Extreme rtpk-ism, the deep dark holes of surveying

Nate The Surveyor

Well-Known Member
I've been using Javad gear now for about 5 years.
In the bad places we go, I believe RTPK is going to replace RTK, for solid answers.
Mr. Matt Johnston asked me in another thread, about how I would suggest that we evaluate RTPK .
Well, this so deeply affects us, that I thought it deserves a thread of its own.
I'd like to propose that we first consider some of rtpk's limitations. One is it requires communication base to rover. Sometimes in these holes, com's are intermittent or poor. This can interfere.
Also, as it is now, if you sit on a place for an extended time period, (say 10 minutes) the whole thing restarts, from zero, and goes to 10 minutes and restarts. Each of these 10 minute periods is repeated, and after each one, scored. It's scored for separation. (Time separation) and ellipse size and shape. After hitting store, the RTPK software uses a weighted average, and averages these 10 minute sessions into one master coord. It may work better to use 6 minute sessions, or some other number.
It should then store this RTPK as a piggy back coord, as it now does.
DPOS should mimic this process, so that if some of the data is MISSING, from the RTPK solution, that it is put back via DPOS. Or, DPOS could be allowed via a form of post processed RTPK, after the base is downloaded to the rover.
This mechanism would allow extended RTPK sessions, post processed RTPK, where signals were low, intermittent, or altogether missing.
Post Processed RTPK, should be allowed, without shutting down the base, but it would only be for selected points, or points with poor or missing base data.
Via Bluetooth.
I want to thank the whole team that's making all this wonderful stuff happen. I especially want to pay respects to Mr Ashjaee's vision to allow us, the users of this equipment to make suggestions, and comments on the development, and mechanisms that we use to perform land surveying. I realize I am using equipment, that I only understand in a limited fashion. So, some of my ideas are not implementable.
That's ok. The whole idea is to get good data fast. And to build an invincible survey machine.
Also, someday, we want automated rod height, even if rod height is changed during an observation.
Someday? Yes. It will become commonplace.
Now, where is my T-3?
Nate
 

Matt Johnson

Well-Known Member
5PLS
One is it requires communication base to rover.
RTPK has the ability to process data when the communication stream is lost. You must download the data from the base and then you can use the "Post-Process Selected Point" option in context menu to process the point. When processed this way sessions are created with the time period for the Extreme category.

Also, as it is now, if you sit on a place for an extended time period, (say 10 minutes) the whole thing restarts, from zero, and goes to 10 minutes and restarts. Each of these 10 minute periods is repeated, and after each one, scored. It's scored for separation. (Time separation) and ellipse size and shape. After hitting store, the RTPK software uses a weighted average, and averages these 10 minute sessions into one master coord. It may work better to use 6 minute sessions, or some other number.
It should then store this RTPK as a piggy back coord, as it now does.
It already works great in the worst location without having to add more complexity but if you are not happy with this there is already a "Max Total Duration" setting in How to Stop that you can limit the point duration to whatever value you choose and you can turn on Auto Re-Start to collect additional points.

It's scored for separation. (Time separation) and ellipse size and shape.
In multipath environments error estimate and ellipsoid values are not valid and you should not rely on them. I find them to underestimate the ellipsoid values to be underestimated by a factor if 3 to 4 in multipath environments.
 
Last edited:

Nate The Surveyor

Well-Known Member
In my dreams, I have integrated forms of rtk, RTPK, and ppk.
I have a total of 60 minutes on a single point.
We achieve 100% confidence that we are within 0.2' at the 8 minute mark.
We achieve 100% confidence we are within 0.10' at the 20 minute mark.
And at the 1 HR mark, we are pretty sure we are down to the 0.03', but we are 100% that we are down to 0.05'.
This is a very challenging point.

I'll share a few pics from some recent jobs, later.
It seems that some challenging locations, it's just got to have a larger error ellipse, than we'd prefer, and that is a part of the deal, with these tools. No amount of time can give you 0.005' accuracy. The idea is to give the user the control over this time on point/accuracy equation. In the field, we have to decide.... Stay another 3 minutes.... (Or revisit this point at another time), or stay another 15 minutes, or such.
Thank you Matt Johnson, and Shawn Billings, and all the others, who have invested so much already.
Nate
 

Nate The Surveyor

Well-Known Member
8346D88C-3543-4E65-8889-52319B601FD5.jpeg
Ive been on a job lately, with all shots like this.
 
Top