Do location beacons have more precision than stickers or proximity beacons?

Indoor Location
1

Hey guys!
I am currently trying to deploy a demo using proximity beacons for indoor positioning, but I am not getting precise results.

If I change these beacons for location beacons, should I get more precise results? In fact I just need to be sure if a user is at an alley or another. Between 2 alleys there are about 5 meters of distance.

Thanks in advance

2

At the end of the day, all of Estimote products are built on the very same foundation: Bluetooth Low Energy, with all its pros and cons. So in the grand scheme of things, it’s fair to say you can expect more-or-less the same accuracy from all the beacons.

But then there’s also those small things that make some models better for certain use cases. For example, Location Beacons have the best battery life, so for our Indoor Location, we can crank up their advertising frequency to the maximum, and then have lots of data points for our algorithms, which makes the accuracy better. For Stickers on the other hand, we have to use very conservative advertising settings, so the computed proximity is expected to be less granular than that of Proximity/Location beacons.

With Location Beacons and our Indoor Location, you should definitely be able to achieve the “up to 5 meters” accuracy you require—our internal tests show accuracy between 1 and 4 meters.

3

I am using “Proximity Beacons” and I am getting pretty good results 1m to 4m on both iOS and Android.

I tested my results also with Trilateration algorithms, and I was able to “guess” the position of a person in a room by about 1 to 4 meters.

One tip (which I mentioned already in the forum): When you do ranging (according to the demo code), you might noticed that the list of closest beacons might change every 300ms. If you implement a “low pass filter”, by buffering the beacons_id + distance + time, and then you are using all values <1000ms, you get a more stable result.

Philipp

Schneeweis.Technology

4

Interesting! Our experiments with trilateration/multi-lateration were mostly complete failures. Fascinating to know you made it work for you, that’s awesome (:

5

I hope I didn’t promise too much. I felt that the distance reported was pretty accurate, and the trilateration algorithms in my test environment work pretty good as well. So if the distance is accurate, trilateration works pretty well. However, the real on site test so far has not happened. If you have a client who wants to try this out, let me know.
All the best,
Philipp

6

Ah, gotcha. Yeah, the problem with beacon-based trilateration is that … the distance estimations are indeed not that accurate (;

(This is not to say that you can’t have indoor positioning with beacons, only that you need something more sophisticated that trilateration. Our Indoor Location for iOS shows 1–4 m accuracy in our internal tests.)