An airplane that can fly itself (Autonomous Vehicle)

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

Unknown_Target

Active Member
Joined
Apr 20, 2017
Messages
38
Location
Somerville
To be fair, Topaz was complaining mainly about the servo weights, which would indeed have to be pretty beefy. Plus any backup battery redundancies in case of power failure - and/or some way of transferring to push/pull linkages if the power goes dead.

That being said, you could probably dramatically reduce the required weight by putting the servos on the stick instead of the control surface, which would allow for power-out redundancy and much smaller servos due to lever effects.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,982
Location
Rocky Mountains
To be fair, Topaz was complaining mainly about the servo weights, which would indeed have to be pretty beefy.
If they are added to an existing system designed only for pilot feedback/feel. Once this type of system is considered from the beginning the servos can be quite small. For example I've been looking at spring servo tabs for use with small servos to reduce power consumption. They add hardware, and weight where it is bad for flutter, so they may not survive the benefit/risk trade process. Point is there are options and everything is a compromise/package deal.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,982
Location
Rocky Mountains
but what's the failure mode for a power loss in a FBW system on a single engine aircraft?
Very small provided you have proper redundancy with no single point failure. That is the reason I'm interested in power reduction. Saves backup battery weight. I've not really considered a hybrid (human power and servo) system. Getting the 2 to live together and still provide good pilot feel may be a job bigger to solve than I would consider attempting.

Getting things to wiggle with FBW is pretty simple. When things quit wiggling the way they are commanded ...... that flow chart gets pretty involved :speechles

For the purposes of the system that started this thread the probability of failure during the short time it is intended to be used is likely quite small. To me this system seems to be conceptually the same as current auto-braking on our cars. Never intended to be used but still a logical backup - provided it doesn't compromise the basic system and the humans don't adapt by becoming dependent on it. See: Peltzman effect
 

jedi

Well-Known Member
Joined
Aug 8, 2009
Messages
2,653
Location
Sahuarita Arizona, Renton Washington, USA
Complaining about the weight of sensors is pretty silly. I built a model for a guy who had developed a system enabling a UAV, among other things, to fly back to a starting point and land, which it did quite well. The whole thing must have weighed 3 or 4 lbs, and most of that was battery, motor, and airframe. Also, consider how small and light quad copters are now. They all depend on fancy electronics to keep them behaving.

I suspect that a fly by wire system for a GA airplane could be lighter than what's used now. No steel cables, just actuators and some wire. And don't a lot of airplanes have autopilots already?

----------
OTOH, there are already a bunch of aircraft that can fly themselves, though there may be some question of where it comes down. ;-)

For example:
https://www.youtube.com/watch?v=I7WW14dYLJI

Of course, it doesn't have to be that hi tech:

https://www.youtube.com/watch?v=BuG63kSRmeA
And a fellow in Seattle bought one. When he tried to fly it around the Space Needle, the Space Needle blocked the signal. On the automatic "return to home" guess what it ran into. No lives lost and I think he is out of jail now so all is well compared to a GA incident.
 

PiperCruisin

Well-Known Member
Joined
Jan 17, 2017
Messages
247
Location
Idaho
For most of us flying single engine put-puts, none of this matters much. If you don't have robust de-icing equipment, I generally recommend staying out of the clouds. Personal lesson learned. I get that icing is not as big a problem in some areas with a lot of clouds where IFR capabilities are useful (CA coast).

Autopilot can help when the pilot gets disoriented/vertigo, but it did not help Air France 447.
 

Unknown_Target

Active Member
Joined
Apr 20, 2017
Messages
38
Location
Somerville
PiperCruisin, you essentially just said "None of this matters as long as you never make mistakes/nothing abnormal ever happens". I admit, if you get VFR into IFR you have made some serious errors that should call your piloting skills into question. If you get a heart attack or whatever then you should have had a better medical exam. But come on - that's like saying seatbelts aren't important as long as you don't get into a car crash.
 

Pops

Well-Known Member
HBA Supporter
Log Member
Joined
Jan 1, 2013
Messages
9,747
Location
USA.
When I first started flying the SSSC, I built a homemade autopilot using Giant Scale RC servos to operate trim tabs on the ailerons and elevator. Had Wing leveler and pitch. Enough control to fly the airplane but easy to override with the stick. Would lock on a pressure altitude within 25'. The wing leveler was a Solid state gyro from Newark One and alone with manual roll trim. Pitch also had manual pitch trim. LED's on the front panel indicating the direction of control for both. Also standard rate turn control for right and left. Wanted to couple to a GPS for heading control but never did, just turned to the heading and used the manual trim the fine trim the course. The autopilot made the square wave pulse and frequency to control the RC servos. You could unplug the servos cables in the rear of the autopilot box and plug in a RC receiver and fly the airplane setting in the cockpit with a RC transmitter if you wanted. Fly-BY wire.
I use the autopilot for about 2 years. Got tired of the autopilot having all the fun and setting and doing nothing. A Cub is meant to hand fly. In the picture, the autopilot is on the lower right of the instrument panel . Picture of trim tabs ( yellow and black) on aileron and elevator in rear view.

Added-- Also could turn the autopilot off and still have electric manual trim on roll and pitch.
 

Attachments

Last edited:

lr27

Well-Known Member
Joined
Nov 3, 2007
Messages
3,822
Using a system that can be overpowered by the pilot is probably a good idea. Also, if the airplane is stable in and of itself, then the autopilot doesn't have to be all that strong. Using servo tabs and existing quad copter equipment, I bet you could make a system that weighed only several pounds and was smart enough to return to the origin by itself. If the turbulence was really severe, it might not be enough, but if the problem was partial incapacitation of the pilot, getting lost, disorientation at night or in fog that didn't freeze, it might be very useful.

For an idea of what's possible:
https://web.archive.org/web/20130502190515/http://www.uthere.com/products/ruby/ruby_index.html
This product is no longer offered, but it was for real. I helped test it in a model of 4 foot span. The system was only a small fraction of the aircraft weight.

I expect, using servo tabs, the servos might weigh only an ounce or two and the battery as little as half a pound! So maybe the whole thing for a pound. Pound and a half with lots of wire?
 

smittysrv

Well-Known Member
Joined
Mar 24, 2009
Messages
71
About 20 or 25 years ago, an experimental (non-TSO’ed), box that was going to solve that problem got a fair amount of press. IIRC, it was essentially a wing leveler that would keep the airplane right side up, and when a button was pushed, it performed a 180 degree turn.
Didn't know about that. Interesting.
 

PiperCruisin

Well-Known Member
Joined
Jan 17, 2017
Messages
247
Location
Idaho
PiperCruisin, you essentially just said "None of this matters as long as you never make mistakes/nothing abnormal ever happens". I admit, if you get VFR into IFR you have made some serious errors that should call your piloting skills into question. If you get a heart attack or whatever then you should have had a better medical exam. But come on - that's like saying seatbelts aren't important as long as you don't get into a car crash.
Not what I was saying. It is interesting tech, but one still needs to make smart choices. You may have an autonomous egg-beater, but if it gets ice, it is coming down and maybe not in a nice controlled fashion. However, de-icing equipment is expensive and heavy and most don't have it.

There is a lot of fabulous tech out there, great news for situational awareness, but de-icing is still a limitation in my eye. I'm hoping the tech does not encourage a lot of bad decision making.
 
Top