Airliner Avionics

[MD83FO]

Will airline avionics ever catch up to general aviation advancements?

[FlyingStone]

Which GA aircraft has triple ADIRU, triple FMC, dual HUD and fail-operational autoland system?

[Capn Bug Smasher]

Not an MD-83! :E

[Pugilistic Animus]

While capabilities aren't the same as a typical modern transport category Airplane, there are many glass cockpit GA aircraft with a lot of capabilities. Look at Cirrus or the newer VLJs. But these capabilities sometimes cause pilots to do stupid things such as trying out " airline quality" IMC which is not smart at all.

[mustafagander]

As I see it, the answer to the question asked is NO.
Airlines have a fleet which is desirable to have as common as possible. A big consideration is certification, is the new whiz-bang kit RPT certified? It is so much easier to upgrade one aircraft than it is to upgrade 25 or so to a standard which really is very little better than the current fit out.

[Uplinker]

It depends what the OP means by 'advancements', perhaps they could specify?

Too much technology could destroy situational awareness - leading to a meaningless video game atmosphere. Real pilots still need to know where the high ground is, what the aircraft is doing, what the speed and thrust are etc. I mean, trains could all theoretically be driverless, some are, but you still need the human element to take account of, say, blind passengers exiting slowly, or landslides or trees across the tracks etc.

I personally think the Airbus A320/A330 family is a very good balance between technology to help, but still keeping the pilots very much in touch and control of what is going on. (I have not flown A350/B787 etc, so cannot comment on those).

I have not flown a HUD but a recent go-around accident involving HUDs make me think that perhaps they are an unnecessary step too far for commercial passenger operations, and not a substitute for properly trained and experienced pilots. Beyond CAT lllC auto-landing, what else do you need?

[sonicbum]

I believe the OP possibly refers to the various synthetic vision tools that are currently widely used in GA and business aviation. I have never flown anything equipped with that kind of gear, but would love to give it a go.

[FE Hoppy]

It's already available but mostly undesirable. There needs to be an obvious commercial advantage for a customer to choose to increase the cost of his new equipment by x$ per unit. Over the last 15 years Ive seen synthetic vision demonstrations on both Honeywell and Collins systems but they offer no advantage over Autoland/HUDLS. Even HUDLS is not that popular compared to AL capability as most products are initially designed with AL and then HUDLS offered as an option. If your fleet is equipped with HUD for a.n. other reason (normally someone advancing their career on the basis of added situational awareness) then HUDLS can be argued to be part of that philosophy. But the OEM isn't going to discount the AL capability if it's standard so you only save on maintenance and waste capability if you ignore it.
Then when it comes to fleet replacement you are trying to offload aircraft with a capability that has a different training requirement so anyone in the market is going to look at the more mainstream versions of the same product.

[Pugilistic Animus]

As a neoludite, that I am, I think that this flight deck picture is state of the art and cutting edge :}

https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.reddit.com/r/aviation/comments/c5oc67/757200_cockpit/&ved=2ahUKEwi1ocSX8avpAhXSG80KHReOD9wQtaYDegQIARAh&usg=AOvVaw06QAdtlDbFZUWrrxKETqBF

[Denti]

In some ways i was surprised to find out that transitioning from the 737NG to the A320 was actually a step back in avionic capabilities. Stuff like vertical situation display, navigation performance scales and IAN (no, managed is not the same, FLS would be) are actually quite good. Not to mention of course, that the alert height of the 737NG is 200ft, not only 100ft. Oh, and it does RNP 0.1 out of the box as well as GLS, and of course the screens are bigger on the 737 and their brightness level can be adjusted in a much wider range. As an everyday working place the A320 is better of course, more automation, the table and a slightly lower noise level and more space in the flightdeck keeps one less stressed over a long day.

Would a synthetic vision be nice? Yes, in a way of course it would be. Is it necessary? No, not at all. I'd rather have the distributed breaking action assessment done by the aircraft itself and then redistributed as information to all other aircraft of my company, tangible benefit for real life winter operation, better EFB integration and stuff like that.

Yes, the flight deck in many GA jets sure looks nice. But i always wonder if it is more about "bling" value, and if it can withstand 6000 hours a year of flying, as many airliners manage.

