RTK vs PPK vs Fast Static accuracies

Posted by: trw

We are presently looking at a drone-based photogrammetry project with a totally unreasonable topo survey support requirement (thousands of points with a 0.01 ft vertical accuracy spec).?ÿ

Break out the LEVEL!

Ain't going to happen with GNSS of any flavor.

Loyal

It maybe can be done but I doubt they would actually pay you to do it for the cost of the survey. Like Loyal says, break out the level, and not just any level a really, really good one, forget any GNSS for this one.

?ÿ

I think we all know that the spec is pulled out of someone's a$$, or maybe you are doing something really hi tech. Elon Musk's secret quantum power plant??ÿ

Posted by: trw

We are presently looking at a drone-based photogrammetry project with a totally unreasonable topo survey support requirement (thousands of points with a 0.01 ft vertical accuracy spec).?ÿ

Send them a realistic proposal and a copy of this?ÿ ??ÿ

Putting aside the idea of .01'?ÿaccuracy,?ÿmy preference for doing topo shots if I want more accurate ones would be long PPK sessions. By that I don't mean long sessions on each point, I would probably take 15s shots, but long fixed PPK sessions, they seem to be more accurate for each shot than RTK. I do not see how fast static is an improvement, it would take way too long per shot to the point you would need a total station. The MOST accurate topo with some reasonable?ÿaccuracy for the time spent is using a robot. ?ÿ

Sounds like your client is asking for you to verify the drone data for an accuracy statement on the plat, as per 2014 National Standard for Spatial Data Accuracy. Which states the measurement procedure and equipment for the test must be at least 3 times more accurate than the best possible accuracy statement for the data. This will basically imply that 95% of the spatial data meets the accuracy statement.

So if your procedure is ?ñ0.01', then the best possible accuracy statement that can be applied to the spatial data will be ?ñ0.03'

Your client is quite confident and overzealous in their work, and they do NOT want to be limited by your procedure.

This ?ñ0.01' accuracy would require you to measure at something like First Order - Class I procedures, which is not RTK, PPK, or Fast Static.

?ÿ

There are a few missing pieces to this puzzle...

Is this dirt topo or hard surface? Are they specific points? Is the client an eccentric rich person or just crazy?

You can work with the first optionn.

There are topos that need higher accuracy than the standard dirt topo. Airport concrete reconstruction, inside buildings; urban street topos typically the tie in points need extra care, no doubt there are many others. But these should be done conventionally not with GPS.

Thanks for the feedback, largely confirming what we already believed.?ÿ I should mention that I?ÿam primarily a hydrographic surveyor, though I have?ÿconducted many?ÿintertidal and upland land surveys over the years.?ÿ We currently own several?ÿR10s?ÿused mainly to support?ÿhydrographic projects, though we can take the gear upland as required (mostly for terrain mapping applications).?ÿ (We also run a lot of levels, primarily for tidal benchmark support.)?ÿ We generally rely on short-baseline RTK using our own local base with a cell modem data link via?ÿNTRIP caster. Though we?ÿuse Applanix POSPac for post-processing our vessel-based GNSS/inertial navigation and orientation solution, we do not use TBC (though are now considering purchase).?ÿ We?ÿhave recently supported a few?ÿsmall near shore drone-based photogrammetry projects, primarily positioning ground targets and providing spot check topo.?ÿ Since the drone-based photogrammetry provides vertical accuracies of around 6-8cm, our 2-3cm RTK vertical accuracies seemed reasonable.?ÿ We have pushed back hard against the .01ft vertical accuracy requirement, and have told the drone operator (our potential client) that there is no way for us to meet the spec?ÿ(for thousands of points).?ÿ?ÿThe survey area is is all beachfront, salt marsh, soft ground, or?ÿdense shrub.?ÿ The prime remediation contractor?ÿclaims that this is the spec that has?ÿbeen used for years.?ÿ We sometimes see this same vertical accuracy?ÿ(0.01ft) specified for a hydrographic survey, which?ÿis?ÿeven more laughable than the topo requirement.?ÿ We usually trace that requirement back to a misguided engineer (a category I might fit?ÿunder).?ÿ Just to circle back to one of my original questions: does?ÿPPK (within TBC) offer any potential accuracy improvements over?ÿRTK (or is it primarily used to fill-in gaps due to RTK data link issues)??ÿ Thanks again.?ÿ ?ÿ ?ÿ?ÿ?ÿ?ÿ?ÿ

I would argue that PPK is slightly more accurate in topo mode than RTK. However, the drawbacks using it?ÿoverwhelm the little gain in accuracy. I use it for two reasons, losing base connection and as?ÿstatic check on points.

As we all know the .01' accuracy is silly, what they really want is the data carried out to .01'. That seems to make everyone happy. You can try what I've done and put in your proposal the equipment you are going to use and the spec shown for each unit. Then declare that you will use this equipment as per the manufactures recommendations.

I agree that often a spec like that was created by an engineer who wants data to the 0.01 feet in the coordinates/elevation. I have argued a few times with specs that wanted manhole elevations/coordinates to an?ÿaccuracy?ÿof 0.01 feet, which is pretty much impossible over any area larger than a block or two.?ÿ