GPS localization issue

Just raise or lower the rod a tenth.....to correct it!

🙂

N

Some localizations work and some bite you in the butt. I haven't done one in years but I've dealt with surveys that did. All localizing does is best fit a project for scale and rotation. Find a way to enter that into your configuration and drop the localizing and you should breath easier. As. For network heights we use geoid 12a and survey the project benchmarks then compare to published and make a best fit vertical transformation to the network base and recompute . After the recomp the checks are almost always 1 cm or less provided the heights are worth a dime to start with.

Haha. It wasn't that big of a deal except that conditions on that site were ideal, and I had planned on staking the curbing with my rover. A tenth isn't anything for me to worry about when I'm roughing a site in, but ADA standards are making it tough on everyone as far as curbing goes. You just can't be out a tenth when you've got some guy checking your work with a 4' level.

You will never find us staking pavement with a network rover. Wrong tool

You shouldn't have a geoid file loaded if you align to local control points.

I've had nothing but success with it. I check in before, during and after staking. It's the wrong tool in the wrong operators hands.

Like I said, I found a tenth, put it in the truck, and got my total station out.

I'm running surv ce and I think it replaces the geoid 12 a file I'm running, and creates a sort of inclined plane geoid model within the .loc file.

Thanks for the advice. I'll try my own wholesale vertical translation next time. That takes any guesswork out of it.

>The last project we did, had great vertical ties on one "half" of the site, and a consistent tenth on the other. I used five looped control points for my localization, and all control lines extended beyond the scope of the project. I'm curious as to what went wrong. I've been told that using one vertical tie for a localization can sometimes produce better results. Also, I'm under the impression that when I localize vertically, the computer creates a new geoid file and uses it instead of the normal one I have loaded. Is this correct? And if so do I need to choose the control points to which I localize with that in mind as well?

Unless you are doing a corridor, I would not advise doing more than one. It essentially creates an inclined plane between all the points and measures off that. When localized to more than one it will slope the plane often leading to unexpected results. It has been my policy to only use one as vertical and as many as I can for horizontal. I did two separate 2 mile level loops through control set off one vertical never missed one by more than about .05' but when I tied them together in the localization I missed the farthest out by about .15'

Also if you do want to continue using more than one. Take care the order you put them in, the localization weighs the order placing more precedence on those first in the localization. So I would do one on one side then the next one on the other etc, etc...

Thanks. We don't do many route surveys, so holding one benchmark near the elevation center of the site will work for us. I thought I had an idea of what was going on, but it is always nice to hear a little confirmation.

I would never used elevs for construction projects with RTK. They are good for positional tolerance but not vertical! Believe me... a private company I worked for got screwed for that!

Ran into the same problem on a site when I was pretty green with networks, using survce. Half the site seemed better than the other side. Trust me.. It's the network; you can check in just fine but at random times you can have that tenth vert pop up. It's spurious. If you don't believe me try laying them all out and do a level loop through them. It's not the localization, it's the method and the network. Put it away and just use the network rover to lay things out for horizontal. You'll be able to rest at night.

The way I look at localization is that it is letting the software make the transformation choices. Why would a surveyor knowingly put an inclined plane or a variable inclined plane on a project? If water flows according to gravity why would you replace a fairly good gravity model with a plane tillted away from gravity by various amounts. The 10 yr NGS plan calls for height to be controlled by a gnss gravity model. We have been using the unmodified NGS gravity model with good results for some time. You may have to adjust your base height up or down to fit local height but it doesn't change the modeled geoid height anywhere on the project.

Back in the early 90's the RTK systems started becoming available and calibration or localization was the big thing. I realized fairly quickly that it wasn't a very good procedure. Horizontals were bad enough, but the verticals were even more frustrating.

There is a way to calibrate and use a geoid model in combination. The problem with doing that is I never got it to work very well. I've found it's better to use site control and adjust my ellipsoid height to shift the geoid model I'm using to match the given elevations. Then check the elevation control points.

Even so, RTK just isn't a fine enough instrument to lay-out hard surfaces or buildings. It's great for slope stakes, rough grade, dirt topo, stock reservoirs, drill rig pads and roads, ect. When it comes down to curb, final blue tops, building lay-out it's time for the total station, instruments, levels, tapes, string lines. You don't want to use a pipe wrench to assemble funiture and no GPS for concrete.

If you can stake 100 curb stakes, then come back and run checks on them with a level and 100, not 95 of them, check .02' or less then we can talk. Until then I'm not even looking at doing it-been down that road too often.

And I don't think the network guys would say you could do that either.

Tried it years ago for an experiment and checked with level loop. Got about a tenth under good conditions.

That's not good enough for me for sanitary sewer or curb (asphalt laid in off that). I'll stake water with GPS. Red tops are set by Machine Control now, not us.

I run in the best vertical I can, especially if there's a benchmark around. Then I do a horizontal control net with GPS for total station. I level through them with a closed loop; helps you sleep at night.

Then I set those with a total station with the standard checks. Got miles of the stuff under my belt, no mistakes. No lip from the curb crew either. You can even shoot the string with a peanut prism.

I'm not sure about the curb guys where you are from, but the day laborers here get out a concrete coated 4' level and come off of my stakes to set their forms. A quarter of an inch isn't in their vocabulary. Column control, building pads, and split walls are a completely different story, but unless you are working with the most accurate curb crew in history, 0.02' is not something I'm worried about.

FYI keep your liability insurance up. DOTs are spending millions a yr on design alone for fear of lawsuits for slope violations. Your tax dollars at work thanks to ADA.

http://www.access-board.gov/research/tolerances/final-report.htm

Initiative on Dimensional Tolerances in Construction
Dimensional Tolerances for Surface Accessibility

Final Report

1 Accessibility for walks, ramps, and stairs
1.1 Measurement protocols

1.1.1 Measurement tools. Distance measuring devices should be capable of measuring to a precision of 1/16 in. (1 mm). Angular or slope measuring devices should be capable of measuring to a precision of 0.1 degree and elevation measuring devices should be capable of measuring to a precision of 0.01 ft (1/8 in. or 3 mm).
For measuring device accuracy, the instrument and technique for measurement should provide for an accuracy of at least one - third the required tolerance. For tolerance measurement requirements of 1/8 in (3 mm) this would mean a device capable of measuring to about 3/64 in. Because this is not realistic in construction, a tape measure with 1/16 in. divisions is a reasonable compromise for inch - pound units. For metric measure, most tapes are marked in 1 mm units, which make them ideal for determining 3 mm tolerances.
For exterior ramps, sidewalks, and intersections, in any direction, the gap below a 10 - foot unleveled straightedge resting on highspots shall not exceed +1/4 inch (6 mm).
The tolerance for the top surface of a slab - on - ground is ± 3/4 inch (19 mm).
The deviation from slope or plane of formed surfaces over 10 feet (3 m) is ± 0.3% of length. This is approximately 3/8 inch in 10 feet (10 mm in 3 m). However, the ACI standards do not permit interpolation or extrapolation for dimension greater than or less than 10 feet.
The tolerance for deviation from slope or plane for a stair tread from the back to the nosing is ± ¼ in (6 mm).
The difference in the height of adjacent risers of stairs measured at the nosing shall not exceed 3/16 inch (5 mm).

Smart, smart man.