"Die Zeit" article on Airbus (longish excerpts)

From:         hrz090@aixrs2.hrz.uni-essen.de (Dr. Erdelen)
Date:         30 Oct 94 21:29:41 
Followups:    1 2
Next article
View raw article
  or MIME structure

Hello, airlinerers & airlineresses,	

in its issue 43 (21 Oct. 1994), "Die Zeit", a major German weekly,
carries an article on the Airbus crashes of recent times. Here's
an excerpt. See end of message for motivation and comment.
[Stuff in brackets is mine. Sorry for probable partial unidiomacy.]

The accidents at the base of the report will be well known to readers
of "airliners" (A300 at Nagoya, Japan, in April; A330 test flight near
Toulouse in June; A300 near-crash at Cheju Island, Korea; Tarom A310
near-crash; Mulhouse 1988 airshow A320 crash; Bangalore 1992 A320
crash; Strasbourg 1992 A320 crash).

More interesting is that the article does not describe glass cockpit /
fly-by-wire as inherently dangerous, but rather emphasizes problems of
human perception of and reaction to computerization, as well as
insufficient training procedures: "All pilots had blindly trusted
their machines." Lufthansa Check Captain Georg Keil is quoted as
stating that the cause of the accidents is not the computer system but
rather the attitude towards technics in general: "It starts with the
question if one should sell planes as fool proof systems, or as
systems which require diligent operation."
The (wo)man/machine interface is seen at the base of the errors.
Keil: "This is the fatal consequence of the misconception that one can
reduce training."

The report states that Lufthansa trains its pilots in-house, and not
at Airbus in Toulouse: "At Toulouse, it is said, there are trainers
who have never flown an airliner with actual passengers, whereas at
Lufthansa, experienced line captains handle the training. They know
all problems which occur in line operations, and also the practical
consequences of design errors in newly introduced airliner types.
Flight trainers without line experience know these only in theory."

Lufthansa is further reported to emphasize human factors during
training. Blind trust in technics is discouraged. "Instead, the
so-called crew resource management (CRM) centers on stress-free
communication between captain, co-pilot, and computer."

Airbus is said to have by now recognized that "pilot training has to
be adapted to the participants' needs, and not vice versa. They, too,
now offer a seminar on communication and teamwork in the cockpit."

The article goes on to quote Christian Kepp, Lufthansa captain since
27 years, member of the Workgroup "Accident Analysis" of Cockpit
[union style association of German airline pilots]. He summarized his
experiences with the Airbus in a paper sent to the Luftfahrtbundesamt
[Federal Aviation Office; German equivalent of the FAA], and  
expresses severe critique. Quote from that paper: "If something goes
wrong, it is [allegedly] always the pilot's fault if the technics he
has to handle turns out to have worked as defined. [However,] The question how
good or bad the technics has been adapted to his needs, and if he
could at all have influenced it in the situation, is then not
even asked."

Cockpit demands a partial reversal of [backstep from] computer use from 
Airbus: "No system indispensable for flight operation (such as engine or brake
control) must be controlled by automatics which the pilot cannot influence."

Kepp also criticizes the glass cockpit: "Man-machine communication
mainly uses the optical channel, hardly the acoustical, and not at all
the tactile (after the introduction of the side-stick). [...] In
complex situations, the optical channel breaks down incredibly early."

The article mentions that the Boeing 777 will have a glass cockpit and
fly-by-wire, but will revert to a traditional style "stick" (yoke)
with computer-generated "feel of resistance". (It also says that even
in pre-computer times, "stick resistance" was not real but rather
simulated by the hydraulic system.)

The "Zeit" report is complemented by an interview with Bernard Ziegler,
Vice President [??, or some such; I'm unfamiliar with business ranks]
of Technics at Airbus Industries. Ziegler denies that design errors
have been involved in the crashes, and says that the problems between
computer and human are "not specific for the computer, but a problem
of the pilot who has to understand how a plane flies. We are not born
to fly as birds are. There is a huge difference between the instinct
of man and of birds. Therefore, there will always be misunderstandings
between man and the flying machine. Three crashes during approach (two
at Katmandu, and one at Strasbourg, with different Airbus types)
basically had the same reason, among others that man has no inborne
instinct for height, for distance from ground.

"Q: Do you have to revise your concept of taking work load off the
    pilot and give more responsibility to the computer?

" Ziegler: No, no, no. It is the responsibility of all aircraft
 manufacturers to deal with human errors and to build into the planes
 technical devices which prevent that those errors llead to
 catastrophes. For errors are in the human nature. We cannot prevent
 pilot errors, but we have to prevent that they lead to catastrophes.
 This is and remains our philospophy.

"Q: Does this not mean that humans are to be banned completely from
    the cockpit?

"Ziegler: No, definitely not! The human is the only one who can react
 quickly to the unexepcted, the un-programmable. [e.g. pilot has to
 decide on aborting a fully automated landing in extremely poor

[Q: on Boeing 777 reverting to traditional type yoke, see above]

"Ziegler: With this, Boeing has overdone [exaggerated] technics. They
 introduce another automatism which is not necessary. This is pure
 cosmetics. We will never introduce a technique which is not

[end of summary/excerpt]

[The above is an excerpt/summary, *not* my opinion (as if I had one on
religious subjects like Airbus...). I started to summarize the Zeit
article because I felt it might indicate that some re-orientation,
much advocated also by members of this list, has finally trickled
through to the industry: namely that the strategy of exposing "pilot's
error" at the base of all accidents, even if that were true, is not a
satisfactory state to leave things in. On second reading, esp. of
Ziegler's statements, I'm not so sure about the re-thinking done.
Maybe Lufthansa got the message, but as to Airbus, I'll leave the
conclusion to the readers. If anyone wants the complete article (in
German, of course), e-mail me privately.]

MArtin Erdelen