[OSM-talk] What's good accuracy for a GPS unit?

[Aun Yngve Johnsen]

Augmentation of GPS can be done in various ways depending on the type  
of unit you have, wether you have access to paid subscription  
augmentation or only free systems, and where you are.

A smaller handheld unit generally have lesser accuracy compared with a  
larger unit.

Higher latitude (60 degree N/S) have generally higher accuracy than  
lower latitude (closer to equator)

Paid subscription augmentation such as Fugro Inmarsat, Fugro Spotbeam,  
HP/XP/G2, Veripos, etc. generally have a much higher degree of  
correction, therefor more accurate than free services such as IALA,  
SBAS (common term for WAAS, EGNOS, etc).

Terms of accuracy also depends on what you are going to use the unit  
for, I am working in the offshore industry, where large expensive  
units with several augmentation systems are used to install large  
structures on the seabed with centimeter accuracy, where we aim to  
have an HDOP and EPE as low as possible, I have observed all the way  
to EPE of 0.1, while a simple car navigation with "snap to road"  
function can be more than happy with HDOP of 2 (which would be totally  
rejected by my contractors).

The offshore industry is probably the most difficult consumer of  
position data, so comparing personal units to our standards is  
probably a bad idea, but if you have a unit that give you a HDOP of  
less than 2 than you probably are covered.

HDOP = Horizontal Dilution of Position (generally meaning that more  
than 95% of the measured positions are within a circle with HDOP meter  
in radius, though some units might use another % for the same)
You will probably not find any information about EPE on your units,  
that is a highly technical term involving more factors such as  
satellite constellation, number of corrections, wether or not both L1  
and L2 signal is corrected, how close to the horizon you allow the  
unit to read satellites, etc.

Sometimes accuracy is measured in %, this is only confusing, do not  
look at this number.

brgds
Aun Johnsen



On 13/04/2010, at 17:47, Phil! Gold wrote:

> Hopefully this isn't too off-topic for the list.  If it is, sorry.
>
> I use a Palm Pre for recording GPS traces.  Once it's warmed up, it
> normally reports an accuracy of 3 meters, which seems to fit  
> reasonably
> with what I've seen when overlaying my traces on the Yahoo WMS tiles
> (although sometimes traces where I was driving in the same place at
> different times can diverge by 4 to 5 meters).
>
> What I'm curious about is how well the Pre's accuracy compares to  
> other
> GPS units.  I've read that GPS units with WAAS (for I am in the US)  
> can
> get results with less than one meter of imprecision, but in a survey  
> of
> GPS unit accuracy online I only found reports of "3 meters or less".
>
> What's the typical degree of accuracy you get with your GPS unit?
>
> -- 
> ...computer contrarian of the first order... / http://aperiodic.net/phil/
> PGP: 026A27F2  print: D200 5BDB FC4B B24A 9248  9F7A 4322 2D22 026A  
> 27F2
> --- --
>  "Yeah, I'm gonna have to go with Dead Boy on this one."
>  "Could you *not* call me that?"
>                       -- Xander and Angel (Buffy the Vampire Slayer,  
> "Lie
>                          to Me")
> ---- --- --
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk