Contractor got a state road job in early 2010. 2 lane road a little over a mile long, straightening out a bunch of bad curves and short sight distances and bad grade, essentially made one long sweeping curve cross country, with all the old road snaking to the south of the new by up to 1000 ft, tied back in at each end.
Designers provided monuments near each end along the old road, with SPC and project coordinates and a scale factor to convert between. (scale factor makes about 11 ft. of difference in the coordinates). Road design was on project coordinates, so all the staking had to be on project coordinates for the centerline stationing and alignment and offsets to work. R/W break points were provided in both SPC and project coordinates on the plans.
I initially set up my GPS on one of the monuments near the beginning of the job, checked a couple of the monuments at the other end within a few hundredths, and staked the R/W using the SPC provided. I also set some control points along the R/W where possible so I could flag clearing lines using my total station. I calced the project coordinates for the control I set, and we built the road off that, never using the SPC file except for setting additional control this summer and restaking R/W when needed. I always used the same monument for my base.
In order to tie in to the existing roadway, detours were built at each end, and the original control monuments were all destroyed. Now that the job is finished, I went back to stake the R/W markers yesterday. I set my base up on one of the control points I had set near the middle of the job. Using a Trimble R8 base and rover, and Survey Controller software. I usually look at the map and pick my setup point from there, which is normally very near the symbol that shows the autonomous position. I noticed that the 2 positions were some distance apart, but that occasionally happens when it's only been there a minute or so, so I just selected the correct point and started the base. I then got the rover going, it initialized, and I headed off. After I set a couple of points, I lost initialization. Waited for a few minutes, nothing happening, checked the base, still running, restarted the base, still no initialization on the rover. Moved to another point and the map showed same offset from where I should have been, so I let it set for a few minutes, recorded the base coordinates, and they were off my point about 120'.
Called my Trimble guy to see if he had any ideas. He came up with a few things to try, everything checked but nothing worked. Looked like only answer was that the plan coordinates were NOT SPC. So I set about staking the markers using the robot. Still using the SPC file, since everythin was hitting the old stakes and nails I was finding.
He came out this morning with his VRS setup, and confirmed that nothing out there was actually on SPC, it was all about 120' off. Even found an original control pin that was set on the old road, and missed the plan SPC coordinates by 120'. Everything is in the correct place, just the coordinates are all off.
So, all that being said, here's my only question.
How was I able to set all the control using the bogus SPC, at different times over the past 2 years?
I had never, not once, seen any noticable difference between the SPC provided and the autonomous position, and had never had any problem with the rover initializing. We are both scratching our heads. It is NOT a scale factor or project coordinate problem.
How far off SPC do you have to be before your system alerts you? Did you confirm SPC with OPUS? Sounds like a translation error.
Isn't is pretty standard practice to always be recording the raw data observations for a base station for work like this? Why don't you submit some of those files to OPUS and see what coordinates you get back? Also, what is the coordinate shift if you convert between us survey feet and international feet for the SPC coordinates you have been given?
I'm sure this one will get well covered by the gurus, but it seems that you just calibrated to that one monument and always used it. It always works for me. So everything should, and did, fit because that is what you were using. Instead of thinking SPC coords, you could just as easily been on 10,000/10/000.
Then your 120 ft shows up after your original monument got lost, and you tried to introduce SPC. I'd be thinking the design guys bollixed it up, and they have a translation hiccup. If you'd have tried multiple setups during original control work, and run those with OPUS (a must do for any project where 'they' provide you SPC's), you likely would have discovered it then. But that likely wasn't in your scope.
Yeah really.
Most GPS gear gives a notice if the computed Lat/Lon differs from the GPS autonomous Lat/Lon by more than 500ft (or is it 500m). Anyway you are within that. But to answer your question. If the GPS is not on a VRS or RTN it will use the autonomous Lat/Lon of the base and assign it to your state plane coordinate projection, then everything will be relative to that point. So long as your bearings are on grid, you'd never a problem. If you had run an OPUS or used a VRS you would have found the error, but still could not use the actual SPC to Lat/Lon position. Most likely you need to stay on the map coordinates, else translate the entire project and take risk of grabbing wrong coordinate later in project from the map or CAD. I have never see 120ft error. Even an autonomous position is within 10-20ft most of the time, never seen 120ft. Sounds like a typo. Hard to believe, no one ever tried to drop this project onto an aerial photo. That would found the error.
Around here I had two projects this year missing SPC by 2ft horizontal, and 0.8ft vertical. I run OPUS on ever base setup and use it to note the difference between map coordinates and actual state plane, but always stick with map. We are always getting new design changes from engineer in original map coordinates. But if I loose every control point on the project I can use OPUS or VRS to get tied back into the projects map coordinates. Also I can tell if my control pole or base station point has moved from the daily static sessions run through OPUS.