[turbidus]

Denti,

there is the Panasonic weather data that is transferred throughout the fleet....especially helpful to the trailing aircraft....so the other data could be ported the same way.

[Denti]

How many airports are CAT III certified? How many airports does GA vs Airlines serve? There is your answer regarding synthetic vision, RNP approach capability etc.

A GA operator will see this situation from a different perspective. What´s the use of autoland if over 90% of airports in the world don´t have it? But, you may get lower approach minima with synthetic vision with some CAAs.

I was surprised that the 787 is not certified for RNP approaches. At least not the one I am familiar with. But, then again in the airline world how often do you need to fly an RNP approach?
Actually, not all that rarely. And since airliners can easily get RNP 0.1 certified, not to mention RNP (AR) in many cases, it is actually quite handy to have. The main thing is crew training in that case, training a few thousand pilots to a higher degree of precision is actually quite an undertaking.

Interestingly for something like CAT I LTS (lower than standard) one does not need any new displays, a 35 year old airbus display system is quite enough. And that allows CAT I landings with RVR down to 400m.

But yes, GA operation is of course quite different, as one of the main appeal is being able to get into a lot more airports than the airlines cover, besides the timing of a trip that can be adjusted to personal requirements of the passengers.

[Denti]

Denti,

there is the Panasonic weather data that is transferred throughout the fleet....especially helpful to the trailing aircraft....so the other data could be ported the same way.
Yes, there are several distributed data attempts currently. I know airbus is now rolling out the breaking report thing after it had a non-connected test run with several selected airlines. Check out https://safetyfirst.airbus.com/using-aircraft-as-a-sensor-on-contaminated-runways/

I know several companies are in the process of developing mosaic weather radar pictures, basically real-time radar pictures from all airplanes in an area distributed to all airplanes to enable a better general weather picture. Which could even allow an airplane with inop radar to get a real time picture of the weather situation, kinda like herd immunity (sorry, couldn't resist).

[FullWings]

Will airline avionics ever catch up to general aviation advancements?
No. Because of certification requirements, cost, problems with commonality and the time between design freeze and the first customer getting their hands on it.

My glider has better instruments than the 777 I fly for work. It has 1,000x the memory in the FMC-equivalent, to start with. The outlay to upgrade an airliner avionics suite to tech that is only 10 years old as opposed to 30 is so eye watering that it is rarely done. iPads have augmented a lot of the functionality of the aircraft in a way that would be prohibitively expensive and/or plain impossible conventionally.

[FlyingStone]

My glider has better instruments than the 777 I fly for work. It has 1,000x the memory in the FMC-equivalent, to start with.

It actually doesn't. It's got better instruments to fly in nice weather and it's fairly trivial to land any light aircraft without any instruments when you've got a bit of experience, so no need for any serious redundancy. Also, if one instrument doesn't work, you always have a chance at postponing your flying for a day or two, whereas airlines would go bankrupt without a solid MEL dispatch availability. Horses for courses.

I was surprised that the 787 is not certified for RNP approaches. At least not the one I am familiar with. But, then again in the airline world how often do you need to fly an RNP approach?

A modern aircraft like 787 can't do RNP? That seems very far-fetched. Even an old 737 can do RNP AR 0.10 if you buy some extra toys from Boeing.

[Airmann]

The main advancements in the GA catagory relate to

1) live weather updates
2) synthetic vision

As for the live weather this is only possible because a network (NEXRAD) has been set up around the US that allows for this. Not all countries have this. In any case it doesn't really matter in commercial aviation as weather forecasting is quite good, and we have a weather radar, ACARS and datalink.

What I would like to see however is a transfer of turbulence data from aircraft to a ground station that could then be relayed to other aircraft. I believe that a lot of work is going into turbulence forecasting and its pointless. Much better to just take the data from g loading sensors on board and have the manufacturers work out a way to catagorize it according to turbulence levels and then have that data sent to a centralized location via ACARS.

As for synthetic vision, like others have mentioned it really doesn't add much if one can do a CAT2 or better autoland. As for terraint awareness we have a TAWS database on board, sure the display of terrain on an ND isn't as 'high tech' as synthetic vision but it does the job. The danger is flying the plane using the synthetic vision like a video game. You don't want that in a commercial setting. Having used a HUD for landing I can honestly say that I see no use for it except for certain airports, and even then you only need one for the PF.

[tdracer]

No. Because of certification requirements, cost, problems with commonality and the time between design freeze and the first customer getting their hands on it.


Part 25 certification requirements are the primary driver - any important flight deck instrumentation is considered to be flight critical, which means it's Design Assurance Level (DAL) A or B. It takes a long time and costs a fortune to certify DAL A or B hardware and software. Further, any change to hardware means re-validating the software, and visa/versa. As a result, once avionics are certified, there is huge incentive to leave well enough alone unless absolutely necessary. Parts obsolescence is a huge, on-going problem with commercial aircraft avionics due to the rapid changes in the consumer electronics - it's often not even possible to manufacture some of the electronic bits from ten or twenty years ago because the changes in technology. Back in the 1980s, most of the electronics were mil-spec, specially manufactured for purpose - now days nearly everything is "Commercial Off The Shelf (COTS) (although typically screened to a higher standard than the stuff in your phone). The manufactures simply are not interesting in making a few thousand parts for aviation when they are producing parts in the millions for commercial applications, so we have to use what they are already making.

[KRviator]

It actually doesn't. It's got better instruments to fly in nice weather and it's fairly trivial to land any light aircraft without any instruments when you've got a bit of experience, so no need for any serious redundancy. Also, if one instrument doesn't work, you always have a chance at postponing your flying for a day or two, whereas airlines would go bankrupt without a solid MEL dispatch availability. Horses for courses..Here, you're talking about two different things, capability and reliability.

My RV-9 is, in some ways, more advanced than the latest A350. If something goes out of limits, it won't just go "PING!" and require you to look at the ECAM to find out what went wrong. It will tell you, in your headset "Oil Pressure" or "Electrical Current" or any number of other things. IT has synthetic vision with terrain shading, a glide ring showing what points on the ground I can reach, based on current environment variables, can have touch-screen's installed if I want, displays a geo-referenced position on the visual or approach chart and is interfaced with an electronic circuit breaker system (though, granted the A350 has the last two as well). Another EFIS manufacturer provides for a camera input on which you can superimpose your PFD symbology, allowing a FLIR camera to be fitted for low-light operations.

That being said, while no light single will have autoland capability, I think many folk would be surprised to see the amount of redundancy built in to your typical IFR Experimental. Dual ADAHRS, triple screen (often with a third standby EFIS), dual GNSS and oftentimes dual batteries or alternators as well.

[Uplinker]

What I would like to see however is a transfer of turbulence data from aircraft to a ground station that could then be relayed to other aircraft. I believe that a lot of work is going into turbulence forecasting and its pointless. Much better to just take the data from g loading sensors on board and have the manufacturers work out a way to catagorize it according to turbulence levels and then have that data sent to a centralized location via ACARS.

As for synthetic vision, like others have mentioned it really doesn't add much if one can do a CAT2 or better autoland. As for terraint awareness we have a TAWS database on board, sure the display of terrain on an ND isn't as 'high tech' as synthetic vision but it does the job. The danger is flying the plane using the synthetic vision like a video game. You don't want that in a commercial setting. Having used a HUD for landing I can honestly say that I see no use for it except for certain airports...…....

An automatic turbulence reporting network is a very good idea - aircraft talking to other aircraft in the background, like TCAS, and only popping up with alerts when necessary. That would be good :ok:

I agree with Airman's other points too. As I said in post #6, there is no point having synthetic vision or whatever just because you can. You still need to have SA and fly an aeroplane, and turning the whole process into a video game could cause its own problems, with barely qualified pilots thinking they can fly. (I am not being a dinosaur about this, I write (simple) code myself, and am well into technology).

Light aircraft instrumentation has been awful though, so improvements here would be good. It was bloody difficult to track or fly a hold around an NDB in a light SEP/MEP, but only really because the NDB pointer and the compass bug were on different instruments - and sometimes, the "compass" card had to be manually aligned to a real compass every 10 mins, I mean for crying out loud, as if you haven't got enough to do just flying the thing !!. When you can overlay the NDB needle onto the track pointer of a proper compass or ND, NDB tracking is child's play - so easy ! (Yes, yes, I know, we don't track NDB's any more :))

Engine FADECs, going slightly off piste here: Frankly having to worry about fuel mixture, carburettor icing, priming pumps, magnetos etc. is absolutely prehistoric !! Like most of us; I can casually reach into my average 2008 car in any weather, any temperature, turn the key and it will start, (while I scrape the ice off the windscreen). But a C150 or equivalent??

