AVweb

« Back to Full Story

Fly-by-Wire Fantasies

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

Here's an astonishing fact: The Air Force recently announced that the B-52H, of which 74 are still in service, will remain in the active fleet until 2045. By then, the design will be 93 years old. Lots of technologies—steam propulsion, the internal combustion engine, stick-built houses, the air brake—have lasted longer than that, but I can't think of a single airframe that betters that. The DC-3 is close, but how likely is to be in meaningful commercial service in 2028, on its 93rd?

Push the playhead forward into the imaginary future and focus it on aircraft subsystems, especially primary controls. Then ask yourself if 90 years from now, primary controls will still be mostly direct cables or if, as Diamond Aircraft now seems to suggest, they'll be fly by wire. It's not quite a chicken-and-egg question, but more like if it ain't broke, why fix it? (The B-52H is loaded with cutting edge avionics, but it still has the same control system it left the factory with a half century ago.)

Technologies persist for several reasons. One is performance. If something works well and its replacement does no better, inertia rules and it endures for decades. Cost counts, too. Some electronic systems are cheaper than the mechanical ones they replace or cheaper than the previous generation of electronics and have greater capability than what went before. That's a split decision for aircraft avionics; for upgrades, they're generally more expensive as a percentage of airframe value than legacy systems, but they definitely do a lot more. It's debatable if "a lot more" means you can fly in weather you couldn't before, the airplane goes further and faster and/or is just safer. Essentially, modern avionics mean you have better situational awareness because you have more information. So far, we have aptly demonstrated that this hasn't moved the accident rate downward. In avionics, we've certainly proved an affinity for bright shiny gadgets.

In aviation, at least general aviation, markets and makers tend toward the conservative. The market has not rewarded innovative airframers or engine designs. (Before you say what about Cirrus, I'd say it's incremental technology, not revolutionary and I'd raise you one Beech Starship in the process.)

Consider the new Rotax 912 iS, a nice little incremental improvement on the 912. But it's hardly a technological tour de force. It doesn't yet have direct injection and it took us all the way to 2012 to get rid of carburetors. These decisions by Rotax were made for sound technical and economic reasons, but they're driven by buyers whose idea of a walk on the wild side is a Cherry Coke.

One reason some perfectly adequate technologies don't persist is that they're overwhelmed by the macro economics of developments in related or parallel fields. In case you haven't noticed, electrical controls and motors of all sorts are replacing mechanical controls in cars. Throttle by wire is already common on motorcycles and cars. Next year, Nissan will offer steering by wire on the Infiniti. Goodbye tie rods, steering racks and power steering pumps. Dynamic, electronically controlled suspension systems are commercially viable. On the drawing board are electric brakes—even for motorcycles—and electric valve actuators for cars, eliminating the camshaft and valve train entirely. These aren't widely economically established yet, but they aren't just wild-eyed concepts, either. Twenty years from now, the world might be awash in cheap digital actuators and servos, just as it is now awash in cheap MEMS gyros made possible by huge volume in automotive braking, sensing and navigation systems. That's why Dynon can make a $1500 battery-operated EFIS.

And to me, that's the big unknown about Diamond's fly-by-wire project. On the one hand, cable-controlled primary surfaces work well, they're light and cheap. With cable or rod-controlled surfaces, you feel the air. Feedback is part of the tactile experience of flying. But is an important part? That steer-by-wire Infiniti will numb the driver to how the tires feel when they're about rattling down a washboard road or perhaps at the slippery edge of hydroplaning.

So fly by wire may not be about feel at all. At Diamond, CEO Christian Dries hopes it will bring along features like autoland and higher performance airframes which can't be achieved without what will eventually become automatic flight control systems. It seems so tantalizing and so obvious, but given the industry's preference for sticking with what works, I don't think anyone has a clue which way it will go. It will pit aviation's hidebound conservatism against harsh economic reality.

