3D Robotics

 

3689413502?profile=originalThe results are in, and we have some great reports from the AUVSI student competition. A Paparazzi-powered UAV came in first, and APM made its best showing ever, with APM-powered UAVs in  7th, 15th and 16th place.

 

Dan Strider reports:

There were 26 teams registered this year, 22 of which showed up to competition and I believe 17 of which had successful flights (two teams had ground troubles and would have otherwise flown). Of the 17 flying teams, I personally saw several autonomous takeoffs and landings and most teams flew on autopilot for a significant portion of the flight. It seemed more teams flew still imagery rather than video, though it still surprised me how many teams did not have in-flight transmission of the imagery. WIFI was quite common, especially the Ubiquity hardware. Several teams were flying 72 MHz, though it seemed most were on 2.4 GHz (some despite the 2.4 WIFI) and several were flown through their comm links. The aircraft themselves were perhaps the most conservative across the field we've had. There was a lone quad-rotor representing rotary wings and one canard (which flew quite well I might add). Many teams flew trainer-style Telemasters/Kadets/Rascals, though the foam and/or composite molded aircraft were popular as well. Sizes ranged from about 4 lb up to 55.0 lb (their quote, heh). There were some crashes on Saturday, but Friday was remarkably a clean day and shows the caliber and professionalism of the teams.


Drumroll please...

The overall winner was Utah State University FOSAM
, with a near sweep of the categories. Their flight was fully autonomous: one mouse-click to put it in flying mode and one mouse click to select the landing mode. The in-flight imagery system was second to none, with onboard automatic target recognition and three imagery operator stations on the ground. Utah won two years ago as well, making them the first two-time winner. 

Second place was North Carolina State University, the only team to find all five targets and figure out the acronym A-U-V-S-I. They also did an autonomous takeoff, but didn't quite pull off the auto-landing. Their flight was fully autonomous except for the last 3 feet before touchdown. NC State was the overall winner last year and is the only team to have competed in all nine years of the competition.

Mecha adds:


I'll make a quick mention of some of the autopilots used in the event this year. Overall I saw an increased number of Paparazzi users and ArduPilot Mega.

At least 3 universities including 1st place winners Utah FOSAM team used Paparazzi. Several Piccolo users, NC State and the Air Force Academy amongs them. There where a few kestrels (can't remember the teams). Several teams with MicroPilot including Florida International University. At least 3 teams with ArduMega: Bucknell University, Embry Riddle Aeronautical University and M.S. Ramaiah Institute of Technology. 

There where no teams this years that I am aware off that used Attopilot, Unav or FeiYu Tech.

Here's is the final list, with autopilots in bold where known:

 

1- Utah State University FOSAM -- Paparazzi
2- North Carolina State University -- Piccolo
3- Sherbrooke University (Canada) -- Paparazzi
4- United States Air Force Academy -- Piccolo
5- California State at Northridge
6- University of California at San Diego -- Kestrel
7- Delhi Technical University (India) -- APM
8- Kansas State University -- Piccolo
9- University of Arizona -- Piccolo
10- M.S. Ramaiah Institute of Technology (India) -- Paparazzi
11- Utah States University ROSAM -- Paparazzi
12- Cornell University
13- Virginia Commonwealth University
14- Mississippi State University -- Piccolo
15- Bucknell University  -- APM
16- Embry Riddle Aeronautical University -- APM
17- Florida International University  -- Micropilot
18- Great Mills High School --- Micropilot
19- Rutgers University  -- APM
20- University of Texas at Austin -- APM?
21- University of Texas at Arlington -- APM?
22- Wentworth Institute of Technology
N/A - University of California at Los Angeles
N/A- Illinois Institute of Technology
N/A- Hampton Roads Area High School -- Homebrew
N/A- Cal State Poly Pomona

 

[Picture at top from Dan Strider]

E-mail me when people leave their comments –

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

Join diydrones

Comments

  • Using a closed source system still has many, many issues. Its taken 3 years of hard work by a dedicated autopilot team to get NC State to a 1st place win in 2010 and a 2nd place win in 2011 at the AUVSI SUAS. Mississippi State failed to fly their aircraft this year after taking 2nd place last year and winning 3 years ago with a similar system to what they have today, all because of a lack of testing and some firmware bugs that often plague new autopilots.

