We have traditionally used a Promark 2 L1 static GPS system to bring control to our sites, but have recently purchased a Epoch 35 LI/L2 VRS system and have noticed it has an option to collect Stop and Go data to be post processed. I have heard this term over the years, but have no first hand knowledge of it being used. Is this a viable option for control? ...or just to be used to add confidence to existing VRS points?. What kind of accuracies can be expected from this technique? Is it somewhere between traditional static and RTK? How should one determine the length of observations times? Thanks for any input anyone can give. I've asked some fellow surveyors and the local salesman and no one can give me much information on the subject.
Thanks
"Stop and Go" refers to stopping the movement of the roving receiver (in your case, your only receiver) on your point of interest, collecting sets of data until some threshold is met (period of time elapsed, precision of results, precision after some period, etc.), and then "Going" to the next point of interest, where you will stop again. Here's a link that might be useful: http://www.gmat.unsw.edu.au/snap/gps/gps_survey/chap5/554.htm
I would take a day or two and write them off, then fiddle experiment with this latest addition to my toolbox. Find a recently recovered monument with some good horizontal and vertical accuracies and fiddle with it. See how quickly you can hit the monument and still get a good position on it.
Set some points around the office and hit them once in a while. Put one right next to the building and see what multipath will look like on your collector.
Good luck, and happy hunting!
Stop and Go is similar to RTK.
You need at least one base receiver but two is better (it can make triangles, not just single vectors like RTK).
You start your base receivers as normal static sessions.
Then you start your rover receiver as a Stop&Go session.
You need to initialize somehow. This is done with a initializer bar (typically 20cm long bar which you attach to your base), or you can initialize on a known point (if it is in the rover), or, finally, you can do an on-the-fly initialization if your system allows it. With on-the-fly you just start collecting shots and you should run the session long enough to ensure the processor can initialize your survey (I use 15 minutes minimum).
Once you are initialized then you must keep the receiver "locked". If you walk under a tree or something like that you will break initialization and have to do it again.
The advantage is you can collect shots in a shorter period of time. S&G shots are timed, never by distance like RTK because you don't have a real time solution. I have PM3s, I tell it how long I want to observe each shot. For normal topo 15 seconds is usual.
Then download all of the receivers and process the survey in the software.
For control work I prefer Static because I want longer sessions anyway, the results generally are better and it is just simpler to do.
You Need Something To Process Against
L1/L2 stop and go requires a second L1/L2 receiver not just processing to a PM2 base.
Depending on your proximity to a CORS site you would not need to buy one.
However in close proximity L1 only vectors are quite good, sometimes even better than L1/L2 vectors. Several times in the past I purposely turned the L2 off in my post processing software and saw better results (3 PM2s and 2 Leica 1200s).
I would not just leave my PM2s on the shelf. Here is a good test for a project near a CORS. Set up your 2 PM2s on points you want to use as control and initialize your L1/L2 on a third point, occupy at least twice as long as your planned S&G occupations. Begin your S&G and in the course of events hit the PM2 points twice for an S&G. That gives 3 separate PM2 observation files on those points, but that is OK. At the end reoccupy the L1/L2 initial control point for twice the S&G time. Now post process bringing in a CORS file and setting your minimum vector time so it excudes the regular S&Gs to the PM2s but includes the double length observations, process adjust, print a report etc. Now in that same project or set up a second project set the minimum vector time to include the S&Gs and compare your results. L1/L2 S&G times can be shorter than L1 only S&G so you want to see if there is a compromise time between the 2 that gives good results processing off PM2 bases.
Alternatively you could use GNSS Solutions to create a VRS (Virtual Reference Station) within your project.
Where you are is totally different from where others are, we can give suggestions, but doing exactly as we do may not quite get your job done at your location.
Personally you may find that it is best to let your PM2s do what they do while the L1/L2 receiver does what it does, RTK VRS or S&G VRS. Processing 2 separate projects with only overlapping static observations.
Paul in PA
Stop and Go is the ability to (after you;re initialized) to go from one point to the next and collect a few seconds of data while still having cm-level accuracy (accuracy is virtually the same as RTK).
For example, if you want to shoot a bunch of manhole cover elevations. You need to initialize on the first one (~20 minutes), then you can move from MH to MH just taking a few seconds of data on each MH. Important: you have to maintain satellite lock between MHs.
Static is different. With static, you treat each MH as a completely new point, so 45-60 min or whatever on each MH.
Some else said it right, Stop and Go is similar to RTK, but post-processed.
