I am trying to make use of the trilateration using beacons and wifi. I know first I will have to take the signal strength of three beacons/wifi then the circles will be drawn from device to router or beacon. But how to train the data for this work. Because the signal strength keeps on changing and how to determine that how much distance to assume on which strength
Trilateration with beacons is extremely hard. Have you tried our Indoor Location technology yet? It also gives you (x,y) position, and you don't have to code any of it yourself. We have a team of full-time engineers working on this for more than a year now, so believe us when we say it's hard (: And I guess you likely don't want to invest several man-years into developing your own algorithm (:
But I am developing on android. The main problem How to cope with it now?
Ah, oops. That's a bit of a problem then (:
What kind of an app do you want to build? Maybe you could use the regular Estimote Android SDK to detect the closest beacon and do a rough-grained indoor location this way?
I am making an indoor navigation app for blinds
@makhshif we invested a lot of man-time into indoor geo-location at my company and with just beacons you're not going to have enough accuracy to actually use that information to drive someone around a closed space, even less blind people. You have to look into a different direction to actually develop something that can be used reliably.
@tanis, just beacons, agreed, which is why Indoor Location is more than just beacons (:
Check out this comment about 1 m accuracy using Indoor Location:
Maybe you'd be willing to re-evaluate Estimote Indoor Location? Let me know if we can help!
@heypiotr please surprise me!
I did try your Indoor Location SDK when we first got your dev beacons and I can't say good things about it as it was way too imprecise for actual usage. I suppose you changed a bit of the technology under the hood then. What are you using? Sensor fusion? Does it still need to map the rooms like it did in the first version?
We've been hard at work in fact, yes (: And we've got one more improvement waiting for release, in our internal tests it showed another non-trivial bump in average accuracy.
Yes, there's still a mapping process … but we've also completely revamped the mapping wizard, so that it's also more accurate. (And it suggests the optimal number of beacons for the given space now, instead of just assuming one per wall.)
Alternatively, we've also come up with a guide on how to map a location programmatically:http://developer.estimote.com/indoor/map-your-location/#the-advanced-way-location-builder
Yes. There is some sensor fusion used here.
Ok I have have had some success with this using Raspberry Pi's, if the Pi's are 10 meters apart from each other it will work because beacon have a signal boost at the 10 meter mark its hard to get an accurate signal (10 meters is say RSSI 78 and 5 meters is RSSI 78). Now I have done RSSI smoothing and I am able to get an accurate RSSI to send to a web server and then use the web server to preform calculations for trilateration using in accurate distances and then display a point on a map. Remeber that in the trilateration you will need a z index because of the height of the Pi's and the height of the beacons could be different. If you would like some help on this feel free to contact me.
Hi Tim, Have I understood your post correctly, you have used Raspberry PI's as scanners to locate a beacon within an area? If so what level of accuracy have you achieved. I've been thinking about using Pi's to do the same thing but the trilateration coding gave me a headache when I look at it Any help you can give me would be much appreciated.
Sure no problem I plan on posting all my code up here soon. I am just in the middle of a big project so it will not be till next week or so.
I will respond here when I put it up so you are alerted.
You are a star!!
Did you get any success on that. Can you share the code.
I am actually doing big university project and I plan to use this kind of indoor localisation as well (we want to help disabled people in transports). Do you still have your code? It would be very helpful if I can have a look at it
Thank you in advance!
@mgoutay, maybe this will prove helpful (:
Indeed!Thank you for that
Hey Guys, I am doing something similar to this I have a question,I had built my data base for finger printing with 4 beacons in each corner of a room. now I want to get the MAC + RSSI from just one access point which is installed at the center of the room. (this one is the university's access point and it will receive all the students MAC+RSSI ) I want t know if I can use my previous data base to estimate the location of the students? or this database is completely useless cause the access point location has been changed?