Demystifying Mobile Mapping

Cool article.

"Can the Pegasus:Two do it all? No, it cannot reach out and open that storm drain cover that has been rusted shut for who knows how long. It won‰Ûªt measure the invert for you either..."

Maybe someday?

Good article. I read it recently in the XY...publication. It looks like it is only for the big BiG dogs to implement and market to governments etc. Price tag off the planet.
Well done, I know that you have scribed an article or two in the past.

What is interesting are the control requirements. If the client specs 0.05 feet vertical accuracy for the final product, rule-of-thumb has been that control should be 2 to 3 times as accurate (0.015 to 0.025 ft). This can only be done by leveling, and not just ordinary leveling but perhaps second order leveling.

Yet I hear of people setting control with RTK. I recently worked on a project where the end client did the control, and their levels had serious problems, I estimate the leveling over the entire project is å±0.20 feet. That was after I adjusted the levels using least squares, they had done simple loop adjustments, not taking into account existing benchmarks or cross ties. But, this would not be seen in the processing of the mobile data because the error between adjacent control targets (500 feet apart) is small. I determined that the error in their levels was systematic and depended on elevation. It was a long linear project, about 10 miles, and the profile went up and down several times along the way. In the plot below, the difference between the levels and VRS (dual occupations, separated by several hours) is the primary vertical axis (blue dots), and the elevation of the targets (orange squares) is on the secondary axis. The horizontal axis is northings, since the highway is aligned N-S.

So, my point is that the new technologies that give high resolution data (mobile lidar, imagery from drones) need much more attention to control.

Sometimes the results reveal things that might be overlooked by legacy means. Brad relates a favorite example: ‰ÛÏOne of the first surveys we checked had been both surveyed and static-scanned. In our data there was one control point that was outside of what we would consider normal for the Pegasus:Two.‰Û

‰ÛÏThe client held that up to us as an example of a problem with the mobile mapping system. In fact, after further investigation it was discovered that there was a rod height bust. They had not caught this during two additional surveys, but we caught it with a single pass of a mobile-mapping system.‰Û

Someone has poor field procedures if they didn't catch a rod height bust after 3 surveys. I almost always catch it on the next setup when I check my backsight deltas. At that point I review the raw data and usually the problem is I forgot to change the target height from the BS to the FS TH on the last setup. I make a note on my paper field notes to fix it later in the raw data.

Gavin: As you know with drones it is now possible to get sub-centimeter resolution in aerial imagery. In the recent past 7.5 cm (3") was high resolution. So, we need to be attuned to the accuracy requirements of these enhanced imagery products.

In the project I mentioned above they used a digital level (with a standard fiberglass rod), and ran from the south end to the north on one side of the highway and back down to the POB on the other side of the highway, 20+ miles total. No cross ties. I asked for the digital level file, but it "doesn't exist" because no one knew how to download the level. No ties to existing NSRS marks. Then i got an excel spreadsheet that had all of the turns in it, and "adjusted" elevations. I inquired about this, they had recorded all of the data in field books (which I never saw), and then typed them in to excel. A few months later at my request they tied in 2 NSRS benchmarks at the north end and 1 at the south end. Then I found out they had long static occupations (with fixed height tripods) on points they set every mile, which were included in the levels. But they said that "wasn't any good, why would I want that data". So you can see the level of knowledge they had. The control for this project should have been done by us, but they (government agency) wanted to keep their surveyors busy.

gschrock, post: 378747, member: 556 wrote: Responsible outfits can and do employ surveyors to establish appropriate control

....

gschrock, post: 378747, member: 556 wrote: Let's get proactive in making sure there is more of the former

....

Scope creep; make sure this doesn't happen to you.

• We just need an approximate boundary, for preliminary purposes.
• We're just going to run the sewer line down the center-line, so we don't need to know where the right-of-way line is, just show it as approximate.
• We're just doing a preliminary study; we don't need that kind of accuracy.
• Etc...

A well defined scope and end product notation will go a long way to providing the right service. The surveyor is the best person for the task.