UAV mapping systems that surveyors like too

My father, who is a county land surveyor, pointed this article out to me. Looks like a neat system.
"asked our readers to select the one they believed would have the most impact on the surveying and mapping professions.The results are in. Out of a total of 697 votes, the Gatewing X100 Unmanned Airborne Vehicle is the winner with 48 percent of the votes"http://www.pobonline.com/Articles/Features/BNP_GUID_9-5-2006_A_10000000000000950297

• The camera has a 200 m swath width (150 m altitude) and has 80 m of leg spacing (60 % overlap). If you would chose a square area of 2 km2 (which is sub-optimal) you end up with 18 lines of about 2 km (including turn and line-up). At 75 km/h - which is the cruising speed - this is about 30 minutes (including a few minutes for take-off and landing). Using our new 9000 mAh battery, the vehicles endurance is currenlty +45 min so it can cover a bigger area and fly for +56 km if needed.

The total number of images of a 2 km2 square area would be 710. Stitching this requires a few hours processing time but this runs automatic.
• T3
There is another issue. a 2km square area taken at 150m with those overlap percentages would result in more than 2000 images. Good luck stitching that. Somebody needs to revisit there marketing material.
• T3
An excellent system.
Unfortunately, if the datasheet area coverage is not overbloated, it flies with 110km/h cruise speed which I feel is totally unrealistic. Here's how:
Datasheet:
Overlap 75% (forward), 60% (side)
Average coverage per flight (30 min at 150 m) 2km2

The most time-efficient way to fly 2km^2 is to fly over square area, with side of the box equal to sqrt(2), i.e. 1.41km. Assuming the camera placed long side along the wings, and normal lens with 60deg viewing angle,
150m altitude, the width on the ground will be 150m*sin(60deg)=130m. With 60% side overlap, we have 40%*130m=52m leg spacing. 1,41km of flying is 27 legs PLUS turns PLUS 2 legs to takeoff and move back.
But because the system has no stabilised camera, if we assume tiny turn diameter of 100m, after each leg it will have to fly AT LEAST 270deg of a circle of radius 52m plus 52m straight (during the turn the photos taken at some 40deg roll are not useful fo fotogrammetry so it must be flown outside the pattern). In short, each turn is (3/2*PI+1) leg spacings. There are NLEGS*2 turns: 27*2*5.71=308 leg spacings will be flown in those turns. This is 16km. 29 legs, 1,41km each is 40km. So total flight length is 56km best case in order to cover 2 km square.
With 30min flight time, this means that average cruise speed of this plane is 56*2=112km/h.
All this uses lipo 3S battery (11.1V) and 6Ah lipo during 30min, what is 12Amp avg current, or 133W average INPUT power to the motor...
• Hi, I am Peter Cosyn from Gatewing. I want to comment on a few questions raised about our X100 system.

The camera is protected during landing. The lens is retracted shortly after the acquisition job. What you see on the bottom of the aircraft is a screw-on filter.
The landings are not so rough as you would think. The vehicle lands at 40 km/h max (typically in no wind conditions). The movie shows a landing on a rough barren field and this includes some bumps and mud.

The X100 is indeed compatible with the software from MosaicMill (and similar automatic triangulation packages). The system includes custom software to register the (calibrated) images for processing with stand-alone software or an online on-demand service (third party). It makes it a tool for accurate terrain mapping and surveying.

The X100 is designed for a professional user with no RC experience. Safety, ease-of-use and reliability were as important for the design than the technical qualities needed for an accurate terrain mapping or surveying job.
• Have a look at the datasheet
• Woah loving the software.

Does MosiacMill do the DTM? The automatic scaling and placing into GoogleEarth is nice :).

< longshot >Sooo.. what open source alternatives are there? :)
• T3
Gatewing has taken the same route pioneered by Scot McTavish more than two years earlier at http://www.accuas.ca/

Not as ground breaking as that magazine would have you believe just sexier.
• That looks interesting! Thanks for the tip.
• T3
They are using a MosaicMill and parsing the log file from the MP and using that to stitch and rectify the images.

They had another video and if I recall correctly the ball park figure mentioned there was 40K Euro for a system. I think there has been some creative license with regards to setup. The airframe looks like it rolls quite a bit I assume that this is not an issue as they correct image distortion in software.

If you need flight planning software for a MP system check out http://www.lentsika.ceped.it/

Great software have been using it for more than a year.
Lentsika by CePED
• Cool looking system, appears to be well packaged with fairly easy setup. It appears to use a MicroPilot with some custom mission planning software to deal with the fun of easily creating flight plans in Horizon. Powerful post-processing options as well, though not in real time.