Ardupilot for the win! APM lead developer Andrew ("Tridge") Tridgell's CanberraUAV team with a 3D Robotics Pixhawk autopilot completed all the challenges and scored an impressive victory in this premier UAV competition. 

Congratulations to Canberra UAV and indeed all the teams that found and rescued Joe. Everybody that qualified deserves a round of applause its a mission just to get to the venue and get your airframe through the strict requirements.

A bottle drop distance of 2.6m that's insane.

I personally believe the Outback Challenge has advanced unmanned aviation at a faster rate than any other competition and we should all be thankful for its existence. 

If they get to keep Joe I hope there will be a series of road trip selfies. He has had quite a wait so I am sure ready to party it up a bit.


E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • Congrats Tridge & the entire dev team !

  • Developer

    Congrats guys!

  • Moderator

    What has to be remembered is when this competition started just keeping a model aircraft the right way up with an autopilot was hard. $10k + autopilots could not win the task, now a $250 autopilot has The next level has to be super difficult to reflect just how much autopilots and airframes have come on. It must seem impossible. Just like this did. 

  • 300km

    Magic, impressive what you guys have done. Well done. I'm looking forward to any demonstration of your computer vision system. It was impressive last time, I'm sure you've refined it further.

    I think the organisers have said once Joe has been successfully rescued they'll make it harder. Any ideas how they'll do that? Would it upset your computer vision if he was standing up?

  • Developer

    Kudos for all the Canberra UAV Team!!!

  • Developer

    @Chris, it was auto takeoff and auto landing. It was a bit fast on the landing, likely because I setup the approach glide slope a bit steeper than usual (that was a mistake). The strong cross-wind (around 14 knots) didn't help either. The Lidar picked up the runway nicely at 40m, and it then needed to descend a bit more than it expected as there had been a bit of barometric drift.

    The actual touchdown was OK, but the flare ballooned more than I would have liked. We obviously have a bit more tweaking to do on the autoland code.

    Another slight issue is we were landing heavy. We still had 3/4 of the fuel in the tank, making us about 13.5kg. It was a pretty heavy plane to land, especially with that cross-wind.

    Cheers, Tridge

  • Moderator

    Awesome work Tridge and team.  Just fantastic!

  • Great going team CanberraUAV. What a fantsatic result!!!

    Did the plane land using AUTO or was it manual ?... It looked like it did a little touch-n-go hop on the landing.

  • Amazing that it separated the 1 white blob from the other white blobs.   For autonomously calculating a ballistic trajectory by entering an autonomous pattern & autonomously measuring the environment, it's better than any military system.  Smart bombs still need an optical sight right until impact to hit a target.  TOW missiles need 30 minutes of manual calibration & require a wired connection to the launcher until impact. 

  • 100KM

    Congratulations to all of the successful teams and of course especially CUAV.  To go from 0 to over 35% success rate for placing teams is an amazing progression.  With the advancements that came as a result this competition sure to spawn future innovations, I wouldn't be surprised if the parachute photo ends up hanging in a museum one day.

This reply was deleted.