Darren Clemons
Well-Known Member
First of all, the addition of post-processing kinematic into the LS has been unbelievable valuable. There's not much the LS won't get great, validated RTK on, but on those it struggles PPK is there to back you up, or even better, to have another check on an already suspected good RTK shot.
My quandary though is this. Almost ALL the points I base/rover process come back saying the PPK solution is "fixed", even though sometimes two stored sessions on the same point do not match. I had two yesterday, both were 900 seconds and both came back fixed, but were about 1.0' horizontally apart. This actually may just be that I need to increase my time to 1200 seconds but I just wonder why if I wasn't on the point long enough, why didn't the solution come back float?
One of those two session I mention above, I had basically no RTK coordinate to compare. The group it stored was similar to 2,1,0 (I've lobbied and still wish we could just "reject" these types of RTK coordinates and only store the PPK data - a coordinate I know is wrong is only confusing to store).
The second session however, I had a group of something like 3,285,2 for RTK so I was confident that was good and one of the two PPK points verified that, within 0.05' - so in the end I had what I wanted, two different solutions that matched.
Again, this is a terrific feature but it does require a lot of careful sifting through the solutions to see/determine whether RTK or PPK (or both) are the correct solution. Of course, we the users are ultimately responsible for being confident that our data is correct and the LS gives us plenty of ways to do that.
I guess when all is said and done my question is; are there any "refinements" being worked on in the PPK process where, hopefully, we could get a better percentage of accurately tagged fixed/float solutions instead of almost all being labeled fixed?
What have any others who've used PPK extensively found? Is 15 minutes in heavy canopy reliable where you are? I'm probably getting 85% with what we've been processing and that's great. It's the other 15% where, as we've always talked about, you must have either a reliable RTK or a third PPK to get the two (or more) matching sets of data you need to feel the Javad 100% confidence.
My quandary though is this. Almost ALL the points I base/rover process come back saying the PPK solution is "fixed", even though sometimes two stored sessions on the same point do not match. I had two yesterday, both were 900 seconds and both came back fixed, but were about 1.0' horizontally apart. This actually may just be that I need to increase my time to 1200 seconds but I just wonder why if I wasn't on the point long enough, why didn't the solution come back float?
One of those two session I mention above, I had basically no RTK coordinate to compare. The group it stored was similar to 2,1,0 (I've lobbied and still wish we could just "reject" these types of RTK coordinates and only store the PPK data - a coordinate I know is wrong is only confusing to store).
The second session however, I had a group of something like 3,285,2 for RTK so I was confident that was good and one of the two PPK points verified that, within 0.05' - so in the end I had what I wanted, two different solutions that matched.
Again, this is a terrific feature but it does require a lot of careful sifting through the solutions to see/determine whether RTK or PPK (or both) are the correct solution. Of course, we the users are ultimately responsible for being confident that our data is correct and the LS gives us plenty of ways to do that.
I guess when all is said and done my question is; are there any "refinements" being worked on in the PPK process where, hopefully, we could get a better percentage of accurately tagged fixed/float solutions instead of almost all being labeled fixed?
What have any others who've used PPK extensively found? Is 15 minutes in heavy canopy reliable where you are? I'm probably getting 85% with what we've been processing and that's great. It's the other 15% where, as we've always talked about, you must have either a reliable RTK or a third PPK to get the two (or more) matching sets of data you need to feel the Javad 100% confidence.
