read this and tell me what you think...

[AirMcNair]

I went and asked one of the Tec guys a question on a system to save our airplanes and people. This was his responce.....

------------------------------------------------------------------------------------------------------------
My question
So I was talking to a buddy last night and thought there must be some way to keep these airplanes from going crazy and flying into a group of people. With a lot of us having 3000.00$ plus giant scale airplans. What if we had some type of system that is GPS guided. Somthing that the servos have to connect to after comming from the RX. This system could have a 5 Axis lazer system just like those Auto Pilot boxes, Connect to a GPS system and gyro with a wire comming from the RX so in the case of PCM Lockout or FM signal loss the System could take over and do what ever is programed. I know it would not be cheep but does anybody think its possible????
------------------------------------------------------------------------------------------------------
First off, consider that if your protection system goes crazy, that could lead to a crash in and of itself, so you have to be really careful with this stuff. The more complicated you make your onboard systems, the more points of failure you have, and you can't always make every system onboard triple redundant and you can't always eliminate every single point of failure. Also, you can't necessarily control things like engine failure, bird strikes, mid-airs, an any number of other unforseen circumstances. Even with all our care and technology, full scale aircraft still unfortunately go down once in a while.

I can imagine that a simple AP that activates when signal is lost, might be good enough for many situations, but what fun is an off the shelf solution. :-)

I've been playing around with ideas in the back of my head for ways to take a programmable autopilot system (like the crossbow unav) and use it to build a safety/protection system.

For instance, you could layout 3 zones in your R/C airspace, green, yellow, and red.

The innermost zone is green and is where you would go crazy and do anything. That would be bordered (encased) by a yellow zone where you are starting to get into dangerous territory ... maybe too high, a bit too low, too close to the trees, too close to spectators, etc. Finally you would have a red zone everywhere else. This would be the do-not-fly zone, over spectators, near or below ground, etc.

You could setup a system that passed control inputs through to the servos directly when you are in the green zone so you have full, normal, manual control over everything, just like flying any standard R/C airplane.

When you go into a yellow buffer zone, the onboard system would start to kick in and limit what you can do. Perhaps it would limit your roll angle to +/- 45 degrees, and limit your pitch angle to +/- 20 degres and possibly right you if you are inverted or some other wierd attitude. You would still have some control, but it would be "cooked." Perhaps if you are in a yellow zone and heading towards the red zone, the system would kick in and force a turn or a climb or a dive until you are pointed back away from the red zone, then return semi-manual control until you guide your airplane back into the green zone.

Then in the red zone, the autopilot would take over completely and expedite your return back into green.

If you detect loss of signal from the transmitter the onboard system would immediately take over, stabalize the aircraft and return it to home and perhaps enter a holding pattern until transmitter signal is recovered.

There are issues though ... if your buffer zones are too small and you hit them at 120mph, you might be able to blow through them before the system and the aircraft can adequately respond and recover. If you buffer zones are too big, then your flying is greatly restricted.

The crossbow unav isn't cheap and it has some of it's own single points of failure, so great care would be needed in selling this as a safety system.

But I think their are a lot of possible directions you could take this. You could put a system like this on a trainer and have an aircraft that a student literally could not crash no matter how hard they try.

You could put a system on a higher end plane as a backup/safety/sanity check.

Another variant would be to program a "fly-by-wire" system similar to an airbus where your stick inputs are commanding a roll rate or a pitch rate, but when you center the sticks, the autopilot holds your current pitch or roll angle. You still have "full manual control" but the results are "cooked" and your sticks are no longer directly connected to the control surfaces. I wonder if a scale or jet pilot might find something like this useful. With on board stability built in, you could make an airplane fly much more stable and "scale like." One of the things I dislike about R/C models is how they waggle around so much because the pilot doesn't feel the onboard forces as they give their control inputs. A smart system onboard flight control system could yield a much more stable, "scale flying" aircraft.

Another thing you could roll in with all of this is that the onboard system knows your stick inputs, knows your aircraft location, attitude, speed, and you could possibly integrate a few other sensors for rpm, engine temp, battery voltages, etc. The onboard system could self monitor the aircraft and cross check different variables to make sure they are in allowable range. The system could flag potential problems before you could detect them from the ground ... dimininshing battery performance, out of range engine temps or rpms for your throttle setting, maybe even be able to detect that you are way out of whack in some control surface input for your current speed and direction (control surface or balance problem?) and then the system could take some action to notify you that there is a problem.

But realistically you are probably looking at somewhere in the neighborhood of $3000 to not take a loss on building something like this so your market wouldn't be too big. But there are people that invest this much or more in their equipment and perhaps a few of them would be willing to invest a bit more for a better or safer flying aircraft?

