8

I'm doing a project which needs to track location indoors. It needs to track where someone is inside a building and show them where to go. I've read quite a lot on the subject for all three main smartphone OS's (iOS, Android and Windows Mobile). I would prefer to do it on Windows Mobile if possible to start with as it is the system I can work with the best.

I have decided I'm primarily going to use QR codes to "reset" the position (i.e., have reset points throughout the building where people can stand, scan, and it gets their current location). How can I then track movement away from this point of origin?

Blumer
  • 5,005
  • 2
  • 33
  • 47
Daniel Case
  • 81
  • 1
  • 2
  • Since you've read a lot already, I'm guessing you've seen http://www.indooratlas.com/ ? You might get better responses if you give some more specifics on what your requirements are and what you do/do not like about what you have seen on the market. – Scott W Oct 08 '12 at 19:39
  • since GPS wont work you could try using the accelerometer to track movement... this would still be fairly inaccurate however... – Austin Mueller Oct 08 '12 at 19:43
  • 3
    This question is overly broad as written: pick a single platform and/or be clearer about your constraints: how much resolution do you need? how much control over the environment do you have? – benzado Oct 08 '12 at 19:43
  • Hi, sorry about being vague. I just want to keep my options open as some things are available on some platforms and I've been told it only needs to be a prototype. I've seen Indoor Atlas but it's a university project so I can't really use anything proprietary, what I'm after is ideas that I can follow up on and research, ideally free and open source. – Daniel Case Oct 08 '12 at 20:23

4 Answers4

1

As far as I know, the only solution for you is to use indoor location fingerprinting. Essentially, you need to map signal strengths on locations for the positions that are known. Whenever a mobile phone is on that location, it should read similar signal strengths from various WiFi antennas. When this reading is compared to fingerprints of the indoor locations, the best match should be the location. Different algorithms can be applied. For further reading I suggest papers like link1 and link2

Nikola Davidovic
  • 8,556
  • 1
  • 27
  • 33
1

One appraoch is to tag indoor positions with barcodes defining a location. But this is only as acurate as users scan the barcodes during crossing and scanning the barcodes.

Another approach uses WiFi signal strength. AFAIK CISCO networks are able to give good results with there indoor location services: http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/wifich1.html

josef
  • 5,951
  • 1
  • 13
  • 24
0

Accurate GPS needs direct line of sight to the satellites. Therefore you cannot track location room by room with current mobile phone technology.

Black Frog
  • 11,595
  • 1
  • 35
  • 66
  • 1
    While it is true that GPS does require line-of-sight, there are non-GPS ways of finding one's location (e.g. indooratlas.com or even simply knowing what WiFi networks are nearby) – Scott W Oct 08 '12 at 19:41
  • Sounds like he wants room by room tracking. And if you have the same Wi-Fi network ID in the whole building, how would that work? – Black Frog Oct 08 '12 at 19:46
  • 1
    @BlackFrog Access points have individual, unique IDs independent of their SSID. Most Wi-fi sniffers will show you these. – ceejayoz Oct 08 '12 at 19:57
  • A dead-reckoning device would work just fine, so to broadly say "you can't do it with today's technology" isn't true. – ctacke Oct 09 '12 at 14:41
0

All of our WiFi hotspots at work are assigned the same network ID. I don't know if this is necessary or something our lazy network administrator has gone with.

However, I do know you can pull site survey style data on your WiFi signals. If your network IDs were all different, you could get some crude triangulation if ID_A and ID_B each had 3 bars of strength while ID_C had 4 bars.

I wanted to do this with our network, but gave it up when I learned that all of our hotspots had the same network ID.