Is anyone utilizing drone mapping? I recently attended a demonstration at a landfill and was blown away by the technology. Via three aerial targets over a 4 acre mounded area, the results were well within what I consider designable topo. 0.05' H and V. This was against control set via network RTK and shots taken via an S6.
I wouldn't design a 0.1% flow pipe with the data, but for the majority of municipal and state construction jobs, it would be a great tool. The more interesting aspect; the technology uses photography, not LiDAR for obtaining their point cloud. But, once that point cloud is created, it is the same.
Nuts and bolts, the system uses off-the-shelf digital cameras (the one demo'd was an RGB Canon that retails for around $2k, iirc), wireless technology, and a laptop. All of the flight control is on autopilot. Using the laptop software, you define the area to be mapped, how much overlap and how detailed a grid you would like dictate the flight lines. Those are then uploaded to the flight software on the drone (a fixed wing one in this case) and off it goes. It then streams the pictures back to the laptop and gives real-time data for elevation, wind speed, yaw-roll-pitch, etc. Once complete, it circles overhead until you indicate where it should land. The drone will then fly out and return into the wind for decent and landing. Back at the office, the photo software will crunch the data and create an orthomosaic. This particular software will export a point cloud, but does not offer modeling of the point cloud. That would have to be done on a CAD software.
If someone is using this, I'd really like to hear more about your real-world experiences. Thanks.
~Raybies
It sounds like you got the gold-plated demo. I'm not using it, but the technology has already matured in Europe and Australia. There's tons of literature on the Web and some vendors are claiming 1cm accuracy while others with more rigorous standards are claiming 2-10cm. I find it difficult to believe only 3 RTK ground control points (themselves only good to +-1.5cm NSV could result in 1cm DTM accuracy) because there's no opportunity to use aerotriangulation to correct residual errors.
This whole flap came up years ago when low altitude helicopter photogrammetry was promoted. Problems? Limited coverage compared to traditional photogrammetry (not a problem on small sites), platform jitter, pitch, yaw, speed parameters not datalogged, big cost to keep the chopper in the air, aeotriangulation of the photos required a lot of ground control, although they claimed 1cm accuracy in ideal condtions. Since then the the 'copter guys have added the sensors but still charge $1000/hr. The new fly-weight airborne platforms with onboard rapid GPS positioning, attitude and speed sensors allow postflight corrections to the camera position (critical to any photogrammetric job and handled with ground control traditionally) for rectification; let us pray those sensors and the camera lens specifications are accurate. But they are cheap to run and should dominate the market for lower accuracy surveys when approved.
It's moot in the USA because the FAA has limited commercial use of drones. Once again we lose out because of our glacial bureaucracy, but mark my words in a few years small site topo with clear ground visibility will be dominated by drones, and the land surveying community should keep their ears up and capture that market.
http://www.pobonline.com/articles/96996-how-can-drones-transform-surveying
We have a mavinci Sirius pro and it is fantastic.
It has an onboard rtk so no ground control is needed and the software that comes with it does all of the processing.
Bit expensive but well worth it.
I use Phantom fc40 quadcopter 400$ (without the fc40 camera), instead I use Canon 200$ camera with built in WiFI radio so I can take pictures remotely using smartphone. It is kind of hard to overlap picture properly but still possible very well if you get used to it. My PC is kind of old one so I can not process images (AgiSoft) but it is very easy to do so, and by including ground control you get all in your coordinate system.
But as always with GPS, trees is the biggest enemy tor this, so the total station still can not be thrown away 🙂
