Dionysus Design UAV Blog
Hello and welcome to the blog that outlines the UAV development for the ARCAA UAV Competition in Australia during the week of Sept. 24th. The contest website is located here: http://www.uavoutbackchallenge.com.au/.Rules are located here: http://www.arcaa.aero/rules.pdf
We chose to compete in the search and rescue mission. We will take off from an airport in Kinaroy, QLD and fly over a pre-defined area. Video from the airplane will be sent back real time so we can look for a 'stranded hiker' which will be a dummy placed in the outback. Once we locate the hiker, the judges will be notified of its location and they will hopefully authorize us to drop a water bottle. Once the drop is complete, the airplane is to return to the airport. Seems simple right?
There are many complications, the first one being the water bottle size. It should be 600ml and openable using just your hands. For those of us in the states, that is 24 ounces of water (a pound and a half). Strapping that to a model airplane along with enough batteries to fly for more than 30 minutes is not easy.
The next complication is the autopilot system. The airplane needs to fly as autonomous as possible. GPS guidance is a must and the circuitry must be custom. There are rules about the circuitry as well. There must be a safety system that will override everything else and cripple the airplane on command. That is to ensure the safety of everything but the airplane... :) The safety system must be on it's own power supply and monitor the autopilot. If the autopilot locks up, then the safety system will cripple the plane. The airplane cannot fly by GPS alone (I proved that point by digging an airplane sized hole at the local field) so other sensors will be needed to stabilize the airplane in flight.
The next complication is the fact that everything will be packed up and shipped to Australia as luggage.
First, to qualify for the contest, we had to put together a team of people and submit a proposal. The proposal is located here: TBD
The team consists of:
Dan Jones: Team Leader / Software Development
Clinton Lazzari: Safety System / Electrical
Adam Wurts: Airframe support and videographer
Chris Lake: Australian Liaison
We really tried to make the whole operation as inexpensive as possible utilizing CNC routers for PCB fabrication and a CNC foam cutter for airframe development.
First concept autopilot photos are shown here:
The CNC foam cutting machine was used to make an 8'4" wing for the prototype airframe:
Here is the airframe in it's first real flight:
Due to it's size, it was impractical for shipping for testing and for the actual competition. So a smaller one was made.
|
|
|
|
|
The smaller UAV had some good flights, but it still flew like the large one. Way to slow! The wing surfaces were way to large and it wasn't aerodynamic. However, when tested at 7000 feet, altitude proved that the wings were too small:
So back to the drawing board! It would take me a week to make an airplane on the foam cutter, and two days to order one and put it together. We actually had more money than time, so out came the plastic!
I found a Super Dimona 2400 airframe in stock and ordered it. The fiberglass fuselage was big enough for the electronics, but having the motor in front complicated the camera placement.
Maiden Flight? Yes, but I flew it to slow and it tip stalled. Luckily the motor and electronics survived! I actually missed the deadline for providing proof that the airframe was capable of flight and dropping a water bottle. Luckily the good folks at the competition let me turn in the following videos a little late.
The next model can be seen flying in right and left hand turns here:
It worked very well and we were able to test drop a water bottle:
Here is a video that outlines the safety system as defined by the rules in the competition:
While the airframe was being tested, the electronics were being tested in a Multiplex Easy Star airplane. Using that airframe proved to be a valuable asset since it was a fast, smooth flying airplane that survived crashes. The first autonomous flight using our autopilot was on Sept. 1st, 18 days before we needed to leave for the competition!
Australia
Luckily the flight over was pretty uneventful. The airplane made it safely and we got past the security screening. Here is a photo of the airplane stuffed into a hard shell golf club case. EPP foam was used to suspend the airframe during transit.
My first tasks were to buy a 36Mhz radio system and some pipe for an antenna stand. Chris helped out by driving us to a hobby shop. We tried to find a hardware store in Brisbane for some PVC pipe but people kept looking at us with a dazed look on their face. Apparently the home makeover phaze hasn't hit Brisbane yet! We eventually found a small hardware store that actually carried PVC pipe and bought enough for a stand.
Competition
We arrived the night before the competition and got a chance to drive around the course. We suddenly realized how big it was! 5 miles doesn't look so bad on paper or google earth pictures, but it took us about 1/2 hour to drive up an down the course in a car.
We also learned that two other teams in the competition suffered devastating crashes a couple of days before. Team Silvertone and us had the only working airframes.
I also learned that I had mis-read the rules and the other teams were using off the shelf auto-pilot systems. This wasn't a problem except for the fact that I couldn't talk with anyone about the gritty details of implementing an autopilot system from scratch. The reason we went o the competition was to show the world that we had made our own, so I never did plan on using an existing one.
The first two days were devoted to the high school students. They had to drop a golf ball on a sand pit in between two barriers. The airplane they used also had to demonstrate automatic takeoff and landings. It was really neat to see their solutions to that task!
Problems
We had a radio channel conflict with one of the high school teams. This wasn't an issue except for the fact that I needed to test the radio in the airplane. The radio I bought was JR and I am used to Futaba gear. Compounding the radio issue was the fact that the radio was Mode 1 for Australia and I am used to Mode 2 in the states. I'm sure I voided the warranty when I switched gimbals on the transmitter so I could use it just like my old Futaba brand radio. The gimbal switch made the channel assignments very confusing and it took a while to determine the right way to read in the reciever signals into the autopilot.
Monday night was spend writing firmware in the closet of the Bed and Breakfast that we stayed at. The safety system had a heartbeat pulse monitor that made sure the autopilot did not lock up. The testing earlier in the day showed that the heartbeat pulse worked, but wasn't reliable. The problem was fixed by initializing the pins on the microcontroller and shutting off all of the extra functionality that we did not need.
I had also fried my drop servo during a previous crash. This became apparent when I plugged it into the safety system and the power supply shut off. So I couldn't demonstrate a drop for the judges. At this point, Team Silvertone was finalizing their setup on the best looking UAV at the competition.
Flight
We passed our air worthiness safety test by demonstrating the safety system and all of the required abort sequences. I was then allowed onto the air strip to take off and do the demonstration flight.
When the plane landed, it hit the camera on the bottom, so the camera would not have been able to be used again. I had another camera with me, but actually being able to fly was a major accomplishment. I was happy with that and it did not matter if I actually attempted the mission. Which was a good thing! During the test flight, I was able to put the autopilot in Autonomous mode twice. Both times resulted in an altitude climb that was not planned. So the autopilot had issues with altitude and I'm sure it had problems with GPS coordinates too, simply because this was the first time it has flown in the south east hemisphere and the GPS coordinates are exactly backwards from the north east hemisphere where I did all the flight testing.
The judging panel consisted of some very esteemed individuals, including the president of Boeing, Australia! It was a pleasure to meet each of them and quite an honor that they came to such an event.
After my flight, Team iSOAR performed a demonstration flight even though they previously dropped out of the competition due to a crash. Then Team Silvertone went out to the airstrip. Unfortunatly, the motor did not shut off when it needed to and they were not able to demonstrate the Flamingo. This was truly disappointing as it was the favorite to win. Everyone watching the event wanted to see it fly. Hopefully next year we can see it in action!
Results
All of the teams that made it to the last 2 weeks of the competition had excellent entries and should be very strong contenders for next year. However, nobody actually attempted to fly the mission. Because of our demonstration flight, we managed to get the most points for the competition. A very nice trophy was awarded for that!
It was truly a surprise to get that far. Everyone else had come so close and if any attempt was made towards actually flying the mission or demo flight, we would not come close to taking home the trophy.
As part of the competition, we submitted a video documentary to outline our progress during the months leading up to the competition. Here is the video:
Unfortunatly the video was too long for YouTube so I broke it into two seperate videos.
We also won the best documentary prize by 1 point! The guys at QUT, iSOAR and Silvertone put together awesome documentaries. Hopefully each of them will be available online for viewing.
Lessons Learned
Helpful Places
Here is a list of websites that we aquired some of our equipment from:
Conclusion
We simply had a blast! I would recommend it to anyone interested in UAVs. The high school kids did a great job with their competition. Hopefully this is an annual event! If so, I expect next year's competition to be very fierce. I also thoroughly enjoyed sharing ideas with the other teams. Everything that we were doing has already been done before, so there was no point in holding back conceptual information. We each took pride in our implementation, since that was unique and the most challenging part. Learning from each other will ultimatly produce a much safer industry that people can respect. It is our duty not to give people a reason to shun UAV development. Safety first!
