Activity Feed › Discussion Forums › Strictly Surveying › To GPS or not to GPS, that is the question
-
To GPS or not to GPS, that is the question
MightyMoe replied 7 years, 8 months ago 25 Members · 63 Replies
-
Bruce Small, post: 406969, member: 1201 wrote: I have repeatedly demonstrated that in open areas an RTK point using a bipod for stability will match a total station to 0.01′ horizontally and 0.025 feet vertically.
In my somewhat limited experience, I’d qualify the above statement by saying that you need a base station within a reasonable distance of the rover to reliably obtain this kind of agreement. A base 10 km away won’t dependably deliver those kind of results.
-
I’d like to see you basically “know what you are doing”, before you decide what to buy.
Unfortunately, thats not often the way it happens.There’s lots of details, in learning:
How certian gear works.
Limitations of certian gear.
Just where those “boundaries” are.
A little investment NOW, will pay off well later.
Now, I don’t know where you live and work, but, I’d like to see you work a day with ANY brand equipment, and then a day in the office, analyzing what happened.
Do this with at least 2 brands of gear. 3 brands would be better.
It’s a bit like choosing a wife…
The BEST one for you, may not be the first one that catches your eye.
A little determined education now, with some practical experience, can put you well onto the right track.
Plus, you will understand, in a hands on way, how a given practicioner works. As well as his equipment.
A bit of investment now, can become the foundation, for your choices, which is what life’s about…
I worked for a day, with another surveyor, several times, before i jumped in the pool.
I think thats wise.
N -
Conrad, post: 406982, member: 6642 wrote: Nate made a comment in another thread regarding localisations. It turned out, if I remember correctly, that he was not using the terminology as Leica does. I submit that localisations are a perfectly good method to reconcile TPS and GPS measurements. the scale correction will be to a precision sufficient to surpass most people’s TPS measuring accuracy. The orientation can be rigorous when on true or grid, or can be taken from a baseline long enough when doing an custom orientation.
I would like to hear a good reason(s) why localisations are not acceptably accurate. Reasons based on the nature of the localisation procedure.
Localizations are generally Helmert, 7 parameter transformations, calculating the least squares best fit Translation (dN, dE, dU), Rotation (around N-axis, E-axis and U-axis), and scale between two coordinate systems. Several issues can creep in on the uninitiated. There are a lot of moving parts to localizations and if the operator is not careful, serious issues can occur.
From a purist standpoint, the proper way to scale a terrestrial measurement to a projection surface, individual distances should be scaled from the earth surface to the projection. The scale factor is used to scale this terrestrial distance to the projection is determined by both the ellipsoid height of the beginning and end point, but also by the location of the line relative to the projection (particularly how far from the standard parallel for Lambert, central meridian for Transverse Mercator, etc.) Part of the 7 parameter transformation solution is scale. This scale factor is based on a comparison of the two coordinate systems and not based on any geodesy. Some software will allow the user to override the Helmert scale factor and use a geodetic factor, but not all. If not carefully considered, the Helmert solution can hide errors in one of the coordinate systems.
Consider a total station traverse in which the ppm was not properly set, or perhaps the wrong prism constant was used and the distances in the traverse were approximately equivalent. The Helmert scale factor will accommodate the scalar error and hid them. If the operator does not have some reality check on the scale factor that should be so that errors can be identified.
Perhaps most significant is the issue of geometry. For distances, issues of interpolation vs. extrapolation are important. If an operator scales an RTK survey that may span thousands of feet to a total station survey that only spans hundreds of feet, observational errors in the total station survey or the RTK survey will compound.
Beyond distances, rotations can also present a dilemma. Again geometry is very important. Tilting the reference plane for elevations away from the Geoid can be dicey. If the points being used for vertical control do not extend well beyond the project, then elevations of points extending beyond the control will be subject to compounded error. Some software will allow the tilt values to be overridden. With multiple points, this results in an average translation based on the multiple points.
In all of this, localizations can be incredibly efficient for mating two systems, but the user really needs to be knowledgeable. Most users I’ve encountered want to use localizations so that they can work in 5000,5000 systems just like they did with their total stations and not worry about projections and geodesy in general. I think this is a dangerous scenario.
-
My idea of “Localization” is not what many others do. For me, localize is to Assign a 50k,50k coord system. This keeps it “Local”. I KNOW they are in fact local coords. And, assign a scale factor. The 3rd thing I can do, is assign a rotation. This is for matching a local system.
Now, there is ANOTHER kind of localize.
Simply ASSIGN “This point equals this point” a number of times, check the residuals, and go along.
This kind of localize, can HIDE mistakes in the first coord system. Of varying magnitude.
IF doing this, you should ALSO, only hold one VERTICAL, and then compare.
