Standard Setup on Resect-Solution after Resection?

OK, I'll bite.

Why not just...start surveying after completing the resection?

That typically will work well. If you have bad coords on one of the points you will be able to detect it by taking check shots to those prisms after you have used the solved location for your set up. As long as your residuals remain within your tolerance and your check shots within your tolerance you know you are good to go.

Why not just…start surveying after completing the resection?

Because you won't be "zeroed" on any one of the resected points. You will be oriented to an "average" of the resected points. Some people can't fathom that. I think of it as a plus. But if the instrument drifts out of plumb enough to need to be rezeroed at some point it can be a conundrum.

Resection is a perfectly valid measurement technique. The pros and cons depend on the quality of the existing control you are using, the quality of the resection solution, and the error budget of the work you are doing. I usually set a point under my resection setup, so that I can use it again in the future, if my resection solution is good enough.

Here are some possible reasons:

- the software I'm using (SurveyPro on a TSC3), requires a "Fixed Backsight" when doing repetition shots. The software does not allow you to set a "Fixed Backsight" in a resection.

- when set up using a resection, the check shots on your control will all be off by the residuals from your solution... (trying to think if this changes when adding a standard setup using the solution from the resection...)

There's also a check shot feature I would like to use that automatically turns to the backsight, switches to the backsight prism, takes a check shot, and then returns to the roving prism, but this also requires a "Fixed Backsight."

So far all this is sounding reasonable (especially the comment about drifting off zero or coming out of level).

We spin out new control from our resections all the time, but I'm curious if this technique would be preferable for setting new control (since we could throw in some repetition shots).

"Because you won’t be “zeroed” on any one of the resected points. You will be oriented to an “average” of the resected points."

SurvCE gives you the option of zeroing on the initial control shot, which is what I do.

I agree. If it's stuff I need to be pretty tight, I will set a point under the TS and backsight the longest point after the resection. Then I can check backsight whenever I want and re zero as needed. If it's just a rough staking for excavation then I won't bother.

Trimble access requires every shot to be checked as "backsight", which threw me off the first time trying to use the s5 to do a resection or it just fails to work. That was a frustrating hour. Because I used to use microsurvey Fieldgenius, I would always leave the furthest point for last..which would work out if I want it backsight that one... because Fieldgenius crashes so frequently if you have to navigate the map to find your points...I would want the crash to happen as I got further away so I would be walking back when shooting a second time after the crash.

SurvCE gives you the option of zeroing on the initial control shot, which is what I do.

That's how Access operates - the first observed point is the "First BS" and the zero orientation.

But of course, if I get a goofy angular check after I've been set up with a resect, I'm going to have to reobserve all the backsights to get as good of a result as before. That's true regardless of what software is being used.

– the software I’m using (SurveyPro on a TSC3), requires a “Fixed Backsight” when doing repetition shots. The software does not allow you to set a “Fixed Backsight” in a resection.

Ugh. Been a long time since I worked with Survey Pro, but if that is the case...that's just weird. Turning sets is turning sets. (Access stores the "first backsight" as the master orientation for a resect and will let you check any point, backsight or not, at any time during the setup, as well as turn sets to any other points too...) Regardless, if that's what you have to do, then that makes sense.

– when set up using a resection, the check shots on your control will all be off by the residuals from your solution… (trying to think if this changes when adding a standard setup using the solution from the resection…)

Consider three points, all of equal quality network accuracy.

Consider a proper resection setup (zero-height, both angles & distances measured) that utilizes all three of those points, versus a single-backsight setup that utilizes only two out of the three.

Observations from the resect setup will have better network accuracy than from the single-backsight setup, as the resection is using an overdetermined solution to compute its position. The single-backsight solution will often look better on paper, but it doesn't give you residuals (other than the distance check). Check shots are still just check shots, and they will always be "off" by your setup errors, whether your setup is a single-backsight or a resection.

The resection will look at the third point and incorporate that into the solution; all you can do for a single-backsight setup is observe that third point and say "yup, that's close enough"....but that check in and of itself doesn't improve the setup.

Going back to your workflow, once you end the resection and perform another station setup, you'll be changing your orientation from a best-fit of three valid points to a single observation of one point. That's probably just fine if all you needed the resection for was to set a new point and you're not looking to get the tightest possible results.

I can definitely see time savings by avoiding having to set up on an existing point and turn sets to the new point, and a resect would get you a better position to start off from for the new setup. I like that aspect of it - but I likely wouldn't add another station setup unless the software forced me to.

I've run into a few that run a tape down to the dirt because they thought they needed a HI to have correct elevations on a resection.