I suspect it's rather like Boeing turning out the last of the B-52H's in 1962. They could not have then imagined that those old control wheels would eventually be horsed around by pilots whose great grandfathers laid hands on the very same airplanes. Nor could they have foreseen the old warhorses upgrading through four or five generations of avionics that might, for all we know, result in B-52s flown to the boneyard via fly by wire. Not sure I'd bet against it.

Comments (39)

From what I read (non-pilot, so be forewarned) there is genuine revolution in digital-assisted pilotage -- from voice warning systems "Your rudder is not coordinated" or "Traffic at 2 o'clock" to complete auto-takeoff and auto-land.

If I have it right that _should_ create much much safer flying -- I am talking Private Pilot and LSA -- and thus a vastly increased number of new pilots. Fulfilling (sorta) the expectations of 1948 of a plane in every garage.

And such digital enhancement can plug into legacy planes. Servos in a C150 may not make sense but a CFI in an iPad -- able to monitor/warn the 50 hour per year pilot -- may be very cost-effective.

Am I mis-reading the emerging techno revolution?

Posted by: David Sucher | December 15, 2012 2:17 PM    Report this comment

Fly-by-wire rests on an absolutely uninterruptable power supply for both the control part and the actuators. Providing double redundancy in an airliner is one thing, in a light aircraft you would quickly start to wonder how far into the thicket you want to chase the rabbit.

In a car loss of throttle control isn’t a huge deal, but vehicle steering calls for the same belt & suspenders approach as fly-by-wire; one wonders how Nissan plans to handle this?

Posted by: John Wilson | December 16, 2012 3:02 PM    Report this comment

Why not ultra capacitors? The power has to last only long enough to pull over after the vehicle warns of a power loss. Airplanes are a little different, but Diamond thinks it can certify to Part 25 standards with current power designs. (They can accommodate up to six alternators and three batteries.)

Lightning cert could be a challenge.

Posted by: Paul Bertorelli | December 16, 2012 4:26 PM    Report this comment

Supercapacitors ( sometimes called ultra capacitors) can deliver a large amount of energy quickly better than batteries. But they only store about a third of the energy (per pound) of lithium batteries. I got this from: Energy For Future Presidents, the science behind the headlines by Richard A. Muller 2012 I recommend this book. Covers everything from global warming to electric cars.

Posted by: Bill Berson | December 16, 2012 4:46 PM    Report this comment

Talk about over engineering! Sure you could do it but at what cost? How about the additional weight of all that "redundant" power. A fighter or a multi-million(or billion) dollar airplane is one thing but a C172/182/206 class of aircraft with fly-by-wire controls, no thanks. How much more would that kind of aircraft cost versus new aircraft and what about repair costs? Remember this class of airplane only has a 500-1200lb payload. And then there is getting it certified by the FAA.

Posted by: matthew wagner | December 16, 2012 5:56 PM    Report this comment

Perhaps a failure of imagination here. In 1994, I wrote quite a convincing little essay on why glass panels wouldn't find their way into light GA aircraft because the technology could never be reduced in cost to be competitive and even if it could, no one could ever get the things through FAA certification. And even if that were possible, they were "over engineered" and provided no benefits over steam gauge panels. No one would trust them enough to buy them.

There were probably a half dozen other reasons why glass would never work in small airplanes. And we all know what happened.

Naysayers and visionaries tend to make the same mistakes, but for different reasons. Many under estimate or don't understand how rapidly some technologies evolve and how it spills over into all market segments. That's the example of the MEMS gyro. Others assume that technology will advance much more rapidly than is remotely possible, to the point of defying the laws of physics. That's the electric car argument.

It's all but impossible to know which is which. I've proven that I certainly don't.

Posted by: Paul Bertorelli | December 17, 2012 5:22 AM    Report this comment

Steam gauges worked. Now we have glass. Paper charts and sectionals worked. Now we have tablets. Pulleys and bell cranks worked. Now we'll have fly-by-wire.

It's called technology and progress. And we have all adopted it.

Posted by: Amy Zucco | December 17, 2012 8:54 AM    Report this comment

There is one difference with FBW vs glass-panel, and that is the potential for it to keep a pilot from doing something he/she wants to do, because it is outside of the programmed safety envelope. I'm not sure if this is a good thing or a bad thing (as plenty of pilots have proven they don't really understand how to control an aircraft and stall/spin themselves into a crater), but it is something no glass panel can do.

Posted by: Gary Baluha | December 17, 2012 8:57 AM    Report this comment

As an owner of a 'new' piece of technology that became obsolete and unsupported a mere 7 years after I bought it (MX-20) I gotta say I'm leery of FBW in GA. Whatever benefit the FBW enables will clearly have to outweigh the cost of maintenance over the life of the airframe. And if the parts in the FBW 'brains' become unavailable, that lifetime may be plenty short, as little as 10 years if it's not properly designed for it. (Look at AVConsumer's own recent article on the higher cost of glass PFDs. for as 'dumb' as the vacuum gyros are, they're still cheap and you can still get parts for them.)

Look at another recent example of a 'new' technology: the Mooney PFM. Electronic engine controls, all the whiz-bangs. But it needed dual alternators, dual batteries, the reduction gearbox, the cooling fan, the composite prop and more wetted area drag from an extended fuselage plug to get the CG back in the right spot. It cost 66% more than the Lycoming 4-banger and flew 4 to 5 kt SLOWER. Oh wow, it started quicker. Big deal. The benefit wasn't there.

So when the real game changer shows up, I may be interested, (but I'll wait for someone else to prove it's reliable, first.)

Posted by: David Rosing | December 17, 2012 9:23 AM    Report this comment

The promise is a huge increase in safety of GA flight and a similar reduction in the skills required to fly a GA plane. Absolutely revolutionary. The time frame for that? Who knows, but it will come and probably with some mistakes along the way. But it's a "when" question, not an "if" question.

Posted by: Malcolm Ruthven | December 17, 2012 9:34 AM    Report this comment

At first, GA FBW paired with an authoritative autopilot will comprise an "electronic parachute" - something I've advocated for Cirrus' SF-50 since before they even announced that vehicle. But the inevitable application is pilotless GA aircraft, which will be the salvation of GA.

Posted by: Tom Yarsley | December 17, 2012 9:42 AM    Report this comment

The very thing about the B-52 is that it is old tech that is still viable. GA planes are typically only supported well for a few years. An old Piper Comanche 180 is a fantastic airframe, but try to find a gear retract motor for it. They stopped making them in 1974 and getting an field approval for an equivalent motor is very, very hard. Makes you like the Mooney M20B with the Johnson bar better. The owner and mechanic can fabricate just about every piece. Good luck on maintaining a custom FBW system 20 years out.

Posted by: Mark Naper | December 17, 2012 9:43 AM    Report this comment

Oh, my bad. M20C. The M20B had the wooden wing....

Posted by: Mark Naper | December 17, 2012 9:52 AM    Report this comment

Oh, my bad. M20C. The M20B had the wooden wing....

Posted by: Mark Naper | December 17, 2012 9:52 AM    Report this comment

David, isn't it odd that you have to future-proof future technology? But it's true. You can almost imagine an STC to replace FBW with cables and rods because the parts are no longer made.

This may happen with the first round of EFIS, eventually. I just ash-canned a 12-year-old LightSpeed headset for the very same reason.

In consumer electronics, you don't have this problem. You'll have your iPhone or Droid for 18 months then, even though it's perfectly serviceable, you'll replace it with something new.

Let's hear it for progress.

Posted by: Paul Bertorelli | December 17, 2012 10:42 AM    Report this comment

Paul,

Ah, you get my point.

This is where I'd love to see Aviation Consumer raise the standard to protect the unsuspecting buyers of the future. In my MX-20 example and your headset example, the units had become obsolete and unrepairable 7 to 12 years after they were sold. If anything, any airframe made with parts that couldn't be supported after that period would mean the airframe would become worthless and, therefore, a lousy investment if it could only be used as a glorified paperweight a mere 12 years after it was sold.

But there are ways to mitigate this. In electronics-land, a Field-Programmable Gate Array (FPGA) can be programmed to mimic just about any microprocessor chip. If the FBW electronics were built around the FPGA architecture then even when the FPGA first used to certify the design becomes obsolete, a follow-on model can take its place later on. Trouble is, you'd have to re-certify the new FPGA to ensure it emulates the original microprocessor. But hopefully that's easier and cheaper than having to completely requalify a different microprocessor and printed wiring board.

Of course, the initial cost may be higher, so that goes back to the cost vs. benefit trade at the start of the program.

Those old cables and pulleys are looking more elegant than ever . . . .

Posted by: David Rosing | December 17, 2012 12:00 PM    Report this comment

David, we have explored this with Garmin, Avidyne and Bendix/King.

As you can well imagine, it is a balancing act between cost of manufacture, sales price, margin and front end investment in replaceable components.

When they build a box, they estimate a service life and stock replaceables accordingly. But they can't buy too many because that will eat into the product margin and maybe even make it unprofitable.

The problem is not just chips, but other components on the boards, plus mechanical things like knobs, buttons and switches. Bose ran into the same problem with the original ANR, which cost a grand.

Buyers thought it should last forever at that price. But it won't and it can't. Progress in component manufacture rapidly obsoletes the ability to support these devices.

I went outside avionics to get a better understanding of this to medical and industrial electronics and they told the same story. So I don't think it's realistic to expect forever performance out of these things. Question is, what is the right lifespan?

I'm sure I don't know. But customers need to know this going into the deal and that's what we told them in our reporting.

Posted by: Paul Bertorelli | December 17, 2012 12:37 PM    Report this comment

When the Whizair FBW1000 wired wonder pops it's last diode and gives the "Game Over" code as it slides down the recycling chute, maybe I can catch a ride home with Paul in his J-3.

Posted by: Richard Montague | December 17, 2012 12:56 PM    Report this comment

Happy to, Richard. By then I'll have my wind generator powered (used) G1000 installed.

But I'm sticking with the hydraulic control surfaces.

Posted by: Paul Bertorelli | December 17, 2012 3:11 PM    Report this comment

In the real (consumer) electronics world cost reduction to reduce price and increase profit is a way of life. As long as the product continues to function the same way - it is the same product. I suspect this is not the case with certified avionics. It should be. Cost reduction allows migration away from obsolete parts and suppliers, and usually reduces part count and weight while increasing reliability. It also means not having to stock a million ancient dinosaur parts long after they've gone extinct. I just had to have a sub-board on my KCF-150 controller replaced. If I could have junked the thing for a DFC-90 I would have in a heartbeat. Why not? Because it'll take another 2 years to certify my particular variation. Thanks FAA...

I, too, will stick with servos and hard-linked

Posted by: neil cormia | December 17, 2012 5:42 PM    Report this comment

... hard-linked controls. Sorry for the split post.

Posted by: neil cormia | December 17, 2012 5:43 PM    Report this comment

Here is the potential problem. Aircraft control cables are still manufactured to original specs. Ditto the bearings, pulleys, etc. to make a conventional control system work. Avionics, even certified avionics, rely on microprocessors (chips) that are developed and manufactured for consumer products. With Moore's law still in effect and processor speed doubling roughly every 18 months, the turnover in chips is very fast, and old obsolete chips are dropped from production.

So, no matter how gee whiz Diamond's fly by wire system is TODAY, it's a certainty that in five years the chips that provide the brains for that system will no longer be manufactured. Diamond will likely stockpile some parts, but how many, and for how long? The reason this is important is because you can't repair a microprocessor, so once that stock of parts runs out that gee whiz fly by wire system is going to be junk, and the pilot will have no option but to pay for an upgrade to whatever Diamond has that is new and shiny.

Lest you think this is just doom and gloom, I suggest you try to get repairs done on a Garmin 155, the first certified GPS.

Jay

Posted by: James Hulbert | December 18, 2012 7:37 AM    Report this comment

.....I suggest you try to get repairs done on a Garmin 155, the first certified GPS.

Or a Lightspeed headset more than 3 years out of production....they can't even supply a replacement cord. Now THAT irritated me.

On the other hand, most of the solid state stuff in the electronics is likely to outlast the owner.

Posted by: John Wilson | December 18, 2012 8:28 AM    Report this comment

I know it's not the main point here but we have to face reality regarding the B-52. A comparison to the DC-3 is interesting, if the B-52 were asked to turn a profit in commercial service, it would have found retirement long ago. Instead, it soldiers on emblematic of our incompetence in acquisition. A replacement bomber has never been freed of the expensive demands of leading edge technology, nor the inevitable death spiral of lower numbers and rising per unit cost (B-2). Like a pickup truck the Air Force can't afford to replace, the old bird is still parked on the active ramp. We are lucky it was built to last, it has to.

Posted by: Lance Johnson | December 18, 2012 6:28 PM    Report this comment

Innevitably during hangar sessions or flying long trips, one wishes for a lot of things for the airplane, more accurate fuel gages, more speed, more comfortable seats, less noise, better fuel economy, more speed, better avionics, better placed switches and knobs, more speed and the list seems endless. However, not once in ove forty years of flying have I heard a GA pilot, or any other pilot for that matter, say, "Damn, I wish I had fly-by-wire controls!"

Posted by: Richard Montague | December 19, 2012 8:23 AM    Report this comment

I recall a comment from an Air Force pilot some years ago, that when the B-2 bomber is eventually retired to the bone yard, the crew will be flown home in a B-52.

Posted by: Jonathan Spencer | December 19, 2012 8:43 AM    Report this comment

I'm on the "wait and see" fence regarding light GA FBW, but it does potentially mean we can get airframes that are less stable (by human standards) but offer better performance. The FBW system can emulate stable behavior to the pilot, who simply flies it like any other airplane. It has worked for the airlines and military jets. The question if it will work for light GA is a matter of weight, cost, and cost-to-performance.

Posted by: Gary Baluha | December 19, 2012 8:49 AM    Report this comment

Back in '63, I was at a USAF tech. school, learning the autopilot and compass systems on the B-52 and KC-135. Got to my first assignment in January the following year, and ended up working on B-47's and C-47's. When we transitioned to the -52 a year or so later, they brought a brand new item into our "bench stock" at the squadron. That item was a transistor. Never seen one before that. In one week, the price of that little part went from $198 down to $2.00, just because the technology costs had advanced so far, thus resulting in cheaper manufacturing costs. We have seen costs go down in such things as computers, but most of that has come about because of transfer of manufacturing costs to off-shore manufacturing, at much cheaper labor rates. Now, look what happened with the C-162 Skycatcher. Prices keep going up, even though most of that airplane is actually produced in China. And that is just an LSA "trainer." A lot of that cost, granted, is because of the "advanced" technology installed as standard instrumentation, but, a good percentage is also to be attributed to the initial development costs.

Posted by: Doyle Frost | December 19, 2012 8:50 AM    Report this comment

Anyone see the chicken/egg problem which has developed? We make airplanes that cost more and which are more complicated. So learning to fly them takes longer and is more complicated and therefore costs more. Maintaining them takes longer and is more complicated and therefore costs more. We change the system to be more complicated, so that even older, cheaper airframes must be upgraded with fancier more expensive avionics which don't make the airplane fly faster or more efficiently, just more expensively.

And then we wonder why there are fewer pilots who can afford to fly. In reality, we're taking careful aim and shooting ourselves in the foot, and then wondering why we're limping.

Posted by: Cary Alburn | December 19, 2012 9:12 AM    Report this comment

One of the things you must keep in mind is that the original reason FBW came into vogue is because of the increase performance that came from aircraft designed with relaxed static or negative stability. The ability to fly with an aircraft with RSS and keep it in control was only enabled by FBW technology, and with that came increased performance for military aircraft. While we have FBW systems on commercial aircraft, I don't believe there are any that are actually statically unstable like the F-16,F-18, F-22, or F-35.

Posted by: Richard Mutzman | December 19, 2012 6:47 PM    Report this comment

Might be relevant? http://www.aopa.org/aircraft/articles/2012/121219autoland-for-general-aviation.html

If this stuff really happens then I suggest that the long ago dream of mass flying (just after ww2) may actually happen. No?

Posted by: David Sucher | December 20, 2012 11:52 AM    Report this comment

Richard, no they're not unstable, they just plow into the ocean or ground like AF 447 and China Airlines 140.

D'oh!

Posted by: David Rosing | December 20, 2012 12:14 PM    Report this comment

The crash of AF 447 isn't a result of fly-by-wire, but rather by a complete lack of understanding of basic aerodynamics by the flight crew. They flew that plane into a stall, all the way from FL390.

What it does highlight about FBW, however, is that pilots can become complacent and just assume the system will always keep the aircraft within the normal flight envelope. But that assumes the system is working fully, and in this case it wasn't. On the other hand, pilots of non-FBW aircraft still fly their craft into stall/spin accidents, so it still comes back down to basic hand-flying skills (even if "hand flying" means having the inputs interpreted by a computer).

Posted by: Gary Baluha | December 20, 2012 12:26 PM    Report this comment

Meant to include this link, which seems pretty amazing:

http://www.aopa.org/aircraft/articles/2012/121219autoland-for-general-aviation.html

Posted by: David Sucher | December 20, 2012 1:30 PM    Report this comment

Gary,

But isn't "a complete lack of understanding of basic aerodynamics by the flight crew" a potential side-effect of FBW systems?

Posted by: David Rosing | December 20, 2012 2:09 PM    Report this comment

There's always the Mulhouse crash, which could be fairly described as "fly-by-wire assisted."

Captain Asseline also reported that the engines didn't respond to his throttle input as he attempted to increase power and that the elevators pushed the nose down.

"During a re-creation of the flight at the Airbus facility at Toulouse, it was established that the configuration the crash aircraft was in, i.e. nose high, low speed, low engine power; the fly-by wire system did push the elevators down to keep the nose down to prevent the onset of a stall."

That's the one where they flew into the trees.

Posted by: Paul Bertorelli | December 20, 2012 3:30 PM    Report this comment

David,

It is indeed (which I meant to be more clear about). However, it seems pilots are good at lacking this knowledge even with conventional control systems.

Posted by: Gary Baluha | December 20, 2012 3:42 PM    Report this comment

"But isn't 'a complete lack of understanding of basic aerodynamics by the flight crew' a potential side-effect of FBW systems?"

Not at all. FBW is just one means of causing the movement of flight-control surfaces. There are many ways to do that, with or without the aid of any autopilot functionality at all.

This should not be conflated with the demonstrably deadly combination of pilots' reliance upon flight automation, and their lack of familiarity - or at least lack of proficiency with - basic flight operations like stall-recovery.

The biggest complication - and the biggest flaw - in so-called "flight-automation" systems, is that they are designed to have human pilots "in the loop" in real time. The only justification for this is human ego, but that's enough to buy a horse in lieu of an automobile.

Posted by: Tom Yarsley | December 21, 2012 7:28 AM    Report this comment

Being in the loop is the whole purpose for me. When I stomp full right rudder, slap in left aileron and full up elevator, I expect my plane to eagerly respond "OK, Stupid, here we go again!" I do not want it to say "sorry, that is not allowed, would you like some tea instead?"

Posted by: Richard Montague | December 21, 2012 11:03 AM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?

Register

Enter your information below to begin your FREE registration

« Back to Full Story