Hi,
Open source autopilots are widely used in the RC toys. I tried 3DR's pixhawk. It seems to be quite okay. There are also other open source autopilots.
The commercial autopilots are always used in commercial unmanned aircraft. They are somehow much more expensive than open source. I’m wondering what are commercial autopilots’ typical advantages compared with the best open source autopilots for conventional fixed wing usage. Could anyone give any hint on that?
Best,
Anna
Replies
Gustav, I get your point. From my perspective I am building UAVs for researchers that have no RC pilot skills and their budgets do not allow for military grade equipment. Almost every time I try to train a beginner to fly one of my planes the result is my UAV gets pile-driven within 30 seconds. I now have a trainer that is more epoxy and packaging tape than EPO. Some people are just naturals at the controls and others can't do it, even with training. So to build the most robust system for a non RC flyer I just go auto-pilot all the way from launch to land.
But obviously if you can fly RCs well this will greatly expand your ability and scope of UAV usage.
Anna,
Imagine that you purchase a small real plane and try to fly it without anyone showing you anything. That is what is happening now in the market, along with very low MTBF hardware.. People learn the hard way. At sellers' advantage of course.
So I would say that lack of proper education and not-that-cheap hardware cheapies are the main reasons behind any failure in this market.
Anna, I program a commercial autopilot system, modifying it's UI to specific customers needs.
I did not write the primary code, but I have studied it, and the main difference being sensors input.
The commercial unit I work with uses TRUE gyro's, not rate sensors, or any data from MEMs devices what so ever.
Also there are at least 2 of everything, including primary core. The signals this unit receives is VERY clean, accurate, and there is a lot of code to detect any kind of problem.
Where as the systems for drones (the ones I have worked with), are really all beta's. No offense, but they are not tested anywhere like the commercial units, nor do they have the back-ups.
In short, they are a different animal :)
Hi Dan,
What do you mean by "true gyros", and what technology are then if not MEM's? Laser ring gyros, or ?
Also how is primary core redundancy managed? Somewhere, there must be one final IC which does the switching. How does it determine which of the two "cores" is the correct one to trust? And how reliable is the final IC?
What I'm getting at here, is it is almost impossible to have true redundancy throughout the entire system.
True gyros.... aka mechanical, they give your true attitude, absolute, same units used for your instruments. MEMs gyros are just rate sensors.
The two cores run same program, if data is different, there is extensive code to determine which is correct. As far as I know, this has never happened. The final IC, is in fact the UI IC, the processor for UI / screen, the program I touch. Data from the two cores, can be processed by integrated displays, outside of the autopilot screen / controls.
Note Rob, these are named aircraft autopilots, not drones, in case it wasn't clear.
About the only thing that is not redundant is the screen, which has failed in past. However with the newer integrated systems, the pilot and co-pilot each have a screen, plus the autopilot screen, or often, just buttons.
We have also had issues with flash memory, which is why we still use battery back up type.
Ok Dan, so you're talking about real autopilots on manned aircraft. Yes, completely different animal indeed.
I wonder how they compensate those mechanical gyros for long-term drift. Over the time scales and distances flown, if the gyro were theoretically perfect, you'd have to account for Coriolis effect, etc. If you flew an airplane half-way around the world along the equator, based on a perfect gyro, it would try to pitch it upside-down. Or if you hovered a helicopter at the equator, facing north, for 12 hours, it would roll it upside-down.
The gyro gimbals must be weighted at the bottom? But then this would also make them susceptible to error when flying in a circle for long periods of time?
What you're talking about Rob is apparent drift. In manned aircraft and with heading gyros, this is traditionally compensated manually either by reference to compass or by an open-loop correction schedule when magnetic north is indeterminate. For attitude indicators, they are biased to average the gravitational field over a long time constant, so your theoretical helicopter would stay upright! :)
I imagine the attitude indicators used in the space program were not biased.
Rob,
I helped to build a BD4C that has two of these fitted.
Not a single mechanical gyro here............
http://www.mglavionics.co.za/iEFIS.htm
just ignore that gyro statement, probably refers to old mechanical gyros that are still used.
The same gyros/accelerometers we use 6050 are used in some missile and weapons systems, which have higher performance than manned aircraft. - Even tho Invensense officially says the standard version is not made for high-risk applications, they know they are used for such as well.
About mechanical gyros, anyone recognise these?
Must have been a bitch to write code for them....
-
5
-
6
-
7
-
8
-
9
of 15 Next