More kudos for Javad

#1
A bit blown away yesterday at the performance of the Javad system.

I was using the RTK setup w/ the LS, the T2 base and a 1W radio--and facing locating a wetland edge in dense woods. I'd gone out the day before and flagged the wetlands, but up here in Massachusetts we're close to full leaf-out and I was expecting a slow go of locating all the points. Trees are densely packed and 60-80 feet tall.

I was fortunate in being able to set the base and radio in a large open meadow beside the woods--about as good as it gets. But I was still leery of what would happen when I slogged into the oaks and maples and began to work..

Instead of long waits to collect points, the LS simply rocked. I collected 134 points, each separated by an average of about 35 feet, in 2.5 hours. That's 54 points an hour, which included walking from point to point. To put this in context, my firm has done this sort of location for a couple decades using total stations. On a good day's field work, a crew would survey 40-50 points, on a spectacular day perhaps 60+. The job I completed yesterday was normally a 2-3 day event. So I saved a couple days work, and got far more precise locations.

As an aside, the farthest distances from the base/radio were about 500 feet, so I wasn't pushing the radio's range. Perhaps performance would have dropped if I were farther out. Also, even in the densest woods the LS was pulling in 10-11 satellites. All in all, noteworthy!

:p
 
#5
I neglected to note I'm using a routine with 2 engines, 40 epochs, for fast wetland location. In reality, these points can be accurate to plus or minus 18-24 inches, so that speeds up location (and I don't care about vert accuracy). It becomes a different animal of course if it's location at high precision. But for my purpose, the Javad is a miracle worker.
 

Matt Johnson

Well-Known Member
5PLS
#6
I neglected to note I'm using a routine with 2 engines, 40 epochs, for fast wetland location. In reality, these points can be accurate to plus or minus 18-24 inches, so that speeds up location (and I don't care about vert accuracy). It becomes a different animal of course if it's location at high precision. But for my purpose, the Javad is a miracle worker.
You should be using RTK Verification and Validation with with a Minimum Duration of 120 seconds if you are in areas under tree canopy or multipath. Are you doing this? If not there is a chance that you save a point with a bad fix and the location may be incorrect by many feet.
 
#7
Min duration, 180 seconds.
120 is nearly gold, but once in a while, one slips through. You can set it to reset rtk with U-1, leave it at 180, and then hit U-1 at 120", and you are okish... In light canopy...mmm hmm but if you want 100% it's 180".
 
#8
We took a CAD plot of last week's work out on a site walk this morning and walked the entire delineation. All points appear good; no outliers. I definitely was not clocking at 120 seconds, let alone 180. I was collecting based on epochs, with 40 as a min but 50+ being typical. Let me know if this is not recommended. I used that criteria because, until acquiring the LS, we were using a Leica (2016 model) that stopped after 10 epochs. (Note again that what I'm describing is solely for wetland delineation and not high precision boundary stuff.)
 

Shawn Billings

Shawn Billings
5PLS
#9
With a bad fix you may be off by several feet. 40 epochs at 5Hz is only about 8 seconds. If you have some verification on the front end and validation on the back end, then you get some added comfort that the fix is correct, but from our experience, multiple fixes under canopy in a short amount of time are highly correlated, meaning the error can be reproduced repeatedly. Only after passage of time do the fixes become less correlated showing a likely good solution when the repeat fixes agree. This is why the boundary profile is 180 seconds, not necessarily to improve precision (which is a bi-product of longer observations times for our purposes). The main reason for 180 seconds is to allow the multipath conditions to change enough for a repeat fix to have some significance in proving a good fix. Matt's recommendation for 120 seconds isn't bad and I have used a similar time for creek meanders.

Also, pay close attention to the Distance to Last white box. If you are locating wetlands delineation the elevations from point to point should be pretty close generally from shot to shot. If you see a big variation in vertical difference it's a clear sign that either the fix you currently have is wrong or the point you previously located is wrong.
 
#10
I've become pretty dependant on the Distance to Last white box and have been using it as you recommend. It's not failsafe to follow the elevations though as many woodland wetlands follow groundwater seeps, and may creep uphill and be different vertically by four-five feet. But generally that's good technique. I'll immediately switch to a 120 second collection for this future wetland locations. (Sounds like I got lucky this round.) And 2-3 minutes per point is far faster than total station surveying.
 

Shawn Billings

Shawn Billings
5PLS
#11
For your accuracy needs, if multiple points tend to be clustered fairly close together, you might consider using the 2-3 minute observation as described above and combine it with radial compass and tape measurements. If possible, you can easily use a good compass and get 20 arc-minute accuracy (if there is no local attraction), which would be sub-foot to for distances of less than 173'. I'd expect the practical range of such an approach would be about 50' or so. This way, the LS can crank out a solid, verified position and the compass and tape (or laser range finder) can be used to capture clusters of points nearby.
 
#12
Not something I'd considered. Ingenious. Thanks for the usual (& superb) recommendations. Wetland points tend to average a horizontal distance apart of 30-35 feet, but not always. Sudden turns or kinks may be far closer. Good stuff. :)
 
#14
For your accuracy needs, if multiple points tend to be clustered fairly close together, you might consider using the 2-3 minute observation as described above and combine it with radial compass and tape measurements. If possible, you can easily use a good compass and get 20 arc-minute accuracy (if there is no local attraction), which would be sub-foot to for distances of less than 173'. I'd expect the practical range of such an approach would be about 50' or so. This way, the LS can crank out a solid, verified position and the compass and tape (or laser range finder) can be used to capture clusters of points nearby.
As I think about this I have a follow up question(s). Can I rotate the LS while it is gathering data? I have visions of setting on a point as long as it takes, and using the compass fot sideshots while the LS is "cooking". If this doesn't work I would like to shoot offsets after the point is accepted, and hang the bearing and distance on the good point. Does any of this work, or will it be necessary to carry a compass and tape to the site, write the data (bearing and distance) down and then process the offset points using something other than the JAVAD system?
 

Adam

Well-Known Member
5PLS
#15
I have used the internal compass on the LS to run short creek traverses. I will start on a solid point and end on a solid point. Nate is absolutely correct. Calibrate prior to doing it.

Steve it works fine for creek locations or wetland flags if calibrated and traverse legs are few. I usually collect one point then do the offsets. You can do multiple offsets from the same anchor point. I don't think you can do them while the anchor point is collecting in the background though.
 
#16
Just a currosity question. Can I rotate the LS while collecting data? I can envision doing this to use the compass to take notes while waiting for the position to fix.
 
#18
I use an 8' twist-lok SECO rod most of the time and just loosen the brass ferrule. I change its orientation without issue all of the time. You can stake a point while it's in collect mode. It may suspend epochs while doing so but I've never had a problem with that--it just continues where it left off. Not to hijack a thread but I have to constantly remind myself how multi-task-friendly the LS is--it's not like the "old" data collectors I'm used to from the past.
 

Darren Clemons

Well-Known Member
#19
Just a currosity question. Can I rotate the LS while collecting data? I can envision doing this to use the compass to take notes while waiting for the position to fix.
Steve, I actually rotate the LS intentionally many times when in tough spots. It “seems” to help sometimes with acquiring another “pop or two” of epochs - slightly different alignment of the signal getting through the trees. Sometimes, when up near a large tree a perpendicular LS gets the shot quicker, sometimes parallel gets it quicker, just depends on time of day and orientation to the 20-30 degree “dead zone” of sats to the north really.
When plumbed up and on a bipod, and as David says, just losen the ferrule and slowly twist the entire pole. If you need to, just pause the session, twist it to desired location and resume....