I do this in the worst places. Rarely is there more than 0.07' difference.
If you come back to the same spot 2-4 hours later are you still within 0.07'?
I'm a Javad fan.
Well you've been playing that awfully close to the vest.
I do this in the worst places. Rarely is there more than 0.07' difference.
If you come back to the same spot 2-4 hours later are you still within 0.07'?
I'm not in the jungle like Nate, but yes, 2-4 hours, 2-4 months, the new engines do work. I've had the R-10's long enough to test them in deep canopy with months and years between locations to the same monument and .07' is easily achievable, even in Ponderosa canopy. My biggest issue is time of day, sometimes it's best early, sometimes it's best later in the day.
Just yesterday I completed a traverse in a local suburban neighborhood. The process was to hit all the points with GPS, then traverse through them all, and simultaneously adjust the vectors and traverse data using StarNet.
The "entered provisionals" are the coordinates as determined by GNSS alone, using a Leica GS18 base/rover pair (2 x 90 second RTK observations on each point) . I've set the GPS vector standard error scaling higher than need be, to more heavily weight the traverse data. The neighborhood is not deep woods, but there is plenty of obscured sky.
Just yesterday I completed a traverse in a local suburban neighborhood. The process was to hit all the points with GPS, then traverse through them all, and simultaneously adjust the vectors and traverse data using StarNet.
The "entered provisionals" are the coordinates as determined by GNSS alone, using a Leica GS18 base/rover pair (2 x 90 second RTK observations on each point) . I've set the GPS vector standard error scaling higher than need be, to more heavily weight the traverse data. The neighborhood is not deep woods, but there is plenty of obscured sky.
That looks deep to me.
If you come back to the same spot 2-4 hours later are you still within 0.07'?
Over and over. The time needed to get that varies. There are several metrics to CHECK things.
Nate
@gary_g Rare footage of me when Carl bought a metric level rod for some Corps of Engineers projects.
We are using R12i receivers and they are surprisingly good in a variety of bad locations. We do redundant shots separated by hours. Last week I did a long static session too. Beats the heck out of traversing, after all those setups in rough mountains it is probably worse.
Yes in an urban environment with the possibility of longer sights TS is probably better. But in slippery slidey mountains with terrain induced short sites and brush RTK is looking much better at this point.
(2 x 90 second RTK observations on each point) .
Curious how you arrived at that observation length?
Using Leica/Geomax gear my testing results (RTK base/rover, base <2km away) were 20-30 seconds was quite an improvement over single shot but had to be more like 5 mins to get any meaningful (but quite small; diminishing returns) improvement. I found what made a bigger difference was observe 20-30 seconds and spin pole 180° and shoot again which I consider the 'first fix' then repeat at least 30 mins later (I'm assuming. For control points like this have the 2m pole braced with spade/brace stick at 90° to each other on pole (as stable as using a bipod).
I also find in beside a tree/power pole sometimes my RTK GNSS double ties agree well but the total station ties can be 20mm (95% CI, not the 1 standard deviation rubbish on GNSS receiver spec sheets) off which is too much for urban boundary control here.
Huge Javad fan. Even in the deepest darkest hollow, I've been able to get shots. My total station work is very tight but I feel better with points located with my Javad unit on a large job with steep terrain better than running a traverse. Most important, there are enough details on the screen to give you that warm fuzzy feeling or not about each shot. Immediate post processing is a great trick. Urban canyons are not good GNSS country because there is literally almost nothing but multipath in between the tall buildings and I try to avoid the big cities anyway. Somehow the GNSS signals filter through the leaves and branches enough and there are enough signals with 4 constellations that move the geometry quickly enough to get reliable shots. I love the built in data collector/antenna in one. Makes for getting around the woods much easier.
