Indoor positioning and context awareness

3 views
Skip to first unread message

Jorge Silva

unread,
Mar 14, 2010, 12:06:47 PM3/14/10
to Geomena
Hola All,

I have been working on WiFi-based indoor location at the University of
Toronto for a while and would like to know what you guys think would
be the best way to contribute what we have done so far. Specifically,
we've been working on the following:

1. tagin! (http://scyp.atrc.utoronto.ca/projects/tagin) a text-based
location tagging service (think crowd-sourced digital signage) and,
2. WIPS (http://scyp.atrc.utoronto.ca/projects/wips), a graphical
version using floorplan images

For demos, you can check:
http://scyp.atrc.utoronto.ca/projects/tagin/demos and,
http://www.youtube.com/watch?v=nTd-mCzlwYM

If you guys think this may be useful to Geomena, I'd like to chat with
those of you who would be interested in guiding me as to how to
proceed. I am a very bad programmer with little experience on gnome/
OSS (learn quickly though) so I will likely need a bit of hand-
holding. It is also likely the code we have will need a re-write
(specially WIPS), since I have never really had any guidance from
experienced programmers.

I personally think this is really cool technology, but since our
funding is getting constrained I am considering joining efforts such
as geomena instead of trying to maintain a separate group ourselves.
I'll be happy to hear your thoughts, concerns and/or suggestions.

cheers!
Jorge

Don Park

unread,
Mar 14, 2010, 2:35:29 PM3/14/10
to geo...@googlegroups.com
Hi Jorge,
Thanks for the email. There are a number of ways to contribute. The
leading priorities right now are a bulk upload feature, and desktop
clients to enter new location data for access points that are not yet
in geomena. Even without programming, there is plenty that can be
thought up, sketched out, and discussed as to how these or other
features should work. If you are certain that your existing
wifi-lat/long data is available to put under the creative commons
license, a simple script in ruby/python can insert that data in bulk.

As far as taking a message and attaching a lat/long to it, I think
thats a great idea. I've seen other android apps that claim to do
that. A standard XML format would help with mass adoption. A
geo-tagged tweet for instance is a message and a location tied
together. A tweet is not long lasting and there are not clients that
display tweets for a given area. So a general database of messages and
lat/longs could be valuable if it has a dead-simple API.

Don

Jorge Silva

unread,
Mar 14, 2010, 3:48:50 PM3/14/10
to geo...@googlegroups.com
Hola Don, thanks for the quick reply... I'll venture some clarifications
because I think I was misunderstood...

> If you are certain that your existing
> wifi-lat/long data is available to put under the creative commons
> license, a simple script in ruby/python can insert that data in bulk.

Yes, we could do the bulk upload, but we don't use lat/long info because
it is insufficient for indoor positioning (too inaccurate and can't use
it to differentiate amongst different floors in a building). We are
doing room-level positioning indoors, not just GPS coordinates that can
be retrieved from inside... I mean the real deal: giving you a specific
coordinate of the floorplan inside the building where you are. AFAIK,
nobody else is doing open indoor positioning, so nobody really has the
tables required to hold our data (which include room number, floor id,
floorplan url, x/y coords, etc...)

> As far as taking a message and attaching a lat/long to it, I think
> thats a great idea.

Yeah, I think that's cool too, but that's not what we've been doing. I
invite you to check out the Android demo called tagin! (available for
free on the market). The UI is horrible, but should still give you a
better idea of what we are trying to do. There is also a firefox
extension with the same name and that UI is much better. For some reason
it has been really hard to explain what we do... my guess is that people
have learned this kind of accuracy or functionality is impossible to
achieve and just ignore it. You can think of it as an add-on or
extension to a geolocation service that may be able to return more
accurate info than the usual lat/long (e.g.: floor, room numbers).

> A standard XML format would help with mass adoption.

We use XML for posting/retrieving location info, see some samples here:
http://redmine.atrc.utoronto.ca/wiki/wips

Anyway, I hope this clarifies a bit better what I was really after.

cheers!

Reply all
Reply to author
Forward
0 new messages