[Check Airman]

The main advancements in the GA catagory relate to

What I would like to see however is a transfer of turbulence data from aircraft to a ground station that could then be relayed to other aircraft. I believe that a lot of work is going into turbulence forecasting and its pointless. Much better to just take the data from g loading sensors on board and have the manufacturers work out a way to catagorize it according to turbulence levels and then have that data sent to a centralized location via ACARS.


Your idea on the sharing of turbulence data has already been brought to life.

https://business.weather.com/news/wsis-taps-weather-system-could-provide-tracking (https://business.weather.com/news/wsis-taps-weather-system-could-provide-tracking)

It has its limitations though.

[tdracer]

Engine FADECs, going slightly off piste here: Frankly having to worry about fuel mixture, carburettor icing, priming pumps, magnetos etc. is absolutely prehistoric !! Like most of us; I can casually reach into my average 2008 car in any weather, any temperature, turn the key and it will start, (while I scrape the ice off the windscreen). But a C150 or equivalent??

A big problem with applying FADEC technology to your Cessna is cost. Compared to carburetors, electronics have rather unforgiving failure modes - carbs tend to wear gradually, slowly shifting - while electronics simply quit, often with little or no warning. When that happens in your car, you pull over to the side - when that happens in your GA aircraft, best case is you make forced landing.
That's why commercial airliner FADEC installations are all dual channel, fault tolerant, built with highly screened electronic components. They are also quite expensive - LRU cost for a FADEC is typically between a quarter million and a half million dollars. Yes, the actual cost to screw one together is a fraction of that but still approaches six figures. Further, while GA isn't as bad a commercial aircraft, the FADEC is subject to far greater and much more frequent environmental extremes than automotive installations (and large, frequent, and rapid temperature fluctuations are hell on electronics).

[FullWings]

It actually doesn't. It's got better instruments to fly in nice weather and it's fairly trivial to land any light aircraft without any instruments when you've got a bit of experience, so no need for any serious redundancy. Also, if one instrument doesn't work, you always have a chance at postponing your flying for a day or two, whereas airlines would go bankrupt without a solid MEL dispatch availability. Horses for courses.
I get the point about redundancy but my sailplane has multiple redundant inertial/nav systems running off independent power supplies and a certified EFIS with backup. I fly IMC quite a bit and could go IFR if I wanted to.

I think the question was not so much about the despatch reliability and ultimate redundancy of an avionics fit, it was about what it could do when it was working and the answer to that is GA instrumentation will always be ahead of airliner kit because GA manufacturers can use the latest tech...

[Uplinker]

A big problem with applying FADEC technology to your Cessna is cost. Compared to carburetors, electronics have rather unforgiving failure modes - carbs tend to wear gradually, slowly shifting - while electronics simply quit, often with little or no warning. When that happens in your car, you pull over to the side - when that happens in your GA aircraft, best case is you make forced landing.
That's why commercial airliner FADEC installations are all dual channel, fault tolerant, built with highly screened electronic components. They are also quite expensive - LRU cost for a FADEC is typically between a quarter million and a half million dollars. Yes, the actual cost to screw one together is a fraction of that but still approaches six figures. Further, while GA isn't as bad a commercial aircraft, the FADEC is subject to far greater and much more frequent environmental extremes than automotive installations (and large, frequent, and rapid temperature fluctuations are hell on electronics).

Yes, an actual FADEC - operating a gas turbine engine is a complicated beast, and is not cheap.

But there are millions and millions of Engine Control Modules driving around in cars and trucks today. They start engines in the depths of Alaskan and Canadian Winters, and in the middle of Summer in the desert. Car ECMs look at atmospheric pressure, OAT, mass airflow etc etc, and it would be simple to extend the engine map for an aviation engine envelope. Car ECMs also allow and adjust for load, and optimise the ignition timing and fuel mixture hundreds of times a second. They control emissions equipment and modulate charging systems by varying the mark-space ratio of the drive signal. They can be quite sophisticated.

