Dave Wilson summarised the discussion in this thread which, incidentally,
started when Barry Forster at Air2000 wrote in to ask about Error Chain
training resources. (I'm not sure he got exactly what he was after!) The
point I found most interesting was:
>>>
Surely then we can train crews to recognize a pending problem. Our teaching
focuses on the decision process and situational awareness. As a decision is
reached, the other possible choices should be examined for their potential
for
incident/accident/danger. If one of them could lead to trouble, one should
progress (mentally) back up the decision tree to see if there was an
earlier
fault and what were the consequences.
<<<
Some work I did last year which involved crew coordination failure
modelling based on hundreds of accident/incident reports suggested that
most crew performance breakdowns start not in the upper cognitive strata of
individual or collective planning, decision-making etc. but further down
the hierarchy, in specific communication failures, error-trapping
deficiencies etc. These are often obvious at the time - missed clearance,
incorrect readback unchallenged, ambiguity tolerated. Low-level failures
then begin to contaminate higher level processes, individual skill
performances and situational awareness. The useful thing about this is
that low-level processes and error-recovery are highly trainable.
Even if the low-level failures are missed consciously, how often do they
leave a residual discomfort? The Time Out interrupt procedure should work
best where (1) the crew are trained to react strongly to unquantified
anxiety and (2) the situation has not reached the point where an interrupt
is ineffective (e.g. too busy to check an assumption that the GPWS is
failing safe.)
My two pence in an interesting debate...
Regards,
Rick Heybroek
LOFTwork ltd.