LeeGreen.com
> I initially set up my GPS on one of the monuments near the beginning of the job, checked a couple of the monuments at the other end within a few hundredths, and staked the R/W using the SPC provided.
Did you calibrate to those project coordinates? Did you do a "Here" position and just check into the other points? That's the way is sounds by the sentence above.
If you calibrated to project coordinates, why are you using SPC coords? Maybe I'm missing something in your translation.
Wow. Just wow.
The Trimble "HERE" button at work. Suggest you re-process your raw data using the "OVER THERE" function?
I'm still trying to figure out what planet you could be on to have a scale factor to deal with 11 (eleven,really?) feet in a little over a mile.
Wow.:-|
Didn't do a here. State Contract is to build the job from the provided control, which is identified on the plans as SPC. Did not do a site calibration at any time, I had what were supposed to be SPC, and the monuments checked with each other. This site is in the middle of nowhere, had a hard time even getting the VRS to work.
What has me confused is that I could see the difference on the map when I set up this time, but had never had that issue before. And the fact I could not get the rover to initialize. I'm assuming that's because of the big difference in the coordinates, but had never had that problem on this site either. Could that be a separate problem, not related to the coordinate difference?
Scale factor is only to convert SPC to Project Coordinates. Statement on plans says:
Coordinates for horizontal control were obtained from GPS methods and adjusted to the NAD83/1994 SPC System. A project datum factor of 1.0000057236 was computed from the central most point to convert SPC to Project Coordinates. To obtain SPC, divide Project Corrdinates by the project datum factor.
The difference is actually about 21', the 11 in the original post was a typo. An example from the control point sheet of the plans:
CP#603
Project Coordinates 1894080.538N,691994.118E
SPC 1894059.6972N,691990.1573E
Didn't calibrate to the site, no need. Had SPC, and they checked with each other. Had project coordinates to stake averything else with the robot. Wooded, hilly site, couldn't use GPS for much until halfway thru the job anyway. Only using SPC to stake R/W with GPS, faster than using the robot. Had to make about 10 setups with the robot to see all the R/W markers, could have staked 90% of them using GPS.
> Scale factor is only to convert SPC to Project Coordinates. Statement on plans says:
>
> Coordinates for horizontal control were obtained from GPS methods and adjusted to the NAD83/1994 SPC System. A project datum factor of 1.0000057236 was computed from the central most point to convert SPC to Project Coordinates. To obtain SPC, divide Project Corrdinates by the project datum factor.
>
> The difference is actually about 21', the 11 in the original post was a typo. An example from the control point sheet of the plans:
>
> CP#603
> Project Coordinates 1894080.538N,691994.118E
> SPC 1894059.6972N,691990.1573E
maybe i'm seeing this wrong, but the above 'project datum factor' (presuming that means scale factor) at one mile is 0.03 feet. maybe it is a combined factor, but on a one mile project, this difference is too many magnitudes beyond scaling. still pondering other possibilities
I'm seeing a possibility here...
The factor on the plans is 1.0000057236. The sets of coordinates check using that factor.
Suppose there is 1 extra zero in there. Assume the original GPS work was correct, and as their statement says, they used the central most point to compute it. Then suppose they converted everything to Project Coordinates (PC) and did the design work. Then they backcalculated from PC to get SPC on each point, but used the extra zero. I'm thinking that could be what happened.
Dave,
I'm guessing he meant that the coordinate difference was 21 feet which means the original surveyor made the cardinal sin of not truncating after applying a project combined factor.
What elevation where you at at the time you were using GPS?
> Wow. Just wow.
>
> The Trimble "HERE" button at work. Suggest you re-process your raw data using the "OVER THERE" function?
>
> I'm still trying to figure out what planet you could be on to have a scale factor to deal with 11 (eleven,really?) feet in a little over a mile.
>
> Wow.:-|
I'm pretty sure you didn't direct this statement at me, but I don't calibrate and the only time I use the HERE button is to start RTK & Infill work.;-)
See my 05:53 post, I think that's where the screwup was. They had to have converted from SPC to project coordinates for design, then back to SPC in order to generate the R/W coordinates in SPC. I suspect they also converted the control with the same push of a button, but had an extra 0 in the project datum factor.
Still not sure how I was able to use it before, but not now.
> Didn't calibrate to the site, no need. Had SPC, and they checked with each other.
You had SPC which is fine AND...you had project coordinates, that's fine as well. Each unto their own. When I see or here the words State Plane Coordinates, I think "GRID" coordinates, not "GROUND" (or project) coordinates.
What was your method to get the GPS working with the coords you were provided. Did you create a LDP(low distortion projection), a single point ground projection(not a calibration or HERE) or did you just assign TGO the SPC zone you were working in?