If a car ECM quits, yes, you just park up and call the tow truck, but in my long electronics experience, most problems come from bad connections - rarely the electronics themselves. In many years of driving, I have only once had a car quit on me through ECM electronics, and it restarted straight away and got me home, (albeit running at reduced performance). I took apart and sprayed all the plugs and sockets around the engine bay, and it was all back to normal, and hasn't missed a beat for 2 years since that happened.

Cars use a lot of plug and socket electrical connections because it makes the build process much quicker, but that also causes the majority of electrical problems down the line. (Gas turbine FADECS use gold plated connectors, which don't corrode and thus are more reliable). So, on your C150, you would not have plugs and sockets, but have screw/crimped/soldered terminals for every connection in the engine bay. You site the ECM in the cockpit, not outside with the engine, thereby protecting the electronics from extremes of temperature and vibration. And, yes, you make it dual channel, screened and fault tolerant. Cars already have a limp-home mode which keeps the engine running at reduced performance if important sensors malfunction. Or they use look-up values in the absence of sensor data. Not difficult, and doesn't have to be expensive.

.

[tdracer]

Uplinker, the difference is auto manufactures make them by the millions. Aviation makes them by the thousands - so massive economies of scale come into play. And that automotive engine that started at +40 C doesn't see minus 40C an hour later - aircraft engines do (granted, not as dramatic on piston engines as on jets, but still dramatic temp changes). That rapid thermal cycling is hell on electronics - cracking solder joints and solid state components. You can't just take an ECU out of your Honda, change the software, and expect it to survive on an aircraft - the operational environments are totally different.
Finding electronics that can survive that is expensive, and maintaining them is even worse. If the ECU goes bad in your car, most likely they throw it away and install a new one. But if the ECU costs $50k, you want to be able to fix it. Encasing all the components in some matrix to protect them from vibration (something that is commonly done on automotive electronics) makes them nearly impossible to repair.
Oh, and not many automotive electronics are protected against HIRF and Lightning. Aircraft have to be.

[Uplinker]

Agreed, but check the last paragraph of my previous post :ok:

[stilton]

Here, you're talking about two different things, capability and reliability.

My RV-9 is, in some ways, more advanced than the latest A350. If something goes out of limits, it won't just go "PING!" and require you to look at the ECAM to find out what went wrong. It will tell you, in your headset "Oil Pressure" or "Electrical Current" or any number of other things. IT has synthetic vision with terrain shading, a glide ring showing what points on the ground I can reach, based on current environment variables, can have touch-screen's installed if I want, displays a geo-referenced position on the visual or approach chart and is interfaced with an electronic circuit breaker system (though, granted the A350 has the last two as well). Another EFIS manufacturer provides for a camera input on which you can superimpose your PFD symbology, allowing a FLIR camera to be fitted for low-light operations.

That being said, while no light single will have autoland capability, I think many folk would be surprised to see the amount of redundancy built in to your typical IFR Experimental. Dual ADAHRS, triple screen (often with a third standby EFIS), dual GNSS and oftentimes dual batteries or alternators as well.


Actually there’s a few light singles now equipped and certified to do completely autonomous autolands

An ILS is not even required

[Machdiamond]

...That being said, while no light single will have autoland capability, ...

Not exactly autoland but Garmin Autonomi can take over in case of crew LOC (Helios 522) and land all by itself to the nearest suitable airport, even announcing intentions to ATC and reassuring passengers.

Already certified on several GA single such as the TBM, Cirrus, Piper,...

Garmin Autonomí: Autoland Activation

[stilton]

Not exactly autoland but Garmin Autonomi can take over in case of crew LOC (Helios 522) and land all by itself to the nearest suitable airport, even announcing intentions to ATC and reassuring passengers.

Already certified on several GA single such as the TBM, Cirrus, Piper,...

Garmin Autonomí: Autoland Activation (https://www.youtube.com/watch?v=IyYxbiZ1FCQ)


It lands the aircraft automatically therefore it is autoland


In fact, more so than traditional airline transport aircraft that require the crew to arm the system