GPS based data acquisition?

[Richard Weait]

On Wed, May 26, 2010 at 12:06 AM,  <phiscock-g851W1bGYuGnS0EtXVNi6w at public.gmane.org> wrote:
> Folks -
>
> As part of an ongoing study of light pollution, I'd like to be able to
> automatically acquire data from a light meter and plot it on a map, while
> driving over a route.
>
> I have the light meter sorted out: it's a voltmeter equipped with a light
> sensor. The voltmeter can be queried from a PC host over a USB connection
> and provide an ASCII data string representing light level. The voltmeter
> appears to be a serial port periferal.
>
> I have a GPS receiver (the microsoft 'streets and trips hardware) that
> talks to its software via USB, and I suspect it looks like a serial port
> device. However, I can buy some other hardware if that simplifies the
> application.
>
> So what is needed is some sort of app that can be programmed to fire
> requests to a 'serial port device' and then simultaneously capture the
> data that comes back and the GPS coordinates.
>
> Anyone have suggestions where to start with this?

Hi Peter,

gpsd is the way to talk to a gps device.  It should handle any device
specific translation issues and allow you to save the location data
that interests you.  You could, for instance, save your location as
trackfile.gpx with one data point per second.  That's probably a good
place to start as you'll end up with an XML file of your progress.
The gpx format is widely understood by programs and is relatively easy
to read for a human.

You'll be doing something similar with the light meter.  Query it
regularly and save the data.  You might be tempted to reconcile light
data and position data by timestamp.  Be aware that the timestamps in
the gpx file come from the gps system, not your system clock.  You'll
want to either note the offset between those clocks, or synchronize
your watches before data collection.

I've always wanted to have an excuse to synchronize our watches.

And an example from the OpenStreetMap world.   Cameras with GPS, or
GPS attachments for cameras are still relatively rare and expensive.
We can use a cheap digital camera and a separate GPS, then reconcile
the directory of photos later.  One trick is to take a photo of the
time on the GPS receiver at some point of the survey.  The
reconciliation software then asks "what time is shown on the GPS" and
calculates the offsets for the survey and places photo thumbnails
along the GPX trackfile in the editor.  As long as the camera
timestamp / clock isn't too terrible, location should be "close
enough."
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists