NASA has announced a new UAS contest that I think might interest people here. It is called the UAS Airspace Operations Challenge and it's being developed as part of NASA's Centennial Challenge Program. Our goal is to hold the competition in the Fall of 2013 and the total amount of prize money available at that time will be $500,000. If the 2013 competition goes well, it will be followed by a tougher competition in 2014 with $1,000,000 in prize money. We're announcing it now, along with a preliminary draft of the rules, to get feedback and ideas from the UAS community and to see if any organizations are interested in partnering with us to run the competition. The rules are incomplete because there are many details that still need to be worked out and we want to be able to improve things based upon your feedback.
This is your chance to give us your ideas on how we can structure this competition to foster significant progress toward solving the technical challenges to integrating unmanned aircraft into the U.S. National Airspace System. The web page lists the kind of information we are looking for and how to submit official feedback to NASA. There is a link to the draft rules on that page.
Garry
Comments
Did the competition is closed also for foreign student university team ?
http://www.aviationtoday.com/av/topstories/R³-Engineering-Tests-ADS-B-Functionality-on-UAS_77591.html#.UIhf66U9l0t
Bill,
You read the instructions better than I did : /
Maybe, if link is lost, then craft has a snap shot of last "image" of the airspace and does some behavior accordingly, meanwhile ground station broadcasts an alert to the ADS-B network that there is an uncontrolled UAS at or around its last position/vector.
Maybe I should go read the contest parameters before going on.
@Gary and @Joshua,
I like the idea of having the ADS-B transponder at the ground station, but it is not entirely clear to me whether that is going to be allowed and/or how it would be scored. The draft contest rules refers to "lost link", so I would think at the very least, there would be a scoring penalty for ADS-B transponder at the ground station.
Also, along this line of thinking, I wonder how much computing we might be allowed to do on the ground? For example, would we be allowed to run a full-blown flight management system on the ground?
Gary, if you are allowed to comment on these questions, could you please chime in?
Best regards,
Bill Premerlani
ADS-B by proxy. In other words, keep the ADS-B transponder on the ground station. This enables high power and proper antennae with no worries about weight, power use, aero drag, and more rf on the craft. The UAS sends its position and vector to the ground station and the ground station broadcasts this out to the world. The pilot on the ground takes care of the sense and avoid. ADS-B has weather info and ground based obstacle info (towers, power lines,etc), all very useful on it's own!
One point to clarify: GPS positions usually have a latency of about 1.25 to 1.5 seconds. That is, if the plane recieves a new GPS position at say time 30 seconds from takeoff, that position will be from time 28.5 seconds.
If all planes and GPS units had the same latency this would not be a problem.
However MatrixPilot takes account of GPS latency and flies on High Bandwidth Dead Reckoning (Integraton of accelerometers and gyros to produce an accurate position 40 times / second). This is an advantage when trying to hit waypoints in 4D, or when flying in close proximity to any object, including the ground.
At 20m/s 1.5 seconds iof latency is 30 meters. Quite significant.
So I think competition organisers will have to clarify whether the CFGPS will have any latency in providing GPS positions, and if so what the latency is. Or altenatively the CFGPS could be designed to have no latency if too uses High Bandwidth Dead Reckoning supported by GPS information (but allowing for the latency).
Furthermore, it is clear that quite a few GPS units have thier own dynamic profiles. These become particular evident when GPS units stop suddenly (as in a crash into a tree). In the susequent logs, viewed in Google Earth, one will often see exactly how the dynamic filters in the GPS respond, sending the positions forward by a further 20 meters or so, before returning to the actuall point of impact. Again, the dynamic properties of the CFGPS will need to be found and described, for the sake of fairness in the competition.
Hey Chris and Jordi,
Here is an opportunity for another new product for 3D Robotics to sell: a small ADS-B receiver. If you hurry, you could probably get one out in time for ArduPlane users to buy for this competition.
Best regards,
Bill
@UAVOZ
then it just a meter of transmitting autopilot waypoints in the same way commercial planes does
Hi Garry,
Thanks for your comments! I'm in discussion with some friends in the US about the possibility of them starting a US team, with members of CanberraUAV being involved remotely. It would be better if non-US people could more directly participate, but otherwise some creative teamwork may do the trick.
I'm also glad that mock up of the CFGPS won't be the final design! If you swap out the satellite modem for a simple ISM band radio (a 3DR radio or RFD900 would do nicely) then you could build the CFGPS very simply with one of the tiny AVR arduino boards, a UBlox GPS and a small radio. You'd just need some encryption in the radio firmware to ensure teams can't eavesdrop on the real GPS data. I'd be happy to add support for encryption in the SiK firmware on the 3DR/RFD900 radios if that is useful. I think the whole CFGPS could be about 100g, and be the size of a pack of playing cards, especially as the distances over which it is communicating are so small.
I sent a quick note to Larry Cooper about the sample missions in the draft rules too. The turn radius shown in those missions is far too tight for normal UAVs (it implies about 8G turns or more). I think you should design the missions around 2G turns, with teams designing their planes to handle 3G or 4G as a safety margin.
Finally, I strongly suggest you talk to Jon Roberts at CSIRO about what they've learnt from running the Outback Challenge. I sent Jon a link to the NASA competition and he was very interested in it.
I'm starting to look at what it would take for ArduPlane to complete the mission for the NASA competition, and in many ways it looks easier than the OBC. I've now got a ADS-B receiver running in GNU radio, and I will be looking at the 4DT path planning over the coming months. I think it would be nice for ArduPlane to support all the key features needed for this sort of SAA mission out of the box.
Cheers, Tridge
A few notes...
Historically, foreign teams have been allowed to compete in previous Centennial Challenges, but legal restrictions on the prize money made them ineligible to win the cash. There are many possible creative teaming arrangements and, if you think of an approach you want to use, ask us about it officially, following the web site instructions. There is also an opportunity here for a foreign organization to partner with NASA and offer a separate prize for non-U.S. teams. We're trying to leave the door open as wide as we can for creative ideas that will maximize the creative energy that gets applied to solving these problems.
With regard to ITAR questions, ITAR is usually described as governing the export of U.S. technologies to foreign entities, but the U.S. State Department is really the only place to get reliable advice on anything related to ITAR. If you submit your concerns to NASA, we'll talk to the State Department and then tell everyone the official policy.
Finally, I apologize for scaring everyone with the mockup of the CFGPS. I don't know how big it will be, yet. I am still deciding exactly what is going to be in there and I'll have to do some compatibility testing to see how closely I can pack things together.
Thanks to everyone who has submitted feedback, so far! I am looking forward to your comments.
Garry
-
1
-
2
-
3
of 3 Next