I've got some xbow unav hardware sitting on my desk, I'm just lacking time to fiddle with it much right now. (Dang day job) :-) The xbow hardware would be capable of achieving all or most or maybe even more than this, it's just a matter of finding the time to develop the code to do it. Se habla dinero. :-)
----------------------------------------------------------------------------------------------------------

[damien]

Hmmm. In the case of the recent crash it was caused by interference. A dual frequency reciever would have prevented that crash. He got a hit on channel 62, so if the reciever had a "backup" channel, say spot 75, and it switched over then the pilot would still have had control.
I agree with the 'tekspert', you can never make a plane uncrashable. A parachute would save the plane and serious injury but the chute could fail or the system that deploys it could fail.
Your idea is very interesting, the technology is up to it, but it too could fail...

[Ion01]

I haven't read the whole thing, I will later as I am at work now but it sounds totally possible but incredibly complicated and expensive. What about a simple system like on the firebird freedom. It has a simple anti crash system that simply has a infared camera that sees the horizon and compensates. I also new someone who put one on a f-86 that had a puller prop up front. It worked good enough to keep him from crashing when he used it. A simple and cheap system like this would make more sense. If you get a hit and it goes into failsafe mode then it cuts the engine, levels out, and glides back on a set glide slope. You could even give it a little rudder so it would glide in a long lazy circle until it was down. A cheap easy way to save the plane and people.

Anyways, that thought hit me as I was reading and I wanted to get it down while it was fresh in my mind. I will read the rest later.

Btw-I read an article about a man who was decapitated on a elevator. Apparantly, 30 die anually in the US alone in elevator and escelator accidents. Our hobby is very safe but it is still nice to see people trying to make it safer.

[jack01]

IMHO, the best backup system would be a second transmitter on a second channel. When you turn this transmitter on, it's reciever would disconnect the first reciever. Then, you can land, and sort out the problems.

I feel really bad that somebody was killed, but the fact remains that this type of accident is extremely rare. Modern RC systems are extremely reliable. Just like full scale, nearly every crash is caused in some way by human error. That could mean a faulty installation, carelessness in frequency control, bad maintenance, and so forth.

[airbusmech]

	Quote: Originally Posted by damien
	Hmmm. In the case of the recent crash it was caused by interference. A dual frequency reciever would have prevented that crash. He got a hit on channel 62, so if the reciever had a "backup" channel, say spot 75, and it switched over then the pilot would still have had control. I agree with the 'tekspert', you can never make a plane uncrashable. A parachute would save the plane and serious injury but the chute could fail or the system that deploys it could fail. Your idea is very interesting, the technology is up to it, but it too could fail...

I have heard of a system like that recently.Can you give more info?

[DKjens]

I already ordered a CPD4 from FMADirect. This auto pilot can control one channel each for two axes. Here is how I am going to hook it up. I use Futaba PCM and a Smart Fly Power Expander. I am hooking this AP up between Rx and PE on right aileron and left elevator. My failsafe will be set to put ailerons and elevators neutral, rudder slightly over and to turn this AP on. The AP will bring the plane to level attitude and the plane will basically fly in a flat turn. The engine can be set to kill, idle or enough power to maintain altitude.

The CPD4 costs $99.95 + $8.00 shipping. This setup will assure that the plane doesn't plow into the ground, but settles in a nise level flight attitude, giving the radio the best possible chance to re-connect.

Another plus with this setup is, that if you are hit by a glitch say on an upline, the plane will try to level off, so there is no doubt about what just happened, and the pilot will know he's being hit, and make appropriate decission after that.

I should get my AP next week, and I can't wait to test it. I will put it on a toggle switch and go up high and turn it on, and see how it works from different attitudes.

[3DFreak576]

This has been going through my head for a while.

http://www.cirrusdesign.com/aircraft/safety/

have it set to deploy after a set amount to time in failsafe so small glitches wont cause the chute to deploy....its a good idea, but not very practical for RC planes

[adrian]

maybe we should talk to weattronics - these guys are pretty up to date with this kinda technology - the weatronic dual receiver systen with gps /gyro might just be the answer - check out the website in my signature - weatronics

[damien]

	Quote: Originally Posted by airbusmech
	I have heard of a system like that recently.Can you give more info?

Sorry but I have also only heard of it, I think it was used in a giant RC boeing747.


Dkjens: Let us know how it works out!

Adrian: One of those weatronics dual systems has been ordered for me, I will be using it in my 34% colombo330L. I am getting the one without gyro and GPS though. I agree that they could be developed into something that is very safe.

[why_fly_high]

I think Cal Orr out in California used to make a system that broadcast on 2 channels. The receiver shifted back and forth to the best one. This was like 20 years ago. I saw it demonstrated on a Kraft radio. Don't know why it didn't really get popular. You had to get 2 pins when it was time to fly.

Dan