Indoor Positioning System

Hello, I'm trying to build a system which will track the location of a wearable tag inside a building. The system doesn't have to be exceptionally accurate, I just need to know if a tag is in a specific room. For example, say I put a tag on my cat, I would just like to know if the cat is in the kitchen or in my room, not specifically where in the room. I'm open to just about any kind of solution which isn't too bulky or expensive. So far I have tried a system using RSSI to approximate distance and triangulation but the RSSi to distance relationship varied quite largely. Any ideas? Thank you!

I don't know, and I didn't search nothing on the web, but one day I heard something about a triangulation with three bluetooth transceiver. But the method was the signal strength too

Simple signal strength is a non-starter with RF for various reasons, varying polarization, near field disturbances,
multipath reception and obstacles. Time of flight works, and there's a new system called DecaWave that
uses ultra-wideband time-of-flight that's resistant to multipath - not sure cheap versions are available yet.

You can use signal strength with a database of measured signal strengths per position for each beacon,
but that requires having a database and setting it up... Indoor positioning system - Wikipedia

Ultrasound beacons might be one method, but will probably annoy any feline as they have much greater
range of hearing than humans.

MarkT:
multipath reception and obstacles. Time of flight works, and there's a new system called DecaWave that
uses ultra-wideband time-of-flight that's resistant to multipath - not sure cheap versions are available yet.

You can use signal strength with a database of measured signal strengths per position for each beacon,
but that requires having a database and setting it up... Indoor positioning system - Wikipedia

I was looking at Decawave and this is the route I'm going to take. I found the receivers are $25 and with some light DIY, it seems I can set up a system for relatively cheap considering how much prebuilt solutions cost. Pozyx and others have open source Arduino libraries which will definitely accelerate the process. As for the database, that was my plan to store RSSI versus distance data and I have one running on my server already. RSSI just isn't reliable enough to track position but it appears that Decawave and UWB is the solution. Plus I think my cat would be much less annoyed this way too :stuck_out_tongue:

archeresque:
So far I have tried a system using RSSI to approximate distance and triangulation but the RSSi to distance relationship varied quite largely. Any ideas? Thank you!

Have you actually tried this, and if so what were the variations you saw, we are interested in real test results ?

Your other posts on this topic;

Did not suggest you had actually tested your original idea at all ?

How about a simple pulsed ir transmitter?

Seems my tv remote will operate in the same room no matter where I point it.

srnet:
Your other posts on this topic;

http://forum.arduino.cc/index.php?topic=498860.msg3405853#msg3405853

Did not suggest you had actually tested your original idea at all ?

Yeah I tried to "spoof" RSSI by counting the packets received within a timeframe but the issue was that with the nRF24 modules I was using either all the packets got through or none of them did. I ordered some 433 mHz radios to try but I haven't gotten a chance to run the code on those but it seems most users are advising me against it essentially saying it's a waste of time. I was able to only very occasionally not get the full amount of packets I was sending and the results didn't seem statistically significant nor were the results easily reproducible.

bluejets:
How about a simple pulsed ir transmitter?

Seems my tv remote will operate in the same room no matter where I point it.

I was thinking in the same lines.
Some wireless transmission that works only in the confines of the room, then can detect which room the tag is in.
Battery powered ATtiny flashing an IR LED (doesn't have to be continuous), ATtiny with TSOP as receiver to send the signal to the central control. Small, inexpensive, and should be quite easy to build.

I found the receivers are $25 and with some light DIY, it seems I can set up a system for relatively cheap

I'm really curious to know why it's so important to you to spend any money on knowing where your cat is.

Mmm... Missed the cat part. IR is out then, doesn't work from under the sofa or inside the kitchen cabinet.

kenwood120s:
I found the receivers are $25 and with some light DIY, it seems I can set up a system for relatively cheap

I'm really curious to know why it's so important to you to spend any money on knowing where your cat is.

Surfing the web I saw a project where somebody tracked their cat using bluetooth and mapped out the cats favorite spots in the house. It looked like a fun project so I decided I'd put my own spin on it by tracking him through the house and logging his position in a database. Once I got started, I learned about location tracking and how expensive it was to implement so I became determined that it could be done for much cheaper. I then fell into the rabbit hole of RF communication etc. and became determined to solve the problem without spending money in expensive modules. The cat was just a catalyst to this project not necessarily the end goal.

wvmarle:
Mmm... Missed the cat part. IR is out then, doesn't work from under the sofa or inside the kitchen cabinet.

Hmm I hadn't thought of using IR, I'll have to take a look at that. You're right that it won't track him when he's hiding but it's still something I could implement to achieve my bigger goal of creating a low cost IPS system.

archeresque:
catalyst

Har-de-har…

Low power, high frequency signals may work quite well. After all WiFi is notorious for not being able to penetrate walls well, low enough power WiFi may work. Or even higher frequencies (do check your locality on spectrum use limitations).

wvmarle:
Mmm… Missed the cat part. IR is out then, doesn’t work from under the sofa or inside the kitchen cabinet.

Is that really true?

If the cat’s IR transmitter can be detected between the time s/he goes through the door and gets under the sofa the system will have recorded which room s/he is in.

In the Original Post the OP said s/he only needed to know which room, not whereabouts in the room.

In some other very recent Thread the OP was looking for a solution that could play a different tune (IIRC) depending on which room it is in and I suggested IR because lower wireless frequencies (eg. 2.4GHz and below) can penetrate walls and doors when you don’t want them to (and not when you do :slight_smile: ).

What had not occurred to me was to have the IR transmitter on the device that moves around (the cat in this case) and I reckon that would avoid most problems of the IR signal spilling from one room to another through doorways.

…R

Cat under sofa would give the same signal as cat sneaked out of the window: none. That may be a problem.

wvmarle:
Cat under sofa would give the same signal as cat sneaked out of the window: none. That may be a problem.

A problem for every solution :slight_smile:

...R

Yup. Otherwise it'd be too easy to be a fun challenge.

The simple solution of course would be to block off the space under the sofa, so you know whether the cat is still in the room or went out of the window.

wvmarle:
The simple solution ....

... is to get a dog.

Indeed. Then you can be sure the cat went out the window :slight_smile: