Techno

Guest Discussion
1

Brian Gruis raised an interesting point several days ago that demands more attention. He questioned the presumption that more advanced avionics means easier accessibility and safer flight. He suggested that increased avionics system complexity will require a more intense training commitment on the part of new GA pilots.

Several related questions come to mind.

 *In GA aircraft, will more sophisticated systems actually increase accident rates as pilots become emboldened by the technology and take on risks inappropriate for their level of skill and training?

 *Safe flying is more about pilot judgement than physical ability. Will new pilots develop reasoned judgement if the perception of risk is minimized?

 *Will pilots understand the new systems thoroughly enough to be able to handle cockpit distractions and still control the aircraft?

If you think these questions absurd, you’d better read on. Commercial aviators have been through this already.

The following excerpt is from the NTSB report # AAR-73-14 and is available in complete form by contacting NTIS. The report covers the investigation of EAL Flight 401 from NY to Miami on Dec 29th 1972. On approach to MIA, the L-1011 flight crew became preoccupied with a malfunction of the nose landing gear position indicator light. Following a missed approach, the flight crew engaged the autopilot to maintain 2000 feet msl and continued to trouble shoot the malfunctioning light. The flight crew failed to detect an unexpected descent caused by an inadvertent disengagement of the altitude hold function. The aircraft hit the ground, killing the flight crew and 96 passengers. The Captain had 29,700 hours total time (TT) and 280 hours in type (IT). The FO had 5,800 hours TT and 306 hours IT. The SO had 15,700 hours TT and 53 hours IT. The L-1011 was newly delivered 3 months before the accident and had 986 hours and 502 landings.

The report states:

“Another problem concerns the new automatic systems which are coming into service with newer aircraft and being added to older aircraft. Flight crews become more reliant upon the functioning of sophisticated avionics systems, and their associated automation, to fly the airplane. This is increasingly so as the reliability of such equipment improves. Basic control of the aircraft and supervision of the flightÂ’s progress by instrument indications diminish as other more pressing tasks in the cockpit attract attention because of the over-reliance on such automatic equipment.”

2

Like all things, there’s a balance to be struck. However, I think you’re doing an apples-to-oranges comparison here. The Cirrus, for all of its electronic wizardry, is no more automatic than any plane with a two-axis autopilot that’s been available for forty years.

Autopilot usage still seems to get people worked up one way or the other. When I started my twin training in a well-equipped Baron, I said that I wanted to split my training time 50/50 between hand-flying and autopilot use, even though it would take longer. My feeling is that both of them have to be intuitive in order for me to be safe. If things go to hell in the proverbial handbasket, I want to be able to quickly and confidently punch in the autopilot to help me out,

and conversely if the autopilot quits on me, I’d better be able to hand-fly the thing. There seems to be ample evidence that insufficient proficiency on one side or the other is unsafe (I hate to bring it up, but JFKjr comes to mind, on both counts.)

The L1011 incident is a red herring, I think–to me this is purely an issue of poor CRM and a fixation on a non-critical item to the exclusion of everything else.

I am a firm believer, however, that the use of graphical devices in the cockpit will reduce the IFR death toll. There is an absurd number of people who die flying their airplanes into obstacles or the ground, presumably after losing track as to where they are and where they are going. (A good autopilot would probably save a bunch of the folks that simply lose it in the clouds, even when rated and current–hopefully they would live long enough to realize they need more dual, and go back and get proficient again.)

This stuff is astoundingly complex, however, and we run headlong into the training issue–there simply are not enough instructors familiar with modern avionics. There are few things more frightening than being pounded in the clouds, pushing a button, having the plane do something unexpected, and having absolutely no idea what is going on (or what the best way of saving yourself would be.)

