RTK idea pile

Nate The Surveyor

Well-Known Member
Why not place a RED button, on profiles, for some ADVANCED settings? Place the "Foreward phase" and "backward phase" there, along with other confusing, complicated, or "expert user" stuff.
Let us experiment with it, and once it becomes understood, then place it for the users.
Thinking out loud here.
Nate
 

Matthew D. Sibole

Well-Known Member
5PLS
View attachment 7971
Well, I guess I did that too. But not on purpose. 42 groups here. How many groups is your record?

I’m set up at the TN conference on the Ky VRS network. I’m 20 miles from the nearest CORS station and I am outside of the triangulation network. This is the most groups I have ever seen.

124B611F-38BD-42AB-8659-C35A6917A2E8.jpeg
BEBC0FB5-389A-495C-8780-A625B0967D74.jpeg
5D52C70D-A8DB-41D8-8351-836AA4D7B5C3.jpeg
 

Nate The Surveyor

Well-Known Member
Foreward phase. Jump fail. Foreward phase again. Jump fail. Foreward phase again.
Now, it makes it to phase 2. Good. Now, it makes it to phase 3. Jump fail. Foreward phase again, and it finishes the shot. Since user is doing manual steps to go to next phase, he is doing this at his own risk. So, shooting twice is advisable.
Nate
 

T.Guisewhite

Active Member
I’ve often thought it would be a good idea to have an option to tighten my tolerance after the full verification has completed. It would be an additional step to weed out some of the tails on the scatter plot that are pulling away from the main cluster center.

In other words the LS would have an larger threshold to allow for points to enter the grouping. But when it comes to actually storing the point the user could tighten up the tolerance to better center the mean on the scatter plot.

You wouldn’t want to do it to every point...just on the odd plots. Here’s an example of one that says the scatter is 0.12’. If you had an option to tighten the group to 0.08’ or less it would trim off some of the tail? Maybe the stray high verticals would disappear?
2525F5F1-839E-4278-ADB4-98F496ADEDDB.png

Not really an extreme example...but here’s another if you had an option on this one to cinch the vertical to 0.10’ the strays would go bye bye:
9736E726-6BDB-4122-8F9B-9F62CA32E55A.png

Anyone else ever wish they could refine a position manually like that?? Thoughts?
 

Nate The Surveyor

Well-Known Member
Yes, I have thought about it.
The LS is designed to do it with the data it collects, from start, to finish. Say, 3-4 minutes.
That represents the best that can be done with that precise data, Sat configuration, and such. If you need it tighter, shoot that point 2-5 times, and average it.
Nate
 

T.Guisewhite

Active Member
Understood... my screen shots above are not the best examples but they do show the average being pulled in-line with the stray tail shots...if it’s using all the data in that group maybe when done there could be one additional option button for “outlier reductions” it wouldn’t be a distance tolerance like I was talking before but maybe a percentage. So it could hone down the statistical outliers in that specific group.

I also see the problems with this maybe over estimating the precision of that vector. Possibly causing it to have an over confident weight going into an adjustment later??

I agree in the multiple location method for sure...it’s quite likely there would be no quantifiable difference to the end result...
 

Nate The Surveyor

Well-Known Member
The advantage of the multiple observation method (among others) is that you can separate observations with time. Ie, shoot the pipe in rock pile 1x, (takes 4.5 min and makes one .085' shot scatter), then go observe another point, then re-observe the pipe in rock pile 20 minutes later. Which takes 6 minutes. And has poorer shot scatter. (Ie, 0.13'). Then re observe 2 hrs later. This time it makes it all in 3 minutes, and has 0.09' shot scatter. The avg. Process will sort it out.
 
Top