Replies to This Discussion

Permalink Reply by Gary Mortimer on August 29, 2009 at 4:49am

Yeah Jorn, I think that will come, right now its enough to get around a course!

Hopefully this competition will drive Ardupilot GCS development to allow simple flight planning and payload operation.

Methods of navigation will come under scrutiny as well.

• ▶ Reply

Permalink Reply by Chris Anderson on August 29, 2009 at 7:29am

Here are some ideas I've shared with Gary:

--Trigger event (take picture at certain lat/lon; extra points for the camera being pointed in the right direction!)
--Geometry class. Most perfect circles, figure eights and triangles.
--Total autonomy! Under autopilot from launch to landing (describe a simple pattern in between).
--Certain airframe types: flying wings only, or high-wing trainers only.
--Heli/quad/tri round?
--Aerial mapping (lawnmower pattern and stitch--largest contiguous region wins)

Things I'm not inclined to do:

--Max payload (too dangerous)
--Max duration (too boring)
--Anything beyond VLOS

• ▶ Reply

Permalink Reply by Krzysztof Bosak on December 16, 2009 at 8:53am

'--Aerial mapping (lawnmower pattern and stitch--largest contiguous region wins)"
-this is exactly Max payload+Max duration as soon as you have working UAV. Definitely beyond vlos as you can do 1x1km during less than one h with easystar - and it will be very overloaded. No guarantee there will be continuous photo at all, but there will be long-range losses.
-we have problems with participation so I discourage from dividing into subgroups.
-I think the overall difficulty is unplanned and quite random: multiclimb is quite easy, precision butterfly was very difficult, now again photo stitching is so much more difficult than just single shot mission. Therefore I can say with such rules we are rather scanning 'what has been done' and setting long-term goals, instead of truly driving development.
-there is also a duality on what to do. sometimes we want to break records (just to stay withing lamentable 121m) now I see somebody recommends aerobatics (yes go and do it in midwinter with your helo between 120 and 70m).

My proposal: single-target photo mission, select your favourite target 500m from launch point (something you wouldn't do in manual), fly 120m or less over it and make a photo from all 4 directions (20-30deg look-down). Big bonus points for additional overhead vertical photo form the same flight. Bonus points for 8-directional photo. All photos must be at similar altitude +/-10m. Max number of photos in the mission is 9.
The mission would have 3 subclasses:
1. 1 photo sideways (least precision required)
2. 1 photo overhead
3. 4 sideways 1 overhead (bonus points for 8 sideways, this is technically similar)

• ▶ Reply

Permalink Reply by Jesse & Jared on September 3, 2009 at 11:22am

Skywriting would be fun. So would night flying. Aerobatics would be sweet!

• ▶ Reply

Permalink Reply by Gary Mortimer on September 3, 2009 at 11:42am

Yes the Skywriting idea is cool, I guess T 3 would be the target.

Don't worry I have not forgotten your twinstar entry. Not decided how to do it yet!

• ▶ Reply

Permalink Reply by Krzysztof Bosak on December 16, 2009 at 8:58am

Night flying is... annoying. I did T3-3 in the night 5 times.
It requires automated landing. At the end nothing to show besides logs.
And why should we try the night if we restrain to VLOS and 121m?

• ▶ Reply