Auto pilot certification?

[Flying_Monkey]

It's understandable that Dynon is using it's engineering resources to do the design work on the types with the highest numbers of possible sales. The frustrating thing is that the experimental version of the autopilot can be pretty much installed on any thing a builder builds so long as they don't mind making their own mounts and control rigging.

I suppose the FAA is putting Dynon through a lot more hoops than common sense would dictate. As far as I know Dynon's autopilots haven't caused a lot of problems for all the experimental types they've been installed on. Unless the physics works differently for aircraft that fly using experimental certificates vs. type certificates. A lot of time spent doing proofs does not seem to be warranted. There are conceivably type designs that exhibit flight properties that might not be safe to fly with an autopilot. I don't know of any TC'd aircraft that are in that category since they would probably not get TC'd because it would be hard for a human to fly them as well.

The other thing that probably slows Dynon down is that they feel like they need to have all the installation parts (mounts and rigging parts) developed and on the shelf before they add a type to the AML. That isn't an FAA requirement. It's fairly common that low volume STC's installation designs that require the A&P to fabricate sheet metal parts like brackets. Most of the rest of it could be solved by calling out standard AN parts.

The only way Dynon is going to make a significant dent in their backlog is to hire an engineer to collaborate with type clubs or other type owner groups to develop designs and approved data and walk it through the FAA. Dynon could charge the groups for the engineer's time collaborating with the outside groups. As part of the contract for that collaboration, the resulting design and data could be signed over to Dynon.

A side benefit of that to Dynon would be that they would be getting their engineering underwritten by groups of owners that are willing to put their resources (read time and money) into the process and are likely to actually result in system sales. That takes it out of the realm of a market projection committing a lot of company time and money.

That all makes sense to me. I totally understand that they need to go after the models that will generate the sales and some of the common birds like PA28 have been dominated by Trutrak already so maybe they figure that is a lost segment? Ether way I am super happy with my HDX system - it is just amazing. The certified aircraft world definitely has its drawbacks. After I dont need the hauling capacity of the 6 I'd love to think about building an RV....

 

[Raymo]

FWIW, the Garmin mounts are also the same. Wiring for each brand is very different.

 

[Dynon]

The other thing that probably slows Dynon down is that they feel like they need to have all the installation parts (mounts and rigging parts) developed and on the shelf before they add a type to the AML. That isn't an FAA requirement. It's fairly common that low volume STC's installation designs that require the A&P to fabricate sheet metal parts like brackets. Most of the rest of it could be solved by calling out standard AN parts.

This part is actually not that bad, relatively. Once we've done the engineering and design, which is a prerequisite, we're pretty good about getting parts and kits to market. There are lots of reasons that it's more complex than E-AB AP installation kits, but a major one is that even though we wished that airplanes should be basically "the same" for their model run, the reality is different, and reconciling those differences to come up with a workable installation is often non-trivial (to say the least). The majority of the time and effort is in the former. The latter makes it easier for everyone to actually install (ie, a business decision that we think minimizes costs for everyone involved, which we care a lot about)

 

[Bill Putney]

You'd be in luck with the Navion. Even though there's been half a dozen models they all have the same control linkages. One design would work on all of them. They've had autopilots from Brittain, S-Tec and Century to name a few. Navions have also been made into remotely controlled target drones by the military and fly by wire test beds by NASA and several universities. The point is that they are aerodynamically stable and straightforward to control by servos.

The problem for me is that we're probably about 600th on the list of types you folks will get to. I'll be too old to fly by the time you get around to it. I'd like to offer some other ways to get your AML to include more of the obscure types that won't make it so difficult financially for Dynon. I know one Navion owner who is taking his airplane into experimental, just so he can install an autopilot. That's a pretty extreme measure to take to do something as simple as add 2 servos and a couple of lengths of control cable. 1000's of experimental aircraft have been equipped with better autopilots than the certificated types can get close to. The setup instructions don't vary which tells me that Dynon has designed these to work across a broad range of aircraft designs with out a lot of hardware hand waving. I know S-Tec has probably tried to convince the FAA that autopilots are so hard to do that only they can do them, but I think we've kind of seen through that by now.

 

[chousley]

I have a C172 with Dynon HDX SkyView w/ Autopilot. Have been having continuous issues with ALT hold. I will have the aircraft perfectly trimmed out for straight and level flight at a cruise power setting. Once I engage the auto pilot and Altitude hold if any conditions other than
completely smooth air I will get a "Trim Nose Down" message followed shortly by a "Pitch Slip" message. Autopilot will aggressively pitch and slow the aircraft by 20 KTS or more while only chasing 20 - 30 ft of altitude. Any ideas on a fix?

 

[GalinHdz]

Do you know how the SkyView was connected to the TruTrak? Was it via RS232 or ARINC-429?

 

[chousley]

Not sure. Brand new owner and was not involved with the initial Dynon install (Thrust Flight / KADS) I will research that?

 

[Dynon]

If you haven't already, it's best to talk to our support team for this issue. Contact info in my signature.

 

[Brent Dana]

Question is: if you buy the tru trac, when dynon finally gets their ap done, does the dynon, use same servos, mounts, etc? Or does it all need to be replaced?

 

[Raymo]

Question is: if you buy the tru trac, when dynon finally gets their ap done, does the dynon, use same servos, mounts, etc? Or does it all need to be replaced?

Same mounts but servos and wiring are different (see posts above).

 

[Bill Putney]

Question is: if you buy the tru trac, when dynon finally gets their ap done, does the dynon, use same servos, mounts, etc? Or does it all need to be replaced?

Electrically, the servos are different because Dynon's autopilot is part of the SkyView HDX EFIS and uses the SkyView network to connect to the servos. Tru-Trac probably has their own way of wiring servos that will be different. The circuit breakers for the autopilot servos will probably be the same because there is a fairly direct correlation between the amount of force the servos need to drive your control surfaces and the amount of electrical power is necessary.

Mechanically, the up side would normally be that all the servos I've looked at have very similar if not exactly the same mounting configuration and the Dynon servos and flight control rigging could all be the same as Tru-Trac's. But, Dynon will likely provide servo mounting plates with specific airframe and flight control attachments in their STC. I'm sure Tru-Trac has done that as well and they may be very different. When your A&P installs the Tru-Trac and your IA returns your airplane to service on an FAA Form 337 it will state that the installation was done "In accordance with STC <whatever the number is>". So, having a Tru-Trac installed doesn't necessarily get you very much of a leg up on a Dynon autopilot install since they will have to do the same procedure with that install.

From a tactical perspective, it would be so much better if the type clubs would invest in designing and getting FAA approval for the data to mechanically mount the servos to the airframe and interface them to the flight controls. Then if Tru-Trac, Dynon, Garmin or <your favorite future avionics purveyor here> would base the mechanical part of their autopilot STC on the type club's already approved FAA data, there would immediately be a number of benefits. First, the avionics manufacture could reduce the engineering and regulatory cost and expedite introduction of their autopilot to any type that already had this data available. They could quickly add many more types increasing their market share. Second, if an aircraft owner decided to change autopilots for some reason, they wouldn't have to drill more holes in their airplane to change the mounts to comply with a different STC and the install cost would be significantly reduced for the follow on autopilots. Third, some types that would be way down the list of airplanes to be added to the STC because of their relatively small market numbers would be more likely to get there type included. Forth (this should have been first), it's been shown that a modern autopilot can reduce the general aviation accident rate for accidents involving loss of control, controlled flight into terrain, and VFR rated pilots finding themselves trapped in IMC. This would make GA safer and we'd loose fewer friends.

With regard to this last point it baffles me that the FAA, who's mission (they say) is principally aviation safety still makes manufacturers jump through so many of the same hoops for every type they add to their autopilot STC AML. Autopilot installations on Experimental aircraft have not killed many (any?) people. These autopilots have been successfully installed with very little FAA oversight by "hobbyist" mechanics. Accidents that a modern autopilot could have helped to eliminate kill people every year. If the FAA just played the odds, they'd be ahead by finding a way to safely expedite the introduction of modern, reasonably priced autopilots into more GA types. It's ridiculous that qualifying the purely mechanical parts of an autopilot STC should take so much time and money to implement. This ultimately delays or excludes so many aircraft types while people continue to die for lack of a modern, affordable autopilot.

