1) Devote a large part of the annual crm class to analyzing IN
DEPTH an accident report, with emphasis of getting the class to analyze the
error chain. Some companies already do this, but often it is done
superficially and casually. What i'm thinking of would require
considerable preliminary work by the crm department. The accidents chosen
should not be those rare instances in which the crew was behaving wierdly,
but rather be cases in which the average pilot might say to himself: "I
could have gotten into that situation". Rather than simply cataloging all
the errors the accident crew made, the instructor should lead a discussion
of what factors can lead an experienced and competant crew to each of the
behaviors (commissions and omissions) that contributed to the accident.
The discussion should also include warning signs that might have alerted
the crew had they been looking and should explore strategies crews can use
for breaking error chains.
The crews will not have read the accident reports before
hand, so the crm department must boil the report down to a brief
description of critical events and behaviors that can get the class up to
speed in the first five minutes.
One problem with this suggestion is that, although we all
are fairly familiar with error chains and how they lead to
incidents/accidents, we only barely understand what causes the various
types of error. (This is partly because there has been little funding of
research on the causes of crew error. However, Jim Reason's work points in
the right direction). I wonder if it would be useful for researchers to
work with crm departments to jointly analyze representative accident
reports in terms of what is known about cognitive processes and how those
processes make crews vulnerable to certain kinds of error in certain kinds
of situation?
2) Do what Doug Daniel did when he was LOS coordinator at Delta (i
suspect some other companies are already doing the same): Focus the
recurrent crm class and loft on a single theme each year that can be
explored in depth, e.g. workload management or decision-making. For the
theme chosen, identify specific warning signs that should alert the crew
that things are starting to get out of hand (e.g., feeling rushed) and
strategies to keep problems from propagating. Follow the class discussion
up with reminders in the briefing and debriefing.
3) During the crm class (initial or recurrent) have the crew rate
the performance of a crew in a videotape of a LOFT or LOE, using the
grading forms instructors use for LOE in AQP fleets. The act of grading
will bring out a lot of discussion of what constitutes good and not so good
crm performance (It brings out lots of discussion in the instructor
calibration classes, so i think it will have the same effect with crews).
Key
At 1:34 PM -0700 7/22/97, Guy M. Smith wrote:
>Jan Davies wrote:
>
> "I'd be interested in hearing your ideas on how to
>familiarise crews with their (individual) cognitive
>thresholds."
>
>Jan, I can speak for one airline that we do not address this issue with
>our flight crews, but it is a topic that warrants exploration. Jan's
>question has an interesting challenge for all of us that work with
>flight crews. It is part of a bigger question, "What is the future of
>human factors training in the airline industry?" I get this question
>often -- just yesterday from my Managing Director. There seems to be a
>general consensus among line pilots that CRM has worn out it's welcome;
>that pilots are tired of hearing this CRM stuff. I'm one who is willing
>to admit that there may be some truth to that. It isn't that pilots
>don't need CRM anymore; it's that they feel that they have gotten the
>"undergraduate" CRM course. Maybe they are ready for advanced training.
>Perhaps "individual cognitive thresholds" is one subject that should be
>considered for advanced human factors training. Of course, "individual
>cognitive thresholds" might work in our vocabulary, but it must be
>packaged differently to be palatable to the line pilot.
>
>My question for the group is, "What are some of the other topics that
>should be included in "Advanced Human Factors" training? A second
>question is, "If we perceive a need to 'return to the basics,' how do we
>package the basics so they are different, interesting, and accepted?"
>--
>Guy M. Smith
>Manager, Human Factors Practices
>Northwest Airlines N7205
>5101 Northwest Dr.
>St. Paul, MN 55111-3034
>Phone 612-727-4255