Hello,
Just a quick note on a subject previously discussed.
We have been using a survey style called RTK & logging for road (truck) surveying. Short baselines; less than 2 km's from base. Weekly surveys. 1 second interval data collection.
We are finding that the post-process results (processing with L1 only option) give consistently better results than RTK.
For example, where an observation reports a vertical precision around 40 millimeters (at 68% confidence) in real time, the post-process data gives us 11 millimeters (at 95% confidence).
Our procedure is to initialize for 10 minutes before the start of line, we run the line and we record a second static observation for 10 minutes at the end of the survey.
This is where I am not too sure about the procedure: Does the software (TBC) use the second static observation in the post-processing? Does this second observation strengthen the data? I read from a manufacturer's manual that the post-processing can also be in backward direction, what is meant by that?
Anyway, I just thought I'd share this information, it appears to be a reliable method, this RTK & logging data collection survey style.
Regards,
Georges
Georges-RTK & Logging
One second intervals may be a bit much. If you use 5 seconds it works well for road surveys and the files don't get so big. But one second data will work. Just be careful if you are collecting a point in PPK to give it a little longer than three seconds, which is what it will tell you is enough time. That kind of point can fail. But in continuous topo, each location is a point in time matching that second in time with the base as the rover passes over the point and it works well from what I've experienced. I have always seen that if PPK or Infill data is collected that it is important to let the rover "see" the sky as long as possible. Stay "fixed" in other words. Before and after the data you collect, the longer the better, so it sounds like what you are doing is just the right thing. Although, I don't think the static session is improving the PPK data. Are you still in the same fixed session? Normally you need to start the survey session over to start collecting static info.
I can also report that you can really extend a PPK session and still get really good results. I have collected roads 50-60 miles from the base (they processed just fine) and later checked the data and was really surprised at the accuracy. Something you just can't do with RTK.
Really though, I wouldn't change anything; sounds like you have it working and you are doing the most important thing which is taking the extra time to make it work. I have not always done that and have paid the price of doing something twice.
Georges-RTK & Logging
Mighty Moe
You must be working in some wide open country to be able to maintain lock to a minimum of 4 SV during a PPK session and go 50-60 miles.
Georges-RTK & Logging
Yes, wide, wide open. The county would rent a unit from us. The county engineer would mount it on a Dodge Durango and drive the County Roads collecting in PPK-picking up intersections, low points, culverts. I would then process it using our base that was also collecting. Worked great and some of the roads were over 50 miles out. Years later while locating the same roads and comparing the results I was just seeing a few tenths in elevation and the horizontals looked good but there was no way to compare them very accurately.
He would have to stop and refix if he went under something but that was rare.
Statistics are one thing, but have you ever verified your vertical results with a level? How much does the RTK and post processed data change vertically when compared to one another?