 

[airguy]

It's the liability aspect. They aren't going to fart in the breeze until it's absolute certain that there is no danger. After the 737Max debacle, they are going to be very touchy for a while.

 

[Bill Putney]

Everyone could have seen the 737Max disasters coming. When you understaff the regulatory agencies that are there to put safety first and decide that you will trust companies who are trying harder to make their shareholders happy above anything else, you get the 737Max. Every system needs balance. If the companies are not driven they don't innovate. If they don't have regulatory checks they take short cuts and the outcomes are predictable. The only way an engineering team that wants to do the right thing can push back against the bean counters is if they have that regulatory oversight to point to and say "I know it takes longer and costs more but we have to get it through the FAA."

That's not exactly what we're talking about here.

The FAA has done extensive paperwork and flight testing of the Dynon autopilot in aircraft from C172 to Bonanzas to now a Seneca twin. The electronics and the software and servos have all been checked out. Most of the aircraft on the Dynon STC that are not certified for Dynon autopilots have had autopilot STCs from other vendors for years. That tends to rule out those aircraft that are intrinsically unsuited to being controlled by an autopilot. If this were any other mechanical installation in a GA aircraft, there would be a chapter in AC 43.13-2B to cover the approved data to do this. It needs a basic structural test and a close inspection to insure that the flight controls are impaired by the rigging.

In all the aircraft that has had these autopilots installed in them, Dynon has never had to change anything to meet the FAA requirements. They've had to do individual testing and paperwork for every type and sub-type (model suffix) for all the Certified aircraft on their AML that allow autopilots. With all those and who knows how many experimental types, if there was going to be a big revelation we'd have seen it by now.

As the saying goes, doing the same thing over and over and expecting a different outcome is the definition of insanity. The FAA keeps putting Dynon through this "procedure" over and over expecting a different outcome. If it was a very new, very complex system (like the 737Max) involved there might be new things to learn and improve on. After 12 years of Dynon making essentially the same autopilot system for every aircraft type they sell to, why do we have to keep killing people so the FAA can feel good about their procedure?

 

[airguy]

After 12 years of Dynon making essentially the same autopilot system for every aircraft type they sell to, why do we have to keep killing people so the FAA can feel good about their procedure?

This last line here - this is the reason.

It is said the all the FAA rules are written in blood - it takes blood on the ground to get them to take action. It took the blood of 346 people from the 737max crashes to get the FAA to sit up and pay attention and make sure EVERYbody is playing by the certification rules that they themselves wrote and enforce. They are not about to back off of it at this point. Without the 737max crashes, you might have a chance of convincing them to lighten up, but it won't happen now.

The FAA does not know, track, or care about aircraft fatalities that occurred and "could have / might have" been saved by a proper autopilot. That will never enter the discussion. They ABSOLUTELY care about the possibility of ONE fatality that could be caused by an improper autopilot design that they approved and said was safe. The FAA as we know it just barely survived the Boeing scandal with their skin intact - another similar incident would result in wholesale turnover and a complete reorg - putting lots of fat bureaucrats out of a comfy job - and they just aren't going to allow that.

I ain't saying it's right - but that's the way it is.

 

[Brent Dana]

With all that being said. I will probably wait for Dynon. i want it all to work together, and i like the Dynon product so much. Also dynon is the reason the other companies are having to compete. And i hate that the other companies took advantage for so long! I will never guy a garmin product again for this reason alone.

 

[airguy]

With all that being said. I will probably wait for Dynon. i want it all to work together, and i like the Dynon product so much. Also dynon is the reason the other companies are having to compete. And i hate that the other companies took advantage for so long! I will never guy a garmin product again for this reason alone.

Preach on, brother.