Also, hold ONE point, and compare the rest. This will TELL you how much it is changing.Hold all verticals, and compare. perform a fairly rigorous look, to SEE how much, and WHAT kind of distortion you are introducing.
Think of performing a “best fit” on a table, with a short leg, (so it is not level) and with a short side, so the scale factor is goofed.Find a knowledgeable surveyor, and spend a day with him. If you have to PAY for this, do it. WELL WORTH it.
Another surveyor, (That I actually taught to hold a plumb bob, when I was a kid!) but, he worked for other surveyors, on his road to licensure, said this to me: “I recommend that you work for AT LEAST 3 different surveyors, before you get licensed.”
In a similar sense, I am recommending that you “invest in yourself” and rub elbows, with a few fellow professionals, as you enter this new (to you) field.
You will make life long friends, learn stuff that is not readily available, and become a “More Professional” surveyor.
All good things.I’m trying to “Help you win”.
And not step in a poop pile.
N
-
Loyal, post: 406946, member: 228 wrote: If you are talking about a å? acre Lot, then the Total Station should kick the RTK’s butt.
You got dat right! 😉
I am in production housing layout and have had every damn GPS, RTK/Base do-dad, known to mankind demoed to show me ÛÏHow to increase my bottom lineÛ and nothing, and I mean nothing, is faster than a two man crew with a total station. -
Conrad, post: 406982, member: 6642 wrote: Nate made a comment in another thread regarding localisations. It turned out, if I remember correctly, that he was not using the terminology as Leica does. I submit that localisations are a perfectly good method to reconcile TPS and GPS measurements. the scale correction will be to a precision sufficient to surpass most people’s TPS measuring accuracy. The orientation can be rigorous when on true or grid, or can be taken from a baseline long enough when doing an custom orientation.
I would like to hear a good reason(s) why localisations are not acceptably accurate. Reasons based on the nature of the localisation procedure.
If the term localization is used the same way a calibration I hesitate to apply it to a job. Basically you are fitting GPS observations to an existing system of some sort, so all the adjustments, misclosures and rotations from the older control is imbedded into your new GPS observations.
It was originally sold to do two things in the early days.
1. To get GPS receivers “onto” existing horizontal control (mainly to stake something, or continue a project that was already started). This was generally ties into NAD27 control or some type of 10k, 10k coordinates.2. The more important function was to get GPS “onto” vertical control by calibrating.
Verticals were a vexing issue cause it was quickly apparent that GPS did a poor job of returning elevations, if you had a benign Geoid Model in the area the calibration will have some issues, but in more undulating Geoid Model areas they are basically useless over any distance; and woe be it to anyone who surveys outside the calibrated area, elevations will quickly degrade.
I will advise that if you have to calibrate, do it with redundant static observations and look at everything in the computer, then apply it with extreme prejudice and never, never update it for any reason.
Localizations (I think) are sometimes used as a synonym for creating a LDP around a point and not any type of calibration; which is a different discussion.
-
Charlie c, post: 406942, member: 6798 wrote: I am considering the purchase of my first RTK unit and am confused by the vendors I have been talking to.
One says RTK cannot be used for urban and suburban lots and is questionable for Elevation Certs. He says it is usable for large projects and control, but the GPS points will not agree with TS results. He says there are ways of making adjustments to compensate for the differences I will get between TS and GPS.
Another says just the opposite, best thing since sliced bread.
Most of my work is suburban lots and acreage, usually 10 acres or less, but occasionally 100 to 500 acres.
The topography is hilly and ranges 400 to 1500 foot elevation with acreage usually having a large number of trees.
There is no other surveyor in my area using GPS, (maybe that should tell me something) so there is no one to ask questions of and advice.
What do you use GPS for?
Do you find differences between GPS and TS distances and how do you reconcile these?
Where would you suggest I go for information?
Happy New Year to all, and thank you for any help you can provide.RTK/VRS is a great tool to use in the proper settings and with the ability to know when and where to use. For instance, on those jobs you call “acreage surveys” that are wooded and change dramatically in elevation, RTK/VRS may not be the most practical tool to use based on tree canopy or topographical obstructions. The same holds true in urban areas where the sky line can provide obstructions to the sats. Visibility is the key.
You will always see differences in baseline measurements between terrestrial and GPS observations (both Horizontally & vertically) for a number or reasons, all of which can be overcome easily with proper data adjustments.
As a general rule, where the distances/elevations need to be tighter than a tenth (usually the extreme in differences) I would go with the TS.
I would suggest renting before buying, getting training on the rental and employing it under your specific site conditions to see if it is the tool that works for you.
I’m sure that you will find that there will be somewhere around a 50/50 split on which collection method you use but that 50% where RTK is the answer could amount in a huge savings in time and efficiency, more than enough to make the purchase practical.
