Bandwagon's Drones Thread

Good call on increasing the overlap. Sounds like you've been getting a lot more experience? FYI, I've seen a significant improvement in results (less noise) with the Phantom4p after slowing it down further. I'm typically flying at ~19mph at 400ft and 14mph at 200feet. I also changed my processing template to disable "surface smoothing", and the results have been much better, especially on asphalt surfaces. Instead of trying to smooth the noise around homogeneous pixels, it doesn't even build those points anymore. Much easier to work with.

I don't understand what you mean about the elevations on your GCPs, though? You should be using the text file & XYZ coordinates straight out of the data collector when you're processing, not adjusting them with an offset of flight altitude?

Here's the point cloud for a recent one flown at max altitude, but much slower than the default speeds. Pix4D Cloud

Take a look at all of the gaps on the road and in the parking lot. This makes surface building MUCH easier with better results than trying to deal with the noisy points & surface smoothing in P4d.

I really haven't had to work in a local system much because the two old surveyors I work with almost always translate into SP since I started. They make it seem pretty effortless and I think I'm spoiled.

I'm working on two landfill sites right now where the client's own surveyors provided me with the stateplane and local coordinates. I'm seeing errors of ~1.5ft in half of the targets, which I've never had before. Kind of a pain in the ass. I knew I should have pushed harder to have them use our own surveyors.

I'm working on my LSIT too, btw. I love this stuff, not sure why.

I'm not sure if I'd like it more or less working in your environment.

Vegetation is a pain in the ass in a lot of areas. The sparse sagebrush in east Washington & Oregon isn't TOO bad, but anything heavier than that and I'm just making guesses.

Smaller areas, lower altitude and denser point clouds in an urban environment sound easier to work with, but then I wonder if terrestrial lidar would end up being the better option?

For the first time since I started, I'm fully booked for the next 6-8 weeks. I'm ready to hire some help, but not sure what to look for. I think an experienced lidar tech/draftsmen would help me out the most in the areas that I need help, but a solid GIS-P would help out a ton with how I want to make all this "drone data" available to everyone at my company.
It's weird being in an emerging market, especially because most of it has ties to longstanding professions. There's a lot of people out there with relevant experience I'd like to tap in to, but hardly any with experience that's directly related. They've worked with lidar, but no experience collecting it or vice versa. Experience with enterprise GIS, but not at the size or pace drones produce. Experience with digital only surface building, but not at the resolution drones collect or with survey experience, etc.

I guess I'd like to hire some one that I can learn from as well, but I also want them to have some experience doing what I do. Honestly, I think I'd like to hire someone to be my supervisor, and a photogrammetrist would probably be the best bet. Too bad they all seem to be 5 years from retirement.

I'm in a desert working on a solar project right now.

Leica makes terrestrial scanners that are used in surveying. Not sure about that model.

If you're making small assets, get a 3d scanner or make one with a 3d printer, then use your iphone.

For buildings, get a DSLR and sone software.

Vepil L Lenardo - Thought you guys would like these ones.

First 10 schools for Portland Public Schools. Flown in 1 day, processed files handed off to draftsman <24 hours later.
* The boundary polygons for the completed schools contain links to external files, like the ortho, virtual tour and point cloud. Click on a completed school site to see the links in a pop up. Also has some ground photos in it. The grids with numbers in them (200, 400, 0, etc) is the new FAA LAANC facility maps.
I sent all of the ortho DWGs over to the draftsman for 2D work, and he's using the linked point clouds to measure building corner heights. Working pretty slick so far.
Portland Public Schools web map


Working on a landfill surface. Need to pick up a couple more roads, but mostly done.

Vepil said:

Very cool stuff, we haven't flown our drone in over a year thanks to the retarded owners son.

Click to expand...

Take it home and don't tell him.
Leave a little pile of Meth in the drone case so he gets distracted if he randomly goes looking for it.

That's a bummer, though. Sorry that's still an issue.

Army builds robot attack tanks and ground war drones Tuco

Lenardo said:

hey bandwagon you see that the FAA is testing a new automatic procedure for getting permission to fly a drone in restricted airspace (mostly for near airports)now? your area (washington/oregon right?) should be live now. supposedly anything under 400' is an almost instant approval.

FAA UAS Data Exchange

Click to expand...

Yea, LAANC just launched in my area yesterday. Should help out for the school project I'm doing in the Portland area. 70 school sites left and a little over half are in controlled airspace. The international airport is easy to work with and they're ready for LAANC. The other airport (Troutdale) is not, and I still have to use the old waiver process (100+ day wait) to fly in their airspace.


I've been doing a lot more drafting myself, which has been going much smoother. I think it's easier for me to learn the basic drafting standards than try to show the draftsmen how to use my software and extract their own vector data as needed. It's also made me focus more on trying to improve detail around hard grade breaks like curbs. I've been slowing the flights down to 8mph-12mph and doing perpendicular lines at 185ft and 215ft with a 75d gimbal angle. Point clouds are looking fantastic....considerably improved in the last month or two.

Here's a funny one from today. Land dispute due to an old legal description that references a road that no longer exists, supposedly running NW-SE through this field. I was surprised to see something that matched the general area in the DSM, so I made the surveyor a couple drawings with the DSM as a basemap and including 0.25' contours rasterized on top. He hasn't spent much time on it yet, but it sounds like they'll be able to use this in the dispute.