3689548830?profile=original

This past week the inaugural NASA UAS competition was held at Kennedy Space Center.  This was an inter-agency competition that pitted teams from Kennedy Space Center, Johnson Space Center, and Marshall Space Flight Center against each other.  The mission concept was to design, build, and fly and unmanned aerial system in a mock search and rescue mission.  We had about 18 months to work on this in our spare time.  The objective was to autonomously fly an unmanned aerial vehicle with an imaging payload to locate targets on a field and report the target locations and provide pictures of the targets.

 

There are six of us that made up the core of the Marshall Space Flight Center team called Aero-M.  Our vehicle was designed around a 3DR Hex frame and the APM 2.5.  We used ArduCopter 3.0.1 for the competition and flew a completely autonomous mission (takeoff, multiple waypoint search pattern, landing).  We used the 880kV motors with 12x38 props for maximum efficiency.  With four 6000mAh batteries, we were able to achieve flight times of over 30 minutes.  Our total vehicle cost was around $2800. 

 

3689548910?profile=original

 

Our competition consisted of a large electric helicopter and a custom airplane.  The helicopter used an APM 1 and custom code, and the airplane used a Piccolo flight computer.  The helicopter cost is around $6000 and the airplane cost is around $16000.

 

The flight day of the competition consisted of 2 flight missions for each team.  The first mission had mannequins in orange clothing to identify.   The second mission had mannequins in street clothing and camouflage clothing to identify.  At the end of the second mission there was an endurance test to see how long you can stay in the air.  We ended up being the only team that flew autonomously.  The other teams could not get their flight computers working and had to revert to manual flight.  We also ended up with the longest flight time since the helicopter could only last 15 minutes and the plane had a malfunction at the 18 minute mark.

[Ed: THEY WON! (see comments)]

 

I wanted to take this opportunity to thank 3DR and all the developers for evolving ArduCopter into a flight computer that can easily compete with professional products.  The other teams and the judges were very impressed with the ease that our vehicle handled the high winds (sometimes over 15mph), the rock solid tracking between waypoints, and the amazing loiter during the endurance phase.

E-mail me when people leave their comments –

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

Join diydrones

Comments

  • Congratulations! That is impressive.

  • The 12x45 were APC from 3DR and the 12x6 were from HobbyLobby.

  • Ok, so you didn't go bigger than 12", that explains that.  Which brand were the 12x45 and 12x6?

  • I'm sure you could get more efficient with larger than 12" props, but we only looked at 9-12" props.  We tested 4 different motors (850kV, 880kV, 900kV, and 950kV) with 7 different props (9x47, 10x45, 10x47, 11x47, 12x38, 12x45, and 12x6).  Here's a graph of the results:

     

    3692831193?profile=original

  • Garrick, what other propellers and motors did you compare to?  I do like those props, but I'd be surprised if that was the most efficient system available.  Though I guess you were constrained in prop size due to the frame size.

    Make sure you show the heli guys these videos:  ;)

    https://www.youtube.com/watch?v=WDIg0Q5jRuk

    https://www.youtube.com/watch?v=a4ve9e0ZVlI

  • R, we were focused on long flight times so we went with batteries that maximized the mAh/g.  The batteries we chose are ThunderPower Pro Lite MS series, 6000mAh 3S.  Each battery weighs 381g for a total battery weight of around 3.4 pounds.  Our total vehicle weight of 8.6 pounds means we had around 40% of our total vehicle weight in batteries.  Besides carefully selecting our batteries we also did a trade study comparing several motors and propellers to determine the most efficient combination.  That led us to the 880kV motors and 12x38 propellers.  Our measured flight time ended up being around 35 minutes.

     

    As far as why the team with the helicopter didn't use the ArduCopter code, it was a decision made last summer when they started their design.  I think they tried to use ArduCopter without any tuning and thought they could do better with their own code.  The results of the competition answer how wise that decision was.  :)

  • I see linear polarized antennas on that plane.  NASA YOU'RE DOING IT WRONG! j/k

    Really neat; Thank you for sharing.

  • And yet another video of NASA MSFC's Mighty Eagle filmed with our 3DR quad.

     

    https://www.youtube.com/watch?v=mT9Yz7zdaCI&t=4m50s

  • Here are some additional pictures from the competition:

     

    https://plus.google.com/photos/101417360358005006642/albums/5924299...

  • Yeah, probably. ;)

    So can I say that NASA engineers can't make a better heli controller than me? ;)

This reply was deleted.