My work is in a largely urban setting with mostly obstructed skies so using GPS is not really practical or efficient. When the need arises for me to be on State Plane or I get job requiring Elevation Certificates I either rent the equipment or hire a friend to set me point pairs for my crews to base their ground work on. -
I’ll only speak to my own experience. Both conventional TS and GPS (static & RTK), are amazing tools if the operator has taken the time to understand their strengths and weeknesses. The real value is using them in combination comes from getting work into a common real world projection like state plane which compensates for convergence. Over time the value of going this route becomes obvious as you can begin to relate older work to new work and gradually build an over all control file for a region allowing for much more efficient recycling of older work. After a decade of this approach there are few areas where I work where I already know within a foot or two where to look for new control to add, saving me a lot of scratching around for something to use to get going. Even if all my control points in area get wiped out, if my data was processed using CORS data, with very little effort I can reestablish that network again within a couple centimeters. The real value is in combining them to rock and roll with whatever gets thrown your way.
Willy -
Localizations are great for containing error within a old Survey done by total station, locate existing monuments and hold their values, makes sense, but it’s just a rubber sheet. If you leave the area within the constraint, you no longer are containing error, but introducing error. I believe that the Lieca localizations had a way to hold ground distance and not scale distances to match coordinate values. A much preferred method.
Residuals for a localization can be misleading if the distances are included.I wouldn’t let the can go wrong issues, discourage you. With a little common sense GPS can be the best tool on your truck.
-
FL/GA PLS., post: 407042, member: 379 wrote: and nothing, and I mean nothing, is faster than a two man crew with a total station.
I located 11,000 points for pylons on a solar array project and worked alongside my two sons who were running a total station and my 2 man crew was easily setting more points then they were. We switched with them because I felt there should not have been such a big difference. Same results. Using a total station requires several shots to set a point the GPS is continuously refreshing it’s position. Although I do remember radially setting building corners with a chain and theodolite at a pretty fast pace but this was a 15 acre site.
-
Over the last year we have started using RTK a lot in the woods especially on your 10 acre and up sites. This time of the year with no leaves we are getting very good reproducible results with just a few minutes spent on each point. When the leaves are out we may have to spend twenty or twenty five minutes on each point. While that seems like a long time on each point, when you take into account there is no instrument set up time or backsights (we use two men crews most of the time), we still do better with rtk if there are not a bunch of corners that can be located from one set up. Another point in favor of getting a system is each manufacturer is pushing the limits and improving every year. Why not go ahead and make the jump and get ahead of the curve on learning how and where you can use it. We are the only one in our county with a rtk system and we are doing more in a day with one person than some are with three.
-
billvhill, post: 407108, member: 8398 wrote: Using a total station requires several shots to set a point the GPS is continuously refreshing it’s position.
A total station in tracking mode — which is typically used during staking — also refreshes continuously. I don’t see any difference between the two technologies there, except that the total station position is likely to be tighter with fewer readings.
-
Shawn Billings, post: 407019, member: 6521 wrote: Localizations are generally Helmert, 7 parameter transformations, calculating the least squares best fit Translation (dN, dE, dU), Rotation (around N-axis, E-axis and U-axis), and scale between two coordinate systems. Several issues can creep in on the uninitiated. There are a lot of moving parts to localizations and if the operator is not careful, serious issues can occur.
Thanks for the long reply.
I would describe transformations as transformations. I wouldn’t say ‘localisations are generally transformations’. At least as a leica user anyway. Localisation by leica refers to a method where the local scale is calculated (probably) rigorously at a point on a job specified by the user. The scaled GPS distances will exactly match properly made TPS observations. If they don’t match within expected tolerances then the TPS measurements were probably sloppy.
From a purist standpoint, the proper way to scale a terrestrial measurement to a projection surface, individual distances should be scaled from the earth surface to the projection. The scale factor is used to scale this terrestrial distance to the projection is determined by both the ellipsoid height of the beginning and end point…
This is done in the leica 1 point localisation method in order to scale the GPS to TPS. It is as good as you’d want.
I feel like your objections are not focussed on the issue of localisations as a valid method, rather on potential ways to misuse them, but this applies to anything misused. Perhaps like saying be wary of anyone trying to offer you a powerful car, because with improper steering and throttle input it’s a good way to die in a fiery wreck!?
-
Conrad, post: 407120, member: 6642 wrote: I would describe transformations as transformations. I wouldn’t say ‘localisations are generally transformations’. At least as a leica user anyway. Localisation by leica refers to a method where the local scale is calculated (probably) rigorously at a point on a job specified by the user. The scaled GPS distances will exactly match properly made TPS observations. If they don’t match within expected tolerances then the TPS measurements were probably sloppy.
