Head to Head tests

Loyal, post: 370648, member: 228 wrote: The only way we will ever see a REAL head to head test of GPS equipment, is for some Surveyor to hit PowerBall (or megamillions), buy a bunch of equipment, and travel around the country "testing" ALL of the various "colors" side by side, head to head, simultaneously under real world conditions (which of course vary a HUGE amount).

Simultaneous testing of several different brands of GPS equipment would be a challenge in itself since the environment in which it would be tested can vary both spatially and over time. I'd think you'd pretty much be stuck with resurveying the same points on different days with the identical SVs in view.

If it is basically just RTK GPS to be tested, that is probably a much cleaner problem. A conventionally-surveyed test array of points with relative uncertainties much smaller than those expected from the manufacturer's claims as to RTK performance strikes me as a pretty good indicator.

Cool idea.
I've thought about it for a while. It's hard to 100% compare them.
Because its hard to place 2 receivers in the exact same place, at the exact same time.
If we had a gauntlet. Through some rough stuff. With a point out in the open, at both ends. And about 30 or 40 points down through the trees. Set up so that several are near impossible. Set up a robot, or total sta. So that all 30 points are visable from one station. Anybody can run the rest course with total sta, or robot. (After all GPS is done, and turned in)
Let all GPS units come from owners, or users of the equipment.
Somewhere in central USA. Complete with pics. And stories.
Since we are the ones to gain, or benefit from a side by side comparison, we would also like it if some, or alot of the test points were near each other.
Then, we all publish the "real" coords, and the differences with GPS.
I'm in.
I think it would be a 3 day event.
One day to organize the site, and event.
One day to do it.
Maybe a third day to compare And analyze.
My feeling is that we will wind up with winners, in various departments.
Fastest
Most accurate (relative to total station or robot coords)
Easiest to use
Maybe we could all turn in our data, to each other.
Then tie it all with robot/total sta.
The surveyors would be the winners.
The manufacturers would have a better understanding of modern performance.
RMS, values, and then the actual difference.
So, it would be considered a good answer, if it SAID it was accurate to .5' And it was off by .5 foot.
If it SAID it was accurate to .5', but was off 1 foot, then that would be an item of high intrest/concern.
I'm in.
I can bring my old legacy E's.

Nate.

Nate The Surveyor, post: 370661, member: 291 wrote: Cool idea.
I've thought about it for a while. It's hard to 100% compare them.
Because its hard to place 2 receivers in the exact same place, at the exact same time.
If we had a gauntlet. Through some rough stuff. With a point out in the open, at both ends. And about 30 or 40 points down through the trees. Set up so that several are near impossible. Set up a robot, or total sta. So that all 30 points are visable from one station. Anybody can run the rest course with total sta, or robot. (After all GPS is done, and turned in)

The real test, though, would be how well an RTK system performed over typical land survey vector lengths. Network RTK would have to be separately tested since it isn't generally applicable and is dependent upon the specific network within which it is operating. So, don't start with network RTK.

The most useful questions of system performance start with what the positional uncertainties of points surveyed at typical vector lengths actually are and how the actual uncertainties relate to those churned out by the RTK controller software. You could test those with an array of test points quite close together as long as the multipath environments were significantly different (and the separation from the RTK base was whatever was typical for the expected application).

Determining the actual control values of relative differences between point surveyed and base should be done by a combination of static GPS and conventional. It should be possible to reduce the uncertainties in those relative differences to maybe 1.5mm without breaking a sweat. Once you had determined the actual differences in the positions of the points occupied by the RTK base and the rover, then you'd have a much more realistic test than simply measuring repeatibility at unknown accuracy.

Loyal years back (33 years) the FGCC which is now FGCS (Federal Geodetic Control Subcommittee) started testing GPS equipment.

In Jan 1983 they tested the Macrometer V1000 Interfermetric Surveyor GPS (everyone remembers that machine). They have continued
up to July of 2001; Trimble 5700 (GPS L1/L2). In June 1986 they tested the Istac 2002 GPS L1 machine (Am I the only one that remembers that one??).

It is too bad that they don't test all machines on the market. I am sure that the manufactures would cooperate as in the past.

For a list of equipment that has been tested see www.ngs.noaa.gov/FGCS/instruments/reports/

JOHN NOLTON

Kent, i'd like to nominate you to be involved in this test.
And to write a report afterward.
How about it?

JOHN NOLTON, post: 370670, member: 225 wrote: Loyal years back (33 years) the FGCC which is now FGCS (Federal Geodetic Control Subcommittee) started testing GPS equipment.

In Jan 1983 they tested the Macrometer V1000 Interfermetric Surveyor GPS (everyone remembers that machine). They have continued
up to July of 2001; Trimble 5700 (GPS L1/L2). In June 1986 they tested the Istac 2002 GPS L1 machine (Am I the only one that remembers that one??).

It is too bad that they don't test all machines on the market. I am sure that the manufactures would cooperate as in the past.

For a list of equipment that has been tested see www.ngs.noaa.gov/FGCS/instruments/reports/

JOHN NOLTON

The FGCS Reports are available here: http://www.ngs.noaa.gov/FGCS/instruments/reports/