Personal anecdote–not long after I got my SR20 (but long enough to feel smug about operating it) I was motoring along in IMC near San Jose CA and managed to get close to a small convective cell (no forecast, no pirep, and I was in California, so what would I need my Stormscope on for? Duh.) A discharge nearby caused my ELT to activate, my autopilot to disconnect, and my ARNAV to lock up. In a fraction of a second I went from relaxed and happy to wetting myself. The a/p disconnect is honking, the ELT is squealing through the radio, and I suddenly realize that it’s my turn to fly the plane. Happily I had read up carefully on the systems and, after making sure I was right side up and counting to ten to slow myself down a bit, I managed to get everything reengaged. It could have been much uglier if I was less prepared.

I’m just rambling here, but I think the gist of it is that you can be safer with all this wizardry if you are trained and proficient in both its use and its non-use. If you’re not up to the task, it can be distracting and more dangerous.

3

Statistics say piloting a piston aircraft is anywhere from 5 to 10 times more dangerous than driving.

But when you factor out: flying into bad weather, which can happen, as Dave says, but still is mostly within one’s control; running a tank dry; low-altitude maneuvers (i.e., buzzing and other idiot stunts); and flying while drunk, the odds improve to 2 times more dangerous than driving.

Just to be slightly smart-assed (because there is little else to do on a wet dreary Sunday) I’d venture to guess that Garmins and Arnavs improve safety because they appeal to nerds and anal retentives — i.e., we who visit this site regularly — who are less likely in the first place to fly into clouds, buzz houses, run a tank dry, and fly drunk.

Except for Jeff.

Brian Gruis raised an interesting point several days ago that demands more attention. He questioned the presumption that more advanced avionics means easier accessibility and safer flight. He suggested that increased avionics system complexity will require a more intense training commitment on the part of new GA pilots.

Several related questions come to mind.

*In GA aircraft, will more sophisticated systems actually increase accident rates as pilots become emboldened by the technology and take on risks inappropriate for their level of skill and training?

*Safe flying is more about pilot judgement than physical ability. Will new pilots develop reasoned judgement if the perception of risk is minimized?

*Will pilots understand the new systems thoroughly enough to be able to handle cockpit distractions and still control the aircraft?

If you think these questions absurd, you’d better read on. Commercial aviators have been through this already.

The following excerpt is from the NTSB report # AAR-73-14 and is available in complete form by contacting NTIS. The report covers the investigation of EAL Flight 401 from NY to Miami on Dec 29th 1972. On approach to MIA, the L-1011 flight crew became preoccupied with a malfunction of the nose landing gear position indicator light. Following a missed approach, the flight crew engaged the autopilot to maintain 2000 feet msl and continued to trouble shoot the malfunctioning light. The flight crew failed to detect an unexpected descent caused by an inadvertent disengagement of the altitude hold function. The aircraft hit the ground, killing the flight crew and 96 passengers. The Captain had 29,700 hours total time (TT) and 280 hours in type (IT). The FO had 5,800 hours TT and 306 hours IT. The SO had 15,700 hours TT and 53 hours IT. The L-1011 was newly delivered 3 months before the accident and had 986 hours and 502 landings.

The report states:

“Another problem concerns the new automatic systems which are coming into service with newer aircraft and being added to older aircraft. Flight crews become more reliant upon the functioning of sophisticated avionics systems, and their associated automation, to fly the airplane. This is increasingly so as the reliability of such equipment improves. Basic control of the aircraft and supervision of the flightÂ’s progress by instrument indications diminish as other more pressing tasks in the cockpit attract attention because of the over-reliance on such automatic equipment.”

4

Actually, the conclusions of the NTSB report were far more comprehensive than I am able to include on this message board ( I lack adequate typing skills to include the whole text here). CRM is an obvious problem, but the function of the new autopilot system was extensively discussed in the NTSB report. The system was able to be engaged from either seat, had different override pressures depending on which control column was moved, had different cockpit ennuciation lights depending on which seat initiated the autopilot, and was not well understood by any EAL pilot when in the CWS mode below 2,500 feet. The NTSB felt there was more to the accident then CRM.

*To the point: The use of advanced avionics has been cited by members of this message board as a way to bring GA to the masses. Your personal experience is an example of why it probably won’t.

5

I simply don’t see any logical reason to assume that more capable and complex avionics will lead to such a problem in the long run. Certainly, there will always be exceptions where a pilot will get confused, and an inability to deal with his/her equipment will not help, and may harm there situatuion.

However, the ability to SEE an IFR approach depicted in boxes on a screen such as a Sierra Flighy Systems, or to follow a highway in the sky through the rockies in IMC will undoubtedly reduce the number of accidents in the long run.

Unfamiliarity with equipment is a pilot’s fault. Let’s not start saying it’s the autopilot’s fault when the autopilot is disconnected and the plane crashes. Also, the 1972 EASTERN example you cite, there was NO aural indication that the AP has disconnected. This was a design flaw that was corrected. In Dave’s example above note how he heard the autopilot screaming. Also note how he had the stormscope turned off, otherwise he might not have even got into that situation in the first place. And let’s not forget, he was spooked, but in control at all times.

In the end, capability is merely that. It is an addition, an aid, a potential to make flight safer. There will always be a few loons who dive into Thuderstorms because they have Stormscopes and figure they can slip through it, yes. But I firmly believe that for every 1 loon, there will be 50 pilots who see their Stormscope light up and decide not to even go near that bunch of clouds that they would have flown right into without their equipment to help them.

Dean

6

Actually, the conclusions of the NTSB report were far more comprehensive than I am able to include on this message board ( I lack adequate typing skills to include the whole text here). CRM is an obvious problem, but the function of the new autopilot system was extensively discussed in the NTSB report. The system was able to be engaged from either seat, had different override pressures depending on which control column was moved, had different cockpit ennuciation lights depending on which seat initiated the autopilot, and was not well understood by any EAL pilot when in the CWS mode below 2,500 feet. The NTSB felt there was more to the accident then CRM.

Sounds like a seriously hosed user interface. The world is filled with anecdotes of bad UI, sometimes with disastrous consequences. Doesn’t take high-tech–the old Beech reversal of the throttle quadrant and the flap and gear controls is a decidedly low-tech example. High-tech provides even greater challenges, as the inherent complexity is much higher, so coming up with a way for most users to be able to intuitively use the equipment in the majority of situations is a bigger challenge than just building the technology.

Positive example–I bought a reasonably nice digital camera yesterday. It has a great deal of flexibility as far as manual exposure and focus control, white balance, capture resolution, etc. However, if you just pick it up, turn it on, and shoot, it looks just like a generic point-and-click, with only a few, familiar, buttons. The majority of users will probably never venture beyond this, but the defaults are good enough for both usability and product differentiation. Get into the menus, and it’s a different world for the Power User.

Negative example–I have polybrominated biphenyls (PBB), a known carcinogen, flowing in my bloodstream. I grew up in Michigan in the 60s and 70s. Some chemical company made, among other products, a cattle feed additive and a fire retardant. The cattle feed additive had the trade name of “Feedmaster.” The fire retardant had a trade name of “Firemaster.” One night, a somewhat illiterate factory worker got the bags mixed up, and put a bunch of Firemaster into Feedmaster bags, which got shipped out, mixed with cattle feed, and ended up in the milk that we were all drinking. Exceedingly low tech.

*To the point: The use of advanced avionics has been cited by members of this message board as a way to bring GA to the masses. Your personal experience is an example of why it probably won’t.

Not in its current form, at any rate. The autopilot is basically always looking for an excuse to disconnect, because the designers decided that it would be much preferable to disconnect and let you hand-fly than it would be to risk some kind of uncommanded action. The autopilot is not considered a survivability item. Further, I found out from inquiry that the first GNS430 (and maybe the transponder, I can’t remember) has extra lightning shielding on the antenna leads and wiring in order to pass part 23 requirements, whereas the rest just has the “normal” shielding (which has to be better than what’s in my aging 172).

For this to be anywhere near flyable by “the masses,” all of this has to be elevated to “must not fail” status, which may or may not be achievable technologically and economically.