Indoor navigation

Question

We would like to develop an indoor navigation with BLE. The goal is to get the position of the mobile device and navigate to a specific location. 
 We have already four nRFBeacon and one nRF51DK. Our idea is to use the nRF51DK as master and the four nRFBeacons as slave. 
 1.) How many Beacons are necessary to get the exact position of the mobile device(for one room)? 
 2.) How should the Beacons be placed in a normal room with four walls? 
 3.) If one Beacon is to far from the master, is it possible to send the information to another Beacon and from there to the master? 
 4.) Are there already project like the indoor navigation? What are the difficulties? How should we proceed?

Dan Danknick · Accepted Answer

I have worked on similar projects and this sort of thing is practically more difficult than many would have you think. Indoor position resolution is the Holy Grail of RF and has been for ten years. So you have some work ahead of you. 
 To answer your questions not in order:
Beacons are typically transmit only. Someone could write code to periodically pause beaconing, flip the chip around to be a Central, collect other beacon data and then re-beacon it with a proper suffix but I'm not aware of anyone that has built such a "beacon relay" device. I'm not sure how it would help though as you're essentially collapsing the beacon data to resolve the position of the relay. 
 Since you're considering the problem in the geometry of a room you must visualize the beacons as spheres where the edge is the encoded RSSI at some distance. Your location data is best when the receiver is in the intersection of N spheres where N > 3 when the assumptions are an X,Y solution. But within that intersection of spheres there is ambiguity. 
 The spherical modeling of above has caveats due to attenuation in many directions. If a person is carrying your receiver that's the worst as 2.4GHz doesn't pass too well through water (like humans). I guess a pet shop full of fish tanks would also be terrible. :) 
 My suggestion on how you should proceed is to construct a well instrumented receiver, buy a bunch of beacons and test test test. If I were putting together a feasibility study I'd buy 10 beacons and set aside six weeks for testing in 20 different rooms. 
 Roximity.com sells an open version of their Model X beacon which allows you to alter the transmit power and configure the broadcast RSSI. I'd want both of those alterable for a testing scenario (disclosure: I wrote the Roximity firmware). 
 Hope this helps,
Dan

Indoor navigation

Top Replies