Unusual DPOS Results

Aaron S

Active Member
#21
I has to be better than using a CORS 50+km away. You need to throw the radios in the truck and jump on cell. Longer baseline RTK with DPOS.
I don't have any kind of cell coverage in much of the area I work in. But if I could just set up the base/radio and broadcast on the lowest power with no intention of maintaining a UHF link, I could still post process base-rover, couldn't I?
 

Shawn Billings

Shawn Billings
5PLS
#22
Yes. You can actually do this without the radio, I believe. It's been a while since I tried this, but I think you will get a warning that there is no radio attached, but it will still complete the base rover setup process and log raw data.
 

Aaron S

Active Member
#26
Your report looked good to me. All of the residuals of position were very low.
That is a pretty typical result doing it the way I have been. I don't remember seeing any over 2 cm overall, which is why it raised my suspicions when I saw such poor results suddenly. I'll likely end up reshooting those points just to be safe.

I thought I tried starting the base with no radio and it wouldn't let me. I didn't try to force it or override the warning though - I'll revisit that option.
 
#28
One of the things that I greatly like about the LS is you get a visual of the coordinate pile when you take a shot. How much trouble would it be to make this also possible with DPOS?.
If it used five different CORS, then, it would have 5 dots. With a line. And the station name, written.
Or, a color code, ie, red is station, Dequeen, blue is LR. and, then the details are in the coresponding color, or just a color legend.
It would be nice this way so that from a brief glance you could find problematic items. Like, a station, that has stability issues, or such.
A fast pretty pic.
N
 

Matt Johnson

Well-Known Member
5PLS
#29
Excluding stations is a great feature when you know there is a problematic CORS, however, I'm speaking about excluding a vector from a solution. If you exclude a station, I suspect DPOS will replace it with another station, perhaps further away with even poorer results in this example. If we could exclude a vector and readjust, we might be able to refine our results. The operator would need to be careful not to exclude too much of course.
This can be done with Justin but not Justin Link. Justin Link only allows data and projects to be viewed, not processed and adjusted.

DPOS projects can now be downloaded from app.javad.com with the JPR download button.

upload_2017-12-21_21-10-45.png


Then unzip and open the project in Justin or Justin Link if you only want to view the project.

From the top menu enable View > Solutions and then you can choose desired solutions to be "Disabled in adjustment".

upload_2017-12-21_19-39-19.png


Then you can perform the adjustment again

upload_2017-12-21_19-54-45.png


and create a new report.

upload_2017-12-21_19-56-51.png
 

Matt Johnson

Well-Known Member
5PLS
#30
One of the things that I greatly like about the LS is you get a visual of the coordinate pile when you take a shot. How much trouble would it be to make this also possible with DPOS?
DPOS processing with a static engine so there is no epoch by epoch solution or coordinate pile. (I know calling it the "PPK" solution is very confusing since it is post processed with a static engine and not a kinematic engine. I have argued many times that we should call it PPS not PPK.)

If it used five different CORS, then, it would have 5 dots. With a line. And the station name, written.
You mean like what is already possible with Justin or Justin Link? ;)

upload_2017-12-21_20-35-8.png
 

Aaron S

Active Member
#31
So for some reason all my standalone DPOS sessions which have been working all year (and really shouldn't have worked as well as they did) are coming up with lousy RMS results recently. No problem, I'll just set up my T2 base and reverse-shift them in. Well, I have coordinates two different monuments I've shot below.

The first one, points #100 and #101 were shot using RTK (RTN internet rover) using the boundary preset method, fully verified. They agree, no issues. Skip ahead to the other station, which I also shot twice #114 and #115 using the same method. Fully verified and agreeing not quite as well, but the canopy was worse. I didn't think that second one was good enough, so I set up the base to check and reverse shift. Point #119 is the check shot from the local base, DPOS'ed, which is very close to the RTK. Point #120 is the check shot for the second station, from the same local base setup (300 seconds) - not as close.

I guess my question is, with so many methods to collect points with Javad (DPOS, post-processed, RTK, local base, etc.), how do you compare "apples to apples" with so many variables - time, satellites, epochs, etc. all factor in. What do you look at to gauge the quality of points from a variety of sources?


Point100.JPG


Point114.JPG
 

Alexey Razumovsky

Well-Known Member
Staff member
5PLS
#32
Hi Aaron, we have redeployed post-processing engine since the end of last December. I did this to meet requirements of some customers. New engine works well for them.
Unfortunately, your approach to DPOSing was lost. Now I have a new engine that combine all benefits of current engine and process very well the files that you sent to us. I expect to deploy new engine asap.
 

Aaron S

Active Member
#36
Alexey - I have some weird DPOS stuff going on, but the archive file is too big to post here. Is there an email I could send a Dropbox link to you?
 

Aaron S

Active Member
#39
I sent an invite via Dropbox to that address. Please let me know if you need anything else.

The problem seems to be that some of the DPOS sessions contain the word "span" on the processing screen. For some of the rover points, it seems like they didn't get attached to the appropriate base file, so the shifts didn't apply.

You'll probably figure this out, but when I've been out over the last couple days, I would shoot a point 3 times using RTK/RTN, then switch the profile to Triumph 2 base and shoot it 1 time for comparison. The whole time I had the T2 base mag-mounted to the roof of my survey vehicle, with no radio attached. I started the base using the LS and just let it log the whole time I was out.
 

Matt Johnson

Well-Known Member
5PLS
#40
The problem seems to be that some of the DPOS sessions contain the word "span" on the processing screen.
This would an issue in the J-Field DPOS session logic for Michael Stazhkov. I suspect it may already be fixed in the PreRelease version of J-Field if you were to try it.