[clug] GPS, GIS, GTS, SDI and FOSS
scott.ferguson.clug at gmail.com
Wed Mar 2 18:15:09 MST 2011
On Thu, 3 Mar 2011 09:41:12 +1000 Andrew Janke wrote:
> To: linux at lists.samba.org
> Subject: Re: [clug] GPS, GIS, GTS, SDI and FOSS
> On Thu, Mar 3, 2011 at 09:34, Scott Ferguson
> <scott.ferguson.clug at gmail.com> wrote:
>> > I was probably (late night post) a little unclear - my requirements are
>> > for reproducible 1/2 meter to 1 metre absolute accuracy in the ACT with
>> > sub-2 minute warm starts at less than $1K per unit.
> Well my surveyor BIL who pays $40k+ for GPS units is firmly of the
> opinion that you can achieve this with 3 nokia phones and a bit of
> averaging. His fidding about with his phone (E72) has shown it to be
> surprisingly accurate compared to his Mega-$$ surveying gear
> especially if you are moving a bit. (ie: walking around the boundary
> of a property)
A-GPS is probably what the phones he's talking about use - I've done
some testing with a Telstra ZTE T165. Using some American data from John
Deere(?), which only partially corrects for local conditions, I can
improve resolution to around 3m absolute.
There are a number of techniques I've read about that use averaging +
GPS <--> GPS corrections to improve accuracy. For my purposes the use of
multiple GPS devices would be a last resort. It's necessary with A-GPS
devices as the corrective data supplied by the phone network is only
designed to resolve positions to police/emergency services areas - so on
their own they're not better than +/-2.5Km (absolute).
I have been looking at RTK:-
and (the home drone folks):-
I believe there are a number of viable solutions to the/my problem.
If you get the chance to ask your surveyor I very much appreciate some
pointers to information on the techniques he's using - though, as
previously stated, my ideal solution is a single GPS device, preferably
Not only are GPS units becoming cheaper and more common - the same is
happening with the satellites. Until very recently cost, size and weight
of the satellites was determined by the need for an atomic clock.
Apparently this is no longer the case.
More information about the linux