-fresnel (bumpy?) lens that creates a “moving shadow” effect on the actual sensor target. Movement across the “visible” field of the sensor fades in and out of focus and that affects the brightness of infrared light on the sensor, thus telling it there is motion.

It’s too bad that these don’t have a “reverse” mode that would project a visible light out from the sensor area through the lens so that you could see a pattern projected on the walls, ceiling, and floor which would approximate the width-scope of the sensing area. But you can possibly guess the area by looking at the lens and picturing what that light pattern would look like. Note: PIR’s don’t project even infrared light or sound, they just passively “see” the heat from people or animals.

It would also be nice if PIRs had interchangeable or focusable lenses, but they would be more expensive. The lens is just a piece of injection molded plastic.

I think that any sensitivity adjustment parameters (dials or configuration options) relate to how much “brightness” that the infrared hitting the sensor must be in order to trigger, which affects the distance from the sensor, but not its width.

Given all that, I’ve not even answered your question :wink:

I think you just have to experiment with placement “map out” the trigger area by testing it while moving throughout the room.

Once you find the edge of the detection area, you know you can turn the sensor away from that area and that will move the edge in the same direction (i.e., picture moving a spotlight, movie projector, or camera viewfinder).

Your idea of putting the PIR inside a box is a good one: You can put a slit or hole in the box in front of the PIR lens, and that could help tune the focal width (the detection width) and let you more precisely “aim” the detection area. Also try different models / brands of PIR, as they have different lenses and distance sensitivities.

I hope I’ve been of some help!
…CP / Terry.

Some people are using the Ubi to voice control their SmartThings. At $299, it’s a bit pricey for my taste.

I built an app called SharpTools for Android and one of the many things I use it for is voice controlling SmartThings. When the phone is sitting on a charger (or if the screen is awake), it is always listening for my voice commands. So I can say things like “Okay Google, lights out” to turn off all the lights before I go to bed (tutorial). Or “Okay Google, set the heater to 72 degrees” (tutorial).

That being said, SharpTools is primarily a set of Tasker plugins (and Widgets) and is designed for tinkerers. You have to setup the voice commands yourself, but I’ve posted a number of tutorials and I’m happy to help.

I guess all that is a long way of saying voice control of SmartThings is possible today! :smile:

I appreciate the response. I know most people would just physically experiment to find what works. But think Stephen Hawkings here: physically moving the sensor two inches to the left requires either assistance from someone else or considerable time and effort. Precise physical experimentation requires more time and more effort. So it’s more efficient for me to limit the number of physical “trials” in any trial and error protocol by getting information in advance. Just a factor in my specific situation.

Knowing that it’s an infrared detector, for example, is helpful because it suggests an insulating material might be a more effective limitation on the range. So maybe we start with an insulated lunchbox rather than a metal cookie tin. :wink:

The specific detector I’m using is the SmartThings brand sold in the shop on this site. So if anyone has specific experience with range factors for that, they’d be much appreciated.

Thanks! I looked into Ubi, but the price/benefit ratio isn’t there for me right now. Anything that has to be plugged in to the wall requires multiple units, greatly increasing the cost.

I don’t have any mobile phone at present as physically they don’t work for me. I have a one button panic button that give me two way voice communication with an emergency service that can call my housemates if needed, so that covers the emergency aspects.

I also have an iPad mounted on my wheelchair which works great for computing, internet, entertainment, texting, etc. Decent voice control and one knuckle touch for most stuff. Excellent battery life. The current SmartThings app has some display problems, but is manageable if annoying.

So I don’t have any android devices at the moment. I’m hoping eventually there will be a smart watch that will work for me, but that may be a ways off.

I’m always interested in new options, especially stuff that actually exists, so I’ll take a look at your software and see how it might fit with what I need. Thanks for the suggestion!

I don’t know of any sensors that have restrictions like this. However, I can think of some options. They would initially require experimentation, but once you have it understood you could deploy more following the same rules.

Two things come to mind

  • use tape on the lens to narrow the area it can see. I believe some sensors allow you to remove the front, such as the Monoprice ones (if I recall). You could put the tape on the inside instead of the outside so it looks better. The SmartThings sensors use a very small lens window, which would make it really difficult to cover up.

  • change the POV completely, not just the “angle on a shelf” What happens if, when you combine it with taping off some of the lens, you place it in a completely different position? For example, if you want something to happen as you approach something, narrow the field of view with tape and then place it on a wall perpendicular to the location. That way, it acts more like a “light beam” or “trip beam” sensor.

As for the IR part, I don’t think you need much to block it. You don’t need “insulating material”. It is still fairly “visual”, so I bet some simple electrical tape on the inside of the lens would work…heck, even masking tape would probably work.

Grab a friend to help you experiment with covering up the lens and narrowing down that field of view and, like I said earlier, you could easily replicate to other sensors.

I’m going to check my one Monoprice motion sensor I have at home and see if what I said before is true. If it is, I’ll do a little tape experiment myself and let you know what I find. The sensors are inexpensive compared to others, so they are good to experiment with. The custom device type that is noted on the boards here works well.

This one (picture below) … or the older original model?

I only have the older original one – it has a “cylinder” shaped lens; the new one has a “dome”.
Either way … good luck! Definitely try to picture what the “visible” area it is by looking at the shape of the lense are where you are pointing it. These are designed to have a very wide area view. Blocking parts of this view with silver duct tape or foil might even be effective, but definitely hard to be precise.

Yeah, I think the larger ones, like the Monoprice, would be better to block off since the lens is further away from the sensor and, hence, bigger. I’m definitely going to experiment when I get home - I’ll even rope my daughter into it as a science project. I’ll let you know what I find.

I have the newer one.