>Had project coordinates to stake averything else with the robot. Wooded, hilly site, couldn't use GPS for much until halfway thru the job anyway. Only using SPC to stake R/W with GPS, faster than using the robot. Had to make about 10 setups with the robot to see all the R/W markers, could have staked 90% of them using GPS.
You used project coords for the robot and SPC coords for the GPS on the same project? Am I understanding this paragraph right? If so, this may be your problem. You mixed ground and grid coordinates or better yet, you mixed GPS observations with terrestrial observations in 2 different systems. (Yes.No?)
GPS doesn't really understand Cartesian coords, it understands Geodetic coords and converts to a Cartesian system strictly based on a set of parameters that you assign to it. I'm speculating because there is a lot of info that would take you pages to explain so I'm trying to gather the pieces right. I'm still curious as to your first procedure when you started the work...?...
ALSO, from a post earlier you made....
>A project datum factor of 1.0000057236 was computed from the central most point to convert SPC to Project Coordinates
Never, never, never from a design standpoint, scale from SPC grid to Project ground coordinates using a central point. Always create the projection and scale from 0,0,0. If your above sentence is true, you now have a point of origin that is not 0,0,0 and EVERYTHING in CAD and design is base upon 0,0,0. Less metadata to deal with and less of a problem for some designer to muck up the coordinate base.
Joe poses an interesting question here. Did you set up on a point that had a ZERO elevation?
I don't use a VRS, but with a traditional base an external radio, if the base was set up on a point that had a ZERO or blank elevation, the base would start but the radio would not transmit. The base point needs an elevation close to the real ellipsoid height to broadcast any corrections.
EDIT: AT least this is the situation with Trimble gear using a projection.
If the SPC coordinates are simply shifted and not rotated and scaled then I could this scenario:
You type in the SPC coordinate for the monument you used for the base. You start the base. Since everything is rotated and scaled correctly all the other coordinates that are shifted the same amount should check out fine (with some systematic error caused by the shift which is probably very small).
Later on you get an SPC for the base monument which is different. Several things could account for this. One is the Plan SPC coordinate was shifted for some reason (such as the original surveyor before you based the whole project on a here position somewhere in the control network). Epoch and datum shifts could account for the shift but probably not 150'.
> > Didn't calibrate to the site, no need. Had SPC, and they checked with each other.
>
> You had SPC which is fine AND...you had project coordinates, that's fine as well. Each unto their own. When I see or here the words State Plane Coordinates, I think "GRID" coordinates, not "GROUND" (or project) coordinates.
>
> What was your method to get the GPS working with the coords you were provided. Did you create a LDP(low distortion projection), a single point ground projection(not a calibration or HERE) or did you just assign TGO the SPC zone you were working in?
>
> >Had project coordinates to stake averything else with the robot. Wooded, hilly site, couldn't use GPS for much until halfway thru the job anyway. Only using SPC to stake R/W with GPS, faster than using the robot. Had to make about 10 setups with the robot to see all the R/W markers, could have staked 90% of them using GPS.
>
> You used project coords for the robot and SPC coords for the GPS on the same project? Am I understanding this paragraph right? If so, this may be your problem. You mixed ground and grid coordinates or better yet, you mixed GPS observations with terrestrial observations in 2 different systems. (Yes.No?)
>
> GPS doesn't really understand Cartesian coords, it understands Geodetic coords and converts to a Cartesian system strictly based on a set of parameters that you assign to it. I'm speculating because there is a lot of info that would take you pages to explain so I'm trying to gather the pieces right. I'm still curious as to your first procedure when you started the work...?...
>
> ALSO, from a post earlier you made....
>
> >A project datum factor of 1.0000057236 was computed from the central most point to convert SPC to Project Coordinates
>
> Never, never, never from a design standpoint, scale from SPC grid to Project ground coordinates using a central point. Always create the projection and scale from 0,0,0. If your above sentence is true, you now have a point of origin that is not 0,0,0 and EVERYTHING in CAD and design is base upon 0,0,0. Less metadata to deal with and less of a problem for some designer to muck up the coordinate base.
Rob,
Grid to ground is not the problem. I created 2 files. One was SPC, Ky South Zone, NAD83. the other one was scale factor of 1, using the project coordinates. As I said in the original post, the only control near the project was some monuments at each end. First thing I had to do was flag the clearing limits. Best thing I had to use was GPS to establish some intermediate control, then convert those to project coordinates so I could use the robot in the thick stuff. Also, as I said, the job was designed on project coordinates, not SPC, so all the alignment info and staking info for the road had to be project coordinates.
As for the project datum factor, that's from the Ky Design Manual. You'll have to argue that with them. I stake from whatever they provide. I didn't design it, or survey it for the design.