     

    Open source is fantastic for learning about the hardware, making improvements and most importantly fixing bugs immediately rather than waiting on a manufacturer to make a correction. Winning a competition such as AUVSI SUAS requires more than just getting one autopilot to work, it requires getting a full blown system to work together reliably. The Piccolo autopilot has proven to be one of, if not the most reliable pieces of hardware that NC State owns, allowing for more in-air autonomous flight time and PID optimization than some open source pieces of hardware have across their entire user base. EE's and CS majors are tough to acquire at many universities that participate in the AUVSI SUAS contest, and placing these limited resources on a project such as autopilot development makes little sense if you have the funding to purchase a system an AE or ME can learn to use with enough time. Our EE's and CS team focuses their energy on writing fantastic imagery code and developing embedded hardware systems that support the imagery pipeline and to some extent the autopilot system.

     

    Going forward, I think its clear that Paparazzi is a very capable system and doesn't look overly complicated to use. The price is very affordable for many of the teams at the AUVSI SUAS contest. NC State may purchase one simply to foster learning of the autopilot and aide in development of the system. I and many on our team are not yet convinced that APM is a reliable enough platform to perform at the top of the heap at most of the UAS contests out there. One day I would like to see it there, but its at least 1-2 years away from bridging the gap in capabilities with the Paparazzi's, Piccolo's and Kestrel's of the world.

  • Maybe I will take part in IMAV2011 competition, which will be in September in the Netherlands this year. I'm preparing my plane with ArduPilot Mega now, and will lay oneself out there to show the best results using APM!
  • Developer

    Whan a University uses a closed, proprietary system  for teaching purposes, the ability for the students to learn must be curtailed. And the opportunity for everyone to benefit from the intelligence and creativity of young students is reduced.  Well done to those Uni's that used either Paparazzi or APM or some other Open Source platform.

  • @Happy, I cant wait to see the latest and greatest. Keep up the good work.
  • Admin

    @Nick,

    Nice URSUS. I just about have mine finally built. I am presently painting it in the original colors that Marek used.

    How did it fly? Can you provide your APM PID values?

    Regards,

    TCIII

  • You guys should have contacted me before the competition. I could have helped you meet your requirements. I'm in touch with some guys from the Outback Challenge. We're making preparations a year in advance....

     

    I could have even sent you my latest and greatest (been in the works for a few weeks).

  • Rutgers here, we're using HiTec's Aurora 9 reciever, and like Nick mentioned, we also had to do a bit of fiddling to pass the safety check. We were a bit...surprised...when about 20-30 seconds into demonstrating failsafe mode (full up elevator, full right aileron, full left rudder, throttle cut) we suddenly went full throttle for no apparent reason. 
  • Bucknell here again, just thought I'd comment on the picture. That's our URSUS airframe (the Flying Bison), and I'm the guy in the blue striped shirt. It was quite a surprise seeing myself on my homepage!

    I spoke to Rutgers and Dehli about the APM, and both told be they were using HappyKillmore as their GCS software. Bucknell switched to the planner after the recent update. Also of note, and something I'll try to add to the APM if someone doesn't beat me to it, was the issues some teams had with the APM's failsafe features. Both Rutgers and our team had to play around with the settings to get the failsafe to kick in properly. It would seem some receivers (such as the Futaba model Bucknell used) have their own failsafe, and thus never reach the correct PWM values to get the termination mode to kick in properly. I believe this is already mentioned on the wiki, but it's certainly a feature which could use some tweaking.
  • Hats off to Paparazzi - in spite of many new open and commercial competitors it has dominated such competitions for such a long time one really has to respect its authors and community for creating such a mature and robust system.
  • 3D Robotics
    Thanks all! Post updated with all that info...
This reply was deleted.