I have pushed the R12 base and rover and done a lot of independent checks. I usually change the default confidence level from DRMS to 95% it gives a more realistic idea of what’s going on in canopy or other hostile environments. What I found was it was pretty reliable if I had 20 plus satellites in canopy. Once It dropped below 20 it proved not to be as reliable. Again this was with independent checks from traverse or other means. Redundancy redundancy redundancy in canopy is what I suggest. I will also say that every time i was able to collect all 180 epochs and not loose a solution and held more than 20 i would have great results. Any time it lost lock or fixed solution if we want to call it that it proved that it was not as good. Also had a brx7 carlson do tremendous in canopy but i have not put it through the ringer yet. Javad is about the best in my opinion from what i have seen. If set up correctly like nate said you can bank on it. We are a Trimble shop so i do multiple observations from different base set ups and time gaps for anything in canopy that matters. Here if it doesn’t meet the alta requirements i re observe or observe in a different way like robot etc. never do I only observe once. Always a minimum of two usually 3 in canopy.
Regardless of equipment colour, my personal opinion is to just have some sort of additional check to quantify that the shot is acceptable to the necessary job requirements. Additional possible checks could be:
-returning at a later time to reshoot
-logging RINEX so that one can calc a check via PPK or static
-traverse/hybrid resect to re-tie the monument
-mix and match constellations (ie. tie with GPS+GLONASS only and then GALILEO+BEIDOU only)
-pop in 2 nails, shoot all 3, tape between, and then run in a LSA software. With the announcement of SurvPC 7 being able to do LSA in the field, on-the-fly one may come up with new checks to harness that quickly and easily.
Checking against a known distance on a plan, to me, is a check but it should not be your only check.
Carlson's SureFix seems to be, to this technological Neanderthal, a "lite" version of the Javad Hybrid RTK. I have never test driven a Javad unit, but they do seem to have a very robust checking system from the literature.
Ublox's service of Point Perfect shows promise for cheap receivers but it's not quite there yet IMO:
https://www.u-blox.com/en/product/pointperfect
I imagine all the manufacturers will be coming out with something similar (ie. PPP&RTK) if they have not already.
https://www.mdpi.com/2072-4292/14/20/5100
Link is for a study showing ublox vs. Javad.......just to, maybe, stir the pot on a long weekend 😉
The above statement is so true. Check check check. I have not experimented yet but it has some good theories to it the time of day is a big factor in canopy. I learned last year between about 1400 to 1500 here was not the best in canopy. Now at that same time I could get in the open and lock 30 plus satellites. Teach your field crews how to understand the satellite azimuth and elevation to have good geometry. I am sitting on a point for 180 epochs or 3 minutes i am looking at the numbers and signal to noise ratios and mentally drawing a picture of those weaker sats and where they are compared to my position. You cannot just have button pushers or at some point you will get burned. Doing a little pre planning like we did in the old days and start to accumulate that data with the redundancy of points in canopy should begin to give a good indicator of when you should and should not be observing in canopy. This separates the it said it was good from having confidence of it is good. Also if you can live with a tenth or 2 tenths or need better than that on a project. I am i. Office now days and i can look at the time of day and rms and just about know if i need to send someone back out or not before i do anything further. Now getting the crews to go back and even giving them the time of day in which you want them to re-observe a certain point is like pulling teeth lol. They tend to repeat in the same order lol. So about the same time. Which is not what i want. I am trying to implement some standards that would solve this but as we all know its just hurry up and get it done. Planning is a bad word sometimes.
I have never test driven a Javad unit, but they do seem to have a very robust checking system from the literature.
The literature does not do them justice.
Javad had a number of receivers, set in crazy places, and feeding data via an internet or other connection, that was testing thousands and thousands of shots.
I've been using them for years.
It automatically checks itself.
There was a time when Mr Javad Ashjaee was alive, and they offered 10,000 $ reward if you could get an LS (their flagship) to lie to you, by some amount. (I think it was 0.16' or more).
Some of the users talked him out of it, because there are methods, using tin foil, and frequency generators, and the like to fool it.
When you get a shot, in a very poor place, it gives you a number of warnings, such as an image that looks like a shotgun shot makes. With a "large shot scatter".
Get a look someday.
Spending about 4-6 weeks is about what it takes to really experience some of its ability.
N
