Re:Error Chains

Sakuma_at_BDZ.crane.jal.co.jp
Wed, 28 Jan 98 19:53:50 +0900


> It would seem the mood-topic is now entering into Error Chains.
> I talk
> to this a lot in my classes. Lately, I've challenged the students to
> consider this perspective:
> We, in the research and teaching activities, are getting really go
> od at
> recognizing an Error Chain through a historical perspective. That is
> to say,
> we go to a crash site, and investigate what happened. Then we publish
> the
> account of the crash as a story that unfolded through an Error Chain,
> or a
> sequence of contributing causes. With this knowledge, we often conclud
> e that
> we understand how the crash occured, now that we know the sequence of
> events.
> We then, in a class, try to teach the next aviator to "NOT do that".
> This forms my question: "How do you know you are in an Error Chain
> ",
> present time. Some of the reports we read fault "poor planning", but
> I ask
> aviators "did you ever go to the airplane with the knowledge that you
> had done
> a lousy job of flight planning?" Most of them say no, and yes, some h
> ave
> admitted they have begun a flight knowing they were not prepared to fl
> y.
> So my question of the day is "How do you know you are in an Error
> Chain?"
> Greg Deen
> Raytheon

Dear Greg :
I like to try to answer your last question.
Please note here that I would express my comment based upon not the
"Error Chain" theory but my "Rain Drop Model".
Normally, human error can be defined as the deviation of human
performance from what is objectively expected in specific situation. In
reality, objective expectations are reffered to as philosophies,
policies, procedures and practices( 4 Ps ) which are prescribed in
regurations, manuals, SOPs, etc. Due to this context, close monitoring
of the deviation from
4 Ps every time has been considered as good situational awareness for
avoiding human errors( Error Resistant Strategy). However, consider the
case when we are driving a car on the high way. We do not monitor the
centerline of the lane( actually does not exist ), but roughly monitor
the distances from edges, other cars, pedestrians, etc by time sharing
our attention( Error Tolerant Strategy ). In case of rain drops falling
on the window glass, distances between each rain drop and the lower
edge of the window may be good cues to prevent them from reaching the
lower egde. Applying this notion to actual flying the airplane, it
seems to me that observing the margines to various boundaries of safety
envelope may be recommended situational awareness for pilots rather
than paying attention to keep exactly on line with SOPs.
Stall Warning, Barber Pole, GPWS, TCAS, Stick Pusher, Stick Nuger,
Alpha Protection are good warnings and automations for supporting this
pilots SA. However, to utilize these equipments and tools more
effectively( CRM )and to make their design more human-centered still
remains to be top priority issues to solve for realizing "Zero
Accident".
Any comment or correction on my idea is requested.
Regards, Hidetake Sakuma, Corporate Safety, JAL