Enforcement is a completely separate issue.
There is not "Measurement Police", fortunately for some.
Rather, barring some case where damage is immediately discernable and attributable, the majority of violations would be based on retracements by other surveyors.
In other words, if I retrace your survey and find that I disagree with your results on, say, 50% of the measurements by more than 0.20’, using the standard in Texas, I could send my survey and yours to the Board with a request for them to review your work.
But, that’s not what we’re discussing here.
I’m completely intrigued by the question Kent has raised. I’m really interested in what others have to say.
Once we get over the harsh method of delivery, I really don’t think Kent is trying to stand on any pedestal, either. To me, Kent is legitimately asking for discussion, not pointing to failings in others.
An extension of Kent’s discussion could well take the path of looking for methods of enforcement.
For now, let’s focus on the question at hand: Based on positional tolerance standards, what level of precision must be met to meet those standards? 100%? 99%? 95%? Why?
In other words: how much can you get away with?
Wow! That’s probably realistic when dealing with the bottom feeders, though.
I hold myself to a higher standard.
So…to those of who deserve the appellation of “professional”, what standard do we apply to ourselves?
Based on positional tolerance standards, what level of precision must be met to meet those standards? 100%? 99%? 95%? Why?
As written I believe it is 100% or absolute. Statistics don't like absolutes, though. From work we've done over the years and have returned to, we've consistently held that standard (occassions have shown different, but they've always been explainable). The problem with relying on repeat observations with the same equipment is that similar methods and similar equipment are likely to give similar answers (even if they are wrong). GPS has really been beneficial from this standpoint because the difference in methodology between GPS and terrestrial measurements are so extreme that agreement between a measurement determined from each method is about as close to absolute as one can get. Is it realistic to measure each vector with GPS and a total station? Of course not. Is it possible to get a sampling (whether measuring a couple of lines with the total station to prove an RTK localization or static pairs to confirm a total station traverse)? Absolutely (pun intended).
As to not using RTK for control or boundary. We just purchased RTK a month ago. We used it on our first real job this past week. We were performing an as-built on six acres of highly developed commercial property. From reading on the old RPLS.com board, experienced RTK users suggested that 3 minutes produces results about as good as one can get from RTK. I've tentatively held to that rule and been pleased with the results. Of the nine monuments we found in 2010, 5 were still intact. Giving each of them a 3 minute observation produced residuals (compared to our original static GPS spined total station survey) of less than 0.03 foot. We also needed to perform some total station work. I established two pairs, again using three minute observations on each of the four points. The total station check on each pair had less than 0.01 foot horizontal and 0.02 foot vertical residuals. I would take this to suggest that on that day, using similar methods (3 minute RTK observations), on that site, that all points positioned with RTK should have similar accuracy, which would easily exceed the State's 0.10 foot minimal tolerance.
I think that the most appropriate way to test compliance with the standard is on a methods and practices level. I wouldn't rule out some uncommon situation where damages result from a misreported distance between two incontestably stable monuments, though.
Surveyors who perform ALTA surveys have to meet a different standard, but one that is defined in such a way (using the length of the semi-major axis of a 95% confidence error ellipse as a measure) that methods and practices are implicitly the subject.
Ah! Now I begin to understand the underlying point of the question! Excellent!
When I perform a boundary survey for a Land Title Survey, I collect the positions of the found monument using fast static GPS methods. That gives me the ability to generate a statistical analysis of the network of points found as well as control points set. Now, I CAN make a statement as to the positional accuracy of my boundary work!
The topographic data I collect, most likely for design purposes, is often collected using RTK methods or photogrammetry. With a single “shot” on, say the top of curb at a BC point, I cannot make any statement as to the positional tolerance of that shot. Nor do I really need to. The effect of that point’s positional accuracy is nothing compared to that of the found and accepted monument marking the corner of the property. That’s why the 2011 MINIMUM STANDARD DETAIL REQUIREMENTS FOR ALTA/ACSM LAND TITLE SURVEYS, §3E sets the bar for “…the monuments or witnesses marking the corners of the surveyed property…”
Oh…and I’m the one who collects identifiable hard points as check points on projects employing photogrammetry. Points are collected in multiples of 20 to make the math straight forward when check the product delivered to me. So, yes, I CAN make a statistical statement about that data.
I get your point regarding methodology, though.
In “days of olde”, results some surveyors reported using a single angle and distance measured by a 20” instrument sloppily set up and a chain pulled without care would not meet the standards. That crew knew that their work was substandard.
Today, the practitioner who has paid $30k for an RTK set up, but has no real training or understanding of the technology likely has no idea that the measurement reported fail to meet the standards.
Unfortunately, the few people I know who have spoken out about methods v positional standards have all given up their crusade.
Anyway, I have work piled up around my ears and need to buckle down. I’ll do what I can, Kent, but…no promises that I’ll be any more successful in keeping the guard rails up on this topic, even as important a discussion as it is!
Excellent post, Shawn, as usual.
"As written I believe it is 100% or absolute. Statistics don't like absolutes, though."
I mulled over Kent's post, hoping to blast him, as others have. However, after re-reading the thread a few times, I began to understand his question. I came to the same realization that you did, Shawn. That's why I posted.
In short, I agree. There is an absolute here.
The problem is that many methods, such as RTK collecting only one data point (not 3 minutes worth of data) are widely used and widely misused.
Hence the question: is ok good enough or do we really have to hold out for true compliance with the positional standard?
OK, if we hypothesize that the Texas rules cannot be rigorously enforced on the majority of monuments, and we are asking how a professional who wants to feel like a professional should restate the rule so the professional can enforce the restated rule on his/her own work, two questions come to mind.
1. In cases where the rule could be enforced, such as monuments set in bedrock, has there ever been a disciplinary case where the rule was enforced, and if so, what were the criteria?
2. What was the original purpose of the rule? Why is it so tight? Was the testimony at the public hearing that preceded rule adoption preserved? Who testified? Was it a bunch of survey equipment manufacturers who wanted to force surveyors to scrap old equipment and buy new equipment?
A current purpose of the rule would be to allow all the monuments to be reset to a satisfactory degree of accuracy if any two monuments, that are a reasonable distance apart, survive. So define for yourself what "reasonable distance apart" means, then decide how accurately the boundary needs to be for practical purposes in a given setting, and see how the error in the location of the surviving monuments would propagate.
> So…to those of who deserve the appellation of “professional”, what standard do we apply to ourselves?
>
> Based on positional tolerance standards, what level of precision must be met to meet those standards? 100%? 99%? 95%? Why?
You are correct, and I sort of "inferred" ashton was a bottom feeder. I don't know ashton at all, and take back the inferrance.
That is how I read the question.....Who's going to enforce it anyway? To me that means that what do I care if I keep within the legislated tolerance (although that isn't what ashton said). My argument for keeping inside of a required tolerance, is mainly because I hold myself out to be a "professional". It's a form of protecting the public for them to hire a land surveyor to do their work and expect a higher level of competence than someone with a measuring machine and no license.
But realistically, I imagine the main time it would be enforced is if a court case was in effect, and it came out that one of the surveyors didn't do his/her work to minimum professional standards. Most the time I think the surveyor can get away with it. I don't envision another surveyor saying "he's 0.2 off, he's 0.2' off to the State Board. I would more expect he might give the guy a call.
Based on positional tolerance standards, I have felt that the standard level of precision that is used in the industry is a 95% confidence level. Some jobs might require a higher level of confidence. I sure don't want to be responsible for moving a building. If I building is designed to be 1' inside of a setback I might not worry about the precision of the location of the overall building to a better precision of +/-0.1 with 95% (or less) certainly; but I would be concerned with the staking between corners at a much more precise level like +/- 0.01 with 99% confidence. Some of that might be a professional judgement. But I would be taking more measurements of the building sides, and diagonals for the walls of the building than for the baseline of the building in that case.
"Some jobs might require a higher level of confidence."
And some WAY less...
Major League Batters making over $100k PER GAME...
Weatherpersons...
😛
Sorry, I couldn't resist lightening the thread a bit.
Perspective!!!
thanks for the kudos. it sucks to be in agreement with Kent.
😉
it is sad that this discussion always brings up the "I'd rather have a loose measruement between right monuments than a tight measurement between wrong monuments" as if it has to be one or the other. If you measure accurately, then you are using the wrong monuments... warped thinking.
in the Bible, James said faith without works is dead. show me your faith without works and I'll show you my faith by my works...
In the same attitude, show me your right monuments with poor measurements and I'll show you the right monuments with accurate measurements. it can be done...
You lost me with the 0.5 factor. If the north and east components have standard error sigma, then they combine as a Rayliegh distribution and the 95% confidence point for radius is 2.447 sigma. So if each component is 0.04 the 95% bound is +/- 0.10 ft.
Probably 99% of the public, lawmakers, and perhaps even regulatory bodies cannot understand statistics. It's one of the biggest gaps in the US educational system. As soon as you raise the point of confidence levels you are comparing yourself in their minds to weathermen, with the associated low respect for reliability.
So regulations get written in absolutes. The measurement must be within limits 100% of the time. A power plant must emit ZERO radiation, not just less than the natural ambient. Etc.
:good:
I think it's generally known that using RTK or GPS is not real tight at short distances. It might be bad practice.
What about the opposite, long distances. Say a two mile traverse that required a lot of legs due to the terrain/vegetation. RTK has about the same 1-2 cm spread at two miles as at 50 feet. So in one measurement I can get the two miles under the spec. Is the same true with a long many leg traverse. So is it bad practice or incompetence to use a total station to measure long many leg traverses? The potential for error adds up with every leg. You got to get the temp and pressure right. Long shots add up the angular error. Lots of setup error adding in along the way.
What about 25 miles? Any body going to traverse 25 miles and claim a LS proves the actual distance is within 0.1 feet. With GPS quite easy.
I suppose there are correct tools for every job. I always used a robot on construction layout just because I didn't have enough confidence with the RTK staking a lot of points close together, particularly the elevations. But I've seen a lot of folks do it and really never heard about a lot of problems. So maybe I just worked harder and longer for my money. But I have done a lot of earthwork, roads and pipelines with GPS that doesn't need to be spot on.
Early on I did some of my own testing such as a curb and gutter layout. Measured the points with the RTK and then ran a digital level loop through the whole thing. Result - unless the C & G has a lot of slope, better leave the RTK in the truck. I checked control both with RTK and robot so I gained some experience on how it worked. From there I just developed a sense of judgment about what the job required and what really mattered for accuracy. Some things are OK at a less than real tight accuracy, it just depends. If you are not building the Swiss watch all your measurements don't need to be spot on. Applying bad judgment leads you towards incompetence. It's not all wrapped up in a statistical number.
Some states boundary specs, especially for rural surveys are easily met with proper use of RTK GPS.
LR, posing questions as you have has been treated like toxic waste around here lately. See what happens in the repsonses.
I agree with some of your points, but going into that would be reheating too many times. One other angle of convtroversy is the assertion by some who say there is no cost or time difference between using GPS and a total station in a situation like your example, provided both are properly done and meet the a standard, liek those rural standards you mention. Do you think the many-leg traverse, properly done, over many miles is the same amount of time/cost as a properly done static GPS campaign? It is hard to see how that can be, but whatever. And I have not really seen anyone say it costs too much to MEET a spec, a spec must be met of course, but do you think EXCEEDING a spec might add costs/time?
Error Does Not Equal Incompetence
"The premise is that it is given that an error above some threshold value that is present in the relative position of two survey markers you report is incompetent practice."
The premise is incorrect.
I agree. Ian just made this question a lot more interesting by wondering if the standard should be checked by statistical means (LSA).
The standard is an old fashioned closed traverse precision check. I think it assumes you will do a traditional closed traverse and check your closure and it should be better than the standard although that may be misleading.
±0.10' + 1/10,000.
So say I measure between two monuments about 100.00 feet apart with my steel tape and do all of the corrections, etc. I determine the monuments are 100.05 feet apart. Who is going to check that for compliance with the standards? Suppose the County Surveyor checks my measurement using a much better method (maybe that EDM NGS uses on baselines) and finds the distances is really 99.90 feet. But say I did a close traverse with my transit and steel tape and my closure calculates to be better than the standard. Am I in violation?
> "Some jobs might require a higher level of confidence."
>
> And some WAY less...
I would think you would want a relatively high level of confidence. To me, some jobs require a high level of precision and others WAY less. ie that 3-foot dia tree's location is precise to +/-1' w/95% level of confidence.
> I would think you would want a relatively high level of confidence. To me, some jobs require a high level of precision and others WAY less. ie that 3-foot dia tree's location is precise to +/-1' w/95% level of confidence.
Probably a more standard term to substitute for "precision" in the above sentence is "uncertainty". Every survey measurement is subject to some uncertainty and so is every quantity calculated from survey measurements. The indicator of quality is what the limits of that uncertainty are at some particular probability.
Saying that a distance is, for example, 101.24 ft. +/-0.04 ft. (95% confidence level) means that there is only a 5% chance that the measurement value of 101.24 ft. is in error by more than 0.04 ft.
Likelihood of...
> Whenever the rally call for folks to help contribute to such efforts gets floated it seems much easier to knock down than to try to help build up.
Ain't that the truth....;-)
I hope you are enjoying the Labor Day Weekend; I know I am!
Esprit De Corps!
Radar
I think your grammatical advise took a bit away from my point, but after rereading, I guess I would agree that uncertainty would probably have been a better term to use. "precision" and "accuracy" are so misunderstood and the difference is only used in statistical circles (if I am not mistaken) that using the term "precision" as I did can only lend itself to possible confuseion.
To Ian's point:
> I would think you would want a relatively high level of confidence. To me, some jobs require a high level of precision uncertainty, and others WAY less. ie: that 3-foot dia tree's location is precise to +/-1' (of uncertainty) w/95% level of confidence.