Re: non man rated booster compared to shuttle?

From: Peter Stickney (peter_at_adelphia.net)
Date: 08/19/04 

Date: Wed, 18 Aug 2004 22:30:55 -0400

In article <herb.schaltegger-F7926C.18272718082004@news.newsguy.com>,
        Herb Schaltegger <herb.schaltegger@gmail.com.invalid> writes:
> In article <74f0gc-0ff.ln@Mineshaft.local>,
> peter@adelphia.net (Peter Stickney) wrote:
>
>> As forthe other
>> folks, let's jet say that I'm not overly thrilled with the idea of
>> flying in an FBW Airbus. The human-airplane interface is, to put it
>> kindly, unnatural, and a good chunk of that appears to be a basic
>> philosophy that the airplane should be flying the pilots, rather than
>> the other way 'round. I can understand that decision if you're
>> planning for you sole customers to be ex-colonial airlines with
>> low-time, inexperienced pilots trained from scratch for your airplane,
>> but to my mind it's a flawed world view. (So much for my ranting - if
>> you're a Flight Controls Guy for Airbus Industrie, don't take it too
>> personally. Human/Machine interface happens to be one of _my_ hot
>> buttons).
>
> Interesting perspective, Peter. One of the other dad's connected with
> my son's baseball teams is a captain for United and a former Navy S-3
> pilot. He's currently type-rated on the 757, 767 and Triple-7 and the
> later-model Airbuses. He just took a slot flying the Airbuses
> domestically after about two years flying them on U.S/London and
> U.S./Peking routes. In his opinion, the Airbus is much more comfortable
> and easy-to-fly plane than the Boeings; I wonder how much of that is the
> "plane flying the pilot" versus "pilot flying the plane" aspect you
> mention.

One of my neighbors transitioned last year from 737s tp A320s. (Then
he got furloughed - now he's flying a Citation X all over the world,
and loving it - with the Concorde out of business, he's in the fastest
Normal Category jet in the World) He wasn't at all happy with the
change. It's not that an Airbus is more automated - on pretty much
any airliner these days, the Flight Director is flying the airplane once
the gear's up. The Airbus's systems and interfaces aren't very good
at telling you what the airplane's doing. FOr example, the throttles
don't move - You have 4 detents (Ground, Cruise, Takeoff, and one
other, Flight Idle, I expect) and that's it. If the airplane's making
power setting changes, you don't see/feel it. (The original story was
that Airbus just wanted a button that said "Start", and a buttan for
"Stop".) For a pilot, especially one shooting an IFR approach and
trying to scan the flight instruments, or peering through the
windshield for the runway. The FBW system doesn't tell you when
you're approaching an AoA or G limit - it just stops at the limit, no
matter how hard you pull. The Boeing systems provide better feedback
to the crew - there's a stick resistance change as limits are
approached, and the controls behave like those of a non
computer-driven airplane. Boeing used input and direction from their
pilots, and their customer's pilots, in designing the FBW systems of
their jets. Airbus didn't.
The Airbus programming interface is not as "Pilot oriented" as that of
the Boeings, and it's possible to select modes different than what you
think they should be. When you get to a point where you're making a
decision, the airplane (Well, the computers, really) makes its own
decision. Usually it agrees with yours, but sometimes it doesn't. One
example is the A320 that flew into the trees at an airshow in europe.
The Pilot was certain that he had in in an auto go-round mode, and was
waiting for the airplane to fly itself away as it reached close
proximity to the ground. The airplane thought it was in some other
mode, and, while the two sat there staring at each other, it mushed
into the trees. (That one, btw, was particularly acrimonious with
Airbus and the French Government pulling all manner of tricks to keep
the airplane from being blamed, such as releasing an investigation
report before actually investigating, and trying to get the Captain
comitted to a Mental Hospital.)
Or the Indian airbus that found a cumulogranite cloud when descent
rate was set instead of descent angle. Or the Romanian Airbus arguing
with the pilot via 45 degree pitch excursions over Paris.

I will grant that an Airbus is a more comforatable airplane. Boeing
hasn't made a big cockpit since the B.377 Stratocruiser just after WW
2, and if you're in one of the sidestick Airbuses, you've got plenty
of room to open your breifcase, eat you lunch, or peruse your
investent prospectuses. But for me, I'd trade that comfort (And I'm
6'4") for knowing what to expect from the airplane.

The statistics say that an Airbus isn't particularly less safe than
any other airliner - they're all pretty damned safe these days - but
I'd feel more comfortable if the people designing the systems had more
of an interest in listening to the people who sit in the pointy part
of the airplane.
Way back in the '80s, when I was working of Nuclear Medical systems,
we had a couple of things happen that really matured my perspective on
human/machine interfaces. A cancer treatment system with 2 operating
modes, (X-Ray and Electron Beam) had a glitch in its menu-driven
interface that allowed the Electron Beam mode to be selected with
X-Ray excitation levels. This gave electron beam strengths normally
associated with putting together bridges, and the machine ended up
killing a couple of patients. (Another example of bad design all
around, I'm afraid. The configuration was so obviously wrong that
there should have been a hardware inhibit that wouldn't allow a
configuration like that to function.) Another was an observation from
one of our Test Pilots - a Brain Surgeon at Harvard Med:
"The only thing worse than not detecting a tumor is detecting one that
isn't there. 'Cause we're going to go in and look, and once we've done
that, the patient's going to be messed up anyway."
Oh, yeah, and our discovery that due to some quirks in RT-11's overlay
handler, our system could jump its tracks and happily execute its data
for quite some time. (And given the small size of a PDP-11's
instruction set, that's quite a feat. Back then, all computers were
RISC compunters, at least in today's terms. I shudder to think what
an opcode-heavy machine would do in the same situation.) That taught
me that if you're going to have an unhandleable failure, fail hard and
let somebody know. (Note that the Airane V that started this thread
wasn't as Henry points out, unhandlable, just poorly handled.). 

-- 
Pete Stickney
 A strong conviction that something must be done is the parent of many
 bad measures.  -- Daniel Webster