I would describe transformations as transformations as well. In this case, I said localizations are generally transformations, and I stand by this generalization.
I’m fairly familiar with a few localization/calibration applications. In the case of a so-called one point localization in which the underlying coordinate system is geographic, the software develops a projection at that point, a sort of on-the-fly LDP creator. No transformation involved.
One software that I am aware of will not allow for the underlying coordinate system to be geographic and obligates the user to select a projection. The one point localization in this case is a transformation. The user then selects Grid/Ground scaling (ostensibly using the combined factor of the selected point as the transformation scale factor) and then selects Grid/Geodetic for bearing relation (ostensibly using the convergence angle at the selected point for the U axis rotation). This is a simple method but has some overhead, namely that the new system is subject to the same rate of change in distortion as the underlying system, whereas a custom projection will have a lower rate of change in distortion from the selected point.
In the case of multi-point localizations, these are transformations. I used the word “generally” because I exclude one point localizations with underlying geodetic systems.
I’m well versed in localizations, and have calculated them by hand. For the right applications, there is no better tool. Too often, it is relied upon as a magic pill to avoid learning geodetic principles and this is the danger, not the math itself. Localizations (7 parameter Helmert Transformations) are extremely well suited for determining a best-fit set of affine transformation parameters between coordinate systems. They are not, in my opinion, the best way to reconcile measurements. I like using localizations to import surveys based on precise total station surveys (usually based on an assumed coordinate system) into my geodetic coordinate system. I then have a very good estimate of where to look for remaining monuments from the older survey.
I don’t know what you mean by:
Conrad, post: 407120, member: 6642 wrote: Localisation by leica refers to a method where the local scale is calculated (probably) rigorously at a point on a job specified by the user.
Are you referring to a geodetically determined scale factor or the least squares, best-fit, scale factor from observations? If you are referring to determining the scale factor from observations, then it may not necessarily be sloppy total station measurements. In a short baseline, the scale distortion of GNSS is typically going to be inherently less accurate than total station, due to a higher constant error in GNSS. This would produce a poor scale factor. Conversely, if a scale factor is determined from a long baseline, the scale distortion of the total station is typically going to be inherently less accurate than GNSS due to compounding set up errors (if the baseline required more than one total station setup) and the generally higher ppm error of total stations compared to GNSS. The best method is to use a geodetically determined scale factor rather than allow observational biases to influence the scale factor that is produced from a Helmert 7 parameter transformation.
-
Bruce Small, post: 406969, member: 1201 wrote: GPS RTK is the greatest gold mine of all time, in the right hands. I have repeatedly demonstrated that in open areas an RTK point using a bipod for stability will match a total station to 0.01′ horizontally and 0.025 feet vertically. I have achieved the same results on the Tucson calibrated baseline. My usual procedure is to take every important shot five times, in succession, and the unit averages them.
Don’t use a bipod, orient yourself any which way on different shots, run the rod up and down using a bubble calibrated a year or two ago, and your results will vary. Not may vary – will certainly vary.
Before you buy any unit, most important is a hands on demonstration on one of your typical survey sites, including uploading and downloading.
“… 0.025 feet vertically”
You get this consistently? We see day-to-day vertical differences using a CORS as a base ranging from 0.00′ to 0.20′. Same points, same calibrated equipment, same occupation time, same data file, different constellations.
-
I have a suggestion, for those of you who are near a calibrated baseline. Some weekend head over there and take RTK shots on the brass caps, compare your results with the published data, and report back to us. Very important to use a bipod or otherwise brace the rod so it is very steady (and I don’t mean with two wood stakes).
-
Bruce Small, post: 406969, member: 1201 wrote: I have repeatedly demonstrated that in open areas an RTK point using a bipod for stability will match a total station to 0.01′ horizontally and 0.025 feet vertically.
Jim in AZ, post: 407188, member: 249 wrote: You get this consistently? We see day-to-day vertical differences using a CORS as a base ranging from 0.00′ to 0.20′.
I believe that the operative words in Bruce’s statement are “in open areas”.
-
Jim Frame, post: 407112, member: 10 wrote: A total station in tracking mode — which is typically used during staking — also refreshes continuously. I don’t see any difference between the two technologies there, except that the total station position is likely to be tighter with fewer readings.
This is probably one of the moments where I should have left alone. But here I go
Ok, so maybe we can attribute it to the instrument man not relaying the message every second or following the rod man fast enough.
The RTK is faster regardless of the reasons.
Yes the total station is more accurate until about 500 feet then the tables begin to turn.I’m not knocking total stations if that’s what one prefers to work with, but relaying my experience to disbute the comment that nothing beats a two man crew with a total station.
Log in to reply.