[ros-users] Standard GPS Message

Ken Tossell ktossell at umd.edu
Sat May 22 00:00:27 UTC 2010

On Fri, 21 May 2010 19:40:18 -0400, Eric Perko <wisesage5001 at gmail.com>
> I don't see any notion of standard deviation of the fix in the gpsd
> message format. That is _really_ important - am I just missing the
> field in the gpsd message (I couldn't find any documentation on what
> each field in that message actually means or what units it has, any
> hints?).

That's epe in GPSStatus (err in gps_common/GPSInfo [below]).

Here's a cleaned-up version of the gpsd_client message, with the UTM
fields and some descriptions added:


Some things that I think need to be cleared up in that: What should we do
with read/reported values other than the basic horizontal/vertical/altitude
readings? GPSd reports ground speed, climb speed and direction (which I
believe are all calculated from a series of position readings), magnetic
XYZ, acceleration XYZ (from an accelerometer), gyro XYZ and temperature [as
well as whether each of those is available]. IMO a standard GPS message
should include GPS-derived speed, climb and track, but I'm not sure whether
the others should be included.

I'd like to have a shared GPS library that implements things like:
 * conversion from lat-long-alt to UTM
 * calculation of Xdop from the directional components
 * calculation of speed, climb and track from a series of lat-long-alt

 - Ken

More information about the ros-users mailing list