Comments

• 

  JB April 23, 2016 at 11:27pm

  @Stephen Gloor

  You say: 

  "They did not like our GPS failure procedure as we still had the dead-reckoning return which is not allowed this time. "

  "Still had" and "allowed this time" indicates that your team had this functionality previously.

  From memory the previous WestcoastUAV D1 and under the previous team name, H2Joe's D2, or the Skyhunter aircraft used in the 2014 OBC, did not use the Pixhawk OBC_module etc. for failsafe anyway, as your current team had manufactured their own failsafe hardware and software from scratch. From what I can tell it only mentions dead-reckoning "as compliant" in the compliance statements but is described nowhere else in these previous documents. 

  However, I know that when you were with PerthUAV, we used Tridges failsafe/termination design, hardware and software, as well as the included dead-reckoning, and it was part of the termination/failsafe setup in the Pixhawk we took to the 2014 OBC. These were obviously included in PerthUAV's D2's etc, with appropriate references to Tridges work.

  Out of curiosity how did your team implement the "dead-reckoning" previously, and now again for the 2016 D1, if your teams custom failsafe was not part of the Pixhawk integrated failsafe hardware and software solution?  
• Developer

  Andrew Tridgell April 23, 2016 at 9:17pm

  we got a 'go' on our D2 in CanberraUAV today. From the website it looks like 10 nteams got a 'go', and 13 got a 'improve and resubmit'

  https://uavchallenge.org/

  

  UAV Challenge

  Flying robot competitions
• 

  JB April 23, 2016 at 9:12pm

  Cool that you got through CUAV Team! (Not that there was any doubt!)

  Looks like we'll see you there, PerthUAV have a "GO" without a review, so I'm thinking we're top ten as well, and will get through into the top 20 regardless of who else makes it through the review process. There's 10 with a go, 13 with a review and 2 with a no-go. I'd assume those 13 with a review will be scored accordingly and will have to battle it out for the remaining 10 places still available. Not that many entries this time with only 25 D2's in total.

  I'd like to know the scores!

  Funnily enough they even showed our our previous 2014 aircraft on the announcement...what does it mean? ;-)

  Items mentioned from the Go email:

  • Geofence altitude is 1500ft
  • Crossing the geofence should result in termination
  • If you don't have GPS you can't detect a geofence - so flight termination is the only solution after a short loiter - no dead reckoning to base ( I don't know how it could given the restrictive geofence and corners, as a simple heading is not possible, depending on where you are on the track)

  We're obviously very HAPPY with the result! Yay! :-D
• 

  Stephen Gloor April 23, 2016 at 8:49pm

  WestCoastUAV got an Improve and Resubmit.  They did not like our GPS failure procedure as we still had the dead-reckoning return which is not allowed this time. 
• 

  Stephen Dade April 23, 2016 at 8:07pm

  CanberraUAV got through.

  The next 5 months are going to be busy!
• 

  Mark Omo April 23, 2016 at 7:39pm

  D2 decisions went out who got in?
• 

  JB April 18, 2016 at 6:26am

  Awesome Tridge! Can't wait to try it! Especially the new windvaning!  
• Developer

  Andrew Tridgell April 18, 2016 at 4:56am

  For those of you using an ArduPilot QuadPlane you might like to know there is a fairly major update coming. VTOL landings are much smoother with a new velocity controller, plus you can now do transmitter based tuning. See this link for the new tuning options:

  http://ardupilot.org/plane/docs/parameters.html#tune-parameters

  We used the new code to tune our QuadStar on the weekend and then did 3 full auto missions in 12 knots wind and it flew very nicely. Landings were very smooth and the tuning really helped raise the stability in wind.

  I'll try to get some tuning docs written soon, but if you are familiar with manually tuning a quadcopter with transmitter based tuning then you should find the above parameters fairly easy to follow.

  I'm also working on an improved weathervaning approach, thanks to suggestions from Jon and Leonard.

  I'm finally starting to think that a QuadPlane really can do accurate landings in high winds. It isn't really fully proven yet, but we're getting closer. You do need a pretty high power quadplane though, as the motors need to work hard to overcome the wind.

  Cheers, Tridge

  Complete Parameter List — Plane documentation
• 

  JB April 18, 2016 at 4:34am

  Warren 

  Shame you didn't make it through in time. I'm a bit surprised seeing you have one of the few quadplane platforms that are showing transitions and VTOL, let alone in high wind. You were nearly there which can't make it any easier to swallow! Scheduling flights nearly caught us out because of bad weather as well, so I'd advise everyone to heed the OBC advice and get the flight logs for D3 done as soon as weather allows rather than last minute, so as not to run into the same issues. 
• Developer

  Andrew Tridgell April 18, 2016 at 3:00am

  @Warren, did you ask the organisers if they could make an exception for the video being late? They may well be willing to let you submit a bit late.

  Nice video btw! Do you have the log available somewhere?