> We have traditionally used a Promark 2 L1 static GPS system to bring control to our sites, but have recently purchased a Epoch 35 LI/L2 VRS system and have noticed it has an option to collect Stop and Go data to be post processed. I have heard this term over the years, but have no first hand knowledge of it being used. Is this a viable option for control? ...or just to be used to add confidence to existing VRS points?. What kind of accuracies can be expected from this technique? Is it somewhere between traditional static and RTK? How should one determine the length of observations times? Thanks for any input anyone can give. I've asked some fellow surveyors and the local salesman and no one can give me much information on the subject.
>
> Thanks
It shouldn't take 20 minutes to initialize.
An initializer bar takes 5 minutes in the Aschtech single frequency world but I think that's just a default; you could experiment with shorter times.
I just use the OTF because with the PM3s it works great. I've gotten initialization with as little as 3 minutes into the session. I was doing stop and go down the center of a road which goes down into a canyon. I didn't need it to be very accurate. SVs diminished as I went down into the canyon. Float-solutions would be OK for this project. There were trees down there and no good way around them. I lost lock under a tree. Then had open sky for several shots (GNSS Solutions can apply an initialization backwards). More shots processed fixed than I expected. I had lock times in the file of as little as 3 minutes. I didn't expect to OTF initialize that fast but it did.
S&G is like really inexpensive RTK topo capability but you have to post process and you don't know what you have until you post process but I never had a problem with it, worked great.
When I use my Locus receivers this way I set up two receivers, initialize on one, let the other collect data, and gather the data with the third, not the way the book instructs, but I tried it to see what would happen, post processing resulted in two vectors to each setup, I deal in vectors so strength of figure is taken into consideration.
jud
That is one of the hidden advantages of stop and go. You can get as many vectors into a shot as you want. You just need the base receivers.
> When I use my Locus receivers this way I set up two receivers, initialize on one, let the other collect data, and gather the data with the third ...
This may not be what you had in mind, but a variation on that if you are using two base receivers would be to:
- set up Base 1 and start it logging observations,
- initialize the Rover on Pt. 2 as if it were a known point (which you can do, I assume, with your post-processed solutions even if the point isn't in the datalogger's database),
- set up Base 2 on Pt. 2 and start it logging observations.
- go survey with Rover already initialized (in post processing).
As long as the post processing solves the vector from Base 1 to Base 2 before attempting to solve the Stop-and-Go vectors to the Rover, things ought to be cool with the only time spent on initialization being the time that the Rover occupation on a known point took.
Init time all depends on the number of satellites being tracked. I've done it in as little as a couple of minutes. But, since you're post-processing, you play conservative. The last thing you want is to get back to the office and have a mess because you didn't spend the extra 5 minutes initializing.
I never used the initialization bar. I suppose it speeds things up, but I've always been busy enough doing other stuff that the initialization time isn't really an issue, whether it's RTK or Stop and Go.
I agree, Stop and Go is a really cheap alternative to RTK, especially if the sky is wide open.
> It shouldn't take 20 minutes to initialize.
>
> An initializer bar takes 5 minutes in the Aschtech single frequency world but I think that's just a default; you could experiment with shorter times.
>
> I just use the OTF because with the PM3s it works great. I've gotten initialization with as little as 3 minutes into the session. I was doing stop and go down the center of a road which goes down into a canyon. I didn't need it to be very accurate. SVs diminished as I went down into the canyon. Float-solutions would be OK for this project. There were trees down there and no good way around them. I lost lock under a tree. Then had open sky for several shots (GNSS Solutions can apply an initialization backwards). More shots processed fixed than I expected. I had lock times in the file of as little as 3 minutes. I didn't expect to OTF initialize that fast but it did.
>
> S&G is like really inexpensive RTK topo capability but you have to post process and you don't know what you have until you post process but I never had a problem with it, worked great.
The problem with S&G is that you don't know what you have until you post-process...
If you're going to use S&G, learn all the tricks you can to make it work.
Keep the antenna in a vertical(like you use it), when you go between points.
Have "special" recoverable points in your field of work and hit them from time to time(with new point numbers), to see if there's agreement(you can make them all the same point number later, if you wish)
Sometime a lock-loss will result in a shift of all your points(a little better than just plain garbage) following a lock-loss, gathering the "extra" points might help you recover a job's data by just shifting back.
MAKE certain that the BASE(or BASES), above all else WILL NOT BE THE CAUSE OF LOCK-LOSS). Make certain that the BASE(s), have an extremely clear view of the sky.
AFTER all is said and done, the S&G on the PM2 and in Ashtech Solutions is very worthwhile. I suspect that the S&G might even be better in GNSS Solutions.
I might add, that when I do S&G, I set the timer(on the rover) to 1,200 seconds(20 minutes), which is more than enough to initiate.
I then set on a point for 5 or so sigma, then manually stop the static. THIS way, if I suspect that I might've had a questionable point, I can always reshoot for a full stop session.