Amen. I am amazed at the office managers that will not allow a resection period. Say they are not as accurate and would rather spin a fly no closure to another point period. Because they can measure rounds. I say i can turn many face 1 and face 2 sets rounds in a resection as well. I just hear no way its more accurate. Well i just grin then and move on. No since and beating a dead horse. For what its worth i have purposely set 3 points on a straight line. Physical moved a tenth or less and done resections with Trimble access just to prove it freaking works. If the control is good. I probably would not do that on a real job but was the online way i could show its not hand comps and truly hard to mess it up. Even if you mess up a point number the residuals will show that. Just re do.

Thats too funny not having zero to a BS. We had a weird instance where are TSC7 did something funky and lost all our custom survey styles. Last week. Finally got everything rebooted and just used the default trimble survey style. Took BS and checked a third point. All good. We set a few and cked into the last one that was staked the previous day. Sent files off. Next day get a call that we were not zeroed on bs and the 3 points needed to be re staked. I said what. So I grabbed tsc7 and cked survey style. It was set to azimuth. I said are the as staked points not falling in correct location. He said yeah But the angle to BS is wrong. I said its the azimuth my bad but its fine. Oh no it has to be zeroed. Heaven forbid if something happens and we all have to go old school and release the bottom plate and do sets by accumulation ofthe angles like many years ago. I never set zero when hand comping mean angles. Zeros was the enemy for that.

The only reason I tape on a resection is if I'm leaving a point behind and want it to have an elevation in case I set up on it in the future. Otherwise no reason you need to tape obviously, but I find with the Trimble s5 robot not having a correct elevation and HI causes the robot to have a lot more difficulty finding me with GPS search or checking a backsight I set up.

In the settings of gps search you can do no vertical if you want so simply do not use elevation for control. Where I see the biggest issue is people historically use 0 for a no elevation. Trimble 0 is an elevation just like real world. This can burn you especially if you are working on a datum projection like state plane with conventional robot. If you do not wish to cary elevation you can leave the elevation as ? Same if you are caring elevation and say doing a topo and you wish to locate something and you don’t want an elevation computed so the office douse it in contouring. Simply put the ? In for rod height. That point will not have an elevation assigned. I think carlson is about the only other software in the field where they had a process to do something like this. Switch from elevation to non elevation throughout a job. But maybe others do that now. If you do this make sure your project height is correct if working on say state plane as it uses that to scale your distance for cogo and not having elevation etc. i can’t remember now if it was TDS or eagle point but it had a way to attach a code in field we used Z along with the actual field code and when processing back in the office it would null the elevation for that shot. Been to long. I used SMI a lot and i can’t remember how it worked for mapping. Road stake out construction it was the best. Back then. I still miss that as a field software. I still have my SMI construction V …. TDS was what Survey Pro came from I believe. They had Foresight as well for office side. Was not a bad software for its day either.

"I’ve run into a few that run a tape down to the dirt"

I do this routinely so I can include the resected point in my terrain model. A zero HI leaves a lump in the fabric.

Thats pretty smart. Get an extra ground shot that way. I have our stuff set up so no control points and other features are used in the surface. I do have it set up so i can use them and other features like utility markings parking stripes still have the elevations so it gives me some spot checks on the contours and surface though. It doesn’t always work but if the crews do it correctly the simple set rod height to no height. Not 0 but ? So no elevation is computed for items i don’t want to use. Like signs bldg corners etc. or they ar locating something and know the rod height would be bad but we need it planimetrically only. I once had one of those lumps on a dumpster lol.

We basically all just started carrying elevation all the time even though we're frequently only using horizontal. The software likes it better and there are other obvious benefits, you never know when the elevations might come in handy later.

Now it's just the calcs that cause annoyance, I wish turn to had an option to turn off vertical like the GPS search....I usually end up turn to point, then turn to VA 90 to get it back to close again. My project manager has caught on and tries to set elevations to 100 or something close so this doesn't happen, but that's only just started happening.

I have used the ? In the past and don't remember if it causes me any issues, but we just prefer to carry elevation whenever possible now. I guess it would make sense though if you were doing topo and then had to tie in something that could not be measured vertically, just to use the ? for that shot. Guys used to use a crazy number to make it obvious it wasn't elevated but the ? seems like the proper way to do it.

Now it’s just the calcs that cause annoyance, I wish turn to had an option to turn off vertical like the GPS search…

That's what this is for:

https://drive.google.com/file/d/1gcuDavxxY-igsJ_F5-HYoayGhBvbdOgD/view?usp=sharing