I understand the angle issue, that’s why we put it lying down facing up on the bookshelf. So it’s easy to reach and wave a hand over, but I had hoped it would not pick up people walking by. But it does still trigger sometimes when someone is just walking in that area.

It only seems to be happening to the left of the unit but I’m now thinking it’s the moving past that is somehow moving a shadow across the lens. Just because of the layout of the room and what’s on those shelves if someone is on the right side of the unit they’re likely continuing off to the right or further into the room. But someone going to the left of the unit is likely to turn and walk to the right. So that’s a crossover.

Tape might help narrow the field so it doesn’t pick up the passers by. The good thing about this whole situation is that when we do want the detector triggered, it’s very intentional. So the person, even me, can wave a couple of times to be sure to hit the sweet spot.

I have used other motion detectors in the past to turn lights on, but it was a different situation, I was fine with it turning on as soon as anyone went in the room. This particular project is to control a device on demand, but I don’t want it coming on everytime someone walks by. It’s really just going to be a touchless switch.

I’m pretty excited about the whole idea. Voice control, however and whenever it gets here, is going to be great, but touchless silent control has its place too, especially when you have housemates. :blush:

My housemate got home. I understand insulation may not matter, but as it happens we had an insulated bento box container no one was using that looked the right size, so we used that. (Obviously part of the trick here is obscuring the motion detection without affecting zwave interconnectability.)

Anyway, it’s on the same shelf in about the same position and seems to be working great. Trigger is limited to when we actually wave a hand over the top of the box. The motion detector is lying flat inside the box facing upwards.

I’ve started a separate topic on a very short range proximity sensor since I think I could use several. Meanwhile, though, the SmartThings motion detector inside an insulated box is working well as a touchless switch.

Edited to add a thank you to everyone for their information and suggestions. I will also keep researching the ideas recommended.

1 Like

Ok, here is what I found with the Monoprice motion sensor. Note - you can take this sensor apart and put the tape on the inside. I put it on the outside so it was more obvious for this test. In all tests, the sensor was 5.5 to 6 feet high. (Ignore the crack, I dropped my sensor during testing…still works fine, though.)

Test One- Perpendicular to location
In this test, I simulated what would happen if the motion sensor was placed on the wall to the left or right of the location (as in the left or right wall, not the same wall as where you want to sense). If I covered the lens like below, it essentially made the FOV straight in front and to the open side of the lens. Notice I covered about 35 to 40% of the lens, not 50%. I tested with 50% covered and the detection boundary was well past the middle of the sensor, which I found interesting.

Test Two- Shorten the range
Next, I wanted to see if I could shorten the range of the sensor. The lens on this sensor has three different sections. I first only exposed the bottom section, as seen below. This test resulted in me having to be within less than 2 feet of the sensor, probably more like 1.5ft.

Test Three- Shorten the Range 2
Finally, I wanted to see if I could make the range a little further. In this test, I exposed the bottom and middle sections. Range was about 6 feet before I was detected.

Another very obvious solution would be to use a Fibaro or Aeon sensor. These sensors are on mounts that allow them to be aimed in very specific directions. You could use it in the perpendicular wall example and then aim them so the outside of their FOV is at the edge of where you want detection to be. I think these seniors offer the most options. Plus, the Fibaro sensor is AWESOME - motion, vibration, temp, lux! Just read different threads up here.

Anyway, I hoped this help. Maybe there is also a short range sensor as well like you posted about. If there is anything else I can do or test, let me know.



Thank you! Very interesting that the range depth changed as you covered different sections. Did you try uncovering only the top 1/3 in case it’s the degree of coverage that matters rather than the placement?

I did not. My guess is that the different sections cover different distance…based on the testing…with the top section looking the furthest out.

Although I don’t know your exact setup, I think the Fibaro or Aeon sensors might be a good way to go since you can control their “aim” very well. I have a Fibaro and I love it. The mount makes it very easy to adjust the direction. There are also sensitivity settings. Although they are not exposed yet by @wackware’s device type that he coded, they will be soon via a smart app he is working on (I believe). He can better answer that. If you can’t find a close proximity sensor, this might be a viable option.

Thanks, Brian. I’ll look into those brands.

Meanwhile, another experiment if you have time. I’m curious if the motion sensor were mounted very low facing down towards the floor, would it detect a foot wave underneath? Or does it need more light? I can see some practical touchless switch applications for both a low wave and mounted under an end table. (Although I still expect a true proximity sensor would be better, if there’s none available it would be interesting to know what the options with the motion sensor are.)

Thanks again!

It won’t require light since they are all infrared. So, you should be good there. Using anything in the environment around it to mask the FOV would work.

I replied in the other thread but: Why not just mount Ubi to your chair. It uses a standard USB power supply so I’m sure you can get power from the chair for it.

It’s an interesting idea, but not practical, I think. The chair has a battery that powers the chair, but no additional power outlets. Battery life is already a huge issue for every power wheelchair user. You’re lucky to get two years out of them, often get only one. If the gel battery runs all the way out, it’s unlikely to recharge, and if the battery dies while you’re out you may have a hard time getting home again. Conserving battery power is a daily priority for most powerchair users. So it’s unlikely anyone would add any extra draw to it.

My iPad runs off its own battery during the day, and is plugged into a regular charger at night.

how about it’s own battery pack? USB battery packs are pretty cheap and small and you could just charge it when u charge your Ipad.

Even a small 12v battery could be easily setup to run it and recharge easily.

I’ll post a question on the Ubi board and find out current draw and see if it’s feasible to run it off a battery.

Thanks, Bernie, that would be very helpful to know! :smile: The UBI is definitely an interesting option.

So far not much info. It uses a 2 amp power supply so it draws a max 10w. Trying to get more specific info. Even so that’s not much and a USB Battery pack or small 12v battery will easily power it for a day.