PPRuNe Forums - View Single Post - Automation dependency stripped of political correctness.
Old 5th Jan 2016, 15:36
  #32 (permalink)  
FDMII
 
Join Date: Feb 2015
Location: Alternate places
Age: 76
Posts: 97
Likes: 0
Received 0 Likes on 0 Posts
RAT 5, re:
Of course a certain profile had a crew coordination process and a correct use of the automatics to achieve a safe result. That is correct as it is a crew procedure and both expect the other to act in a certain way & certain times. No confusion. Solid SOP required. Pass/fail?
What many forget is that it is an SOP to deviate from SOP's when safety dictates. What that requires is the realisation that something is going awry, why the normal SOP will not be the best, and what other FCTM technique is a solution. IMHO this is what is missing in basic TR, and more critically in command training. It can be that SOP focus is even more intense in that phase.
I like that series of observations. I think it captures what is meant by "guidance of wise men,...", etc. I particularly agree with you where automation and SOPs is concerned, (Solid SOP required), and I would add the same comment where the activation of protections (for both Boeing & Airbus) are concerned; Follow the book!

That said, I recall ignoring the ECAM Abnormal drill on an A330 flight, for what I considered to be very good reasons, a decision which proved to be the correct one but only after all was said and done. That's what the captain is for. And, glancing at the thread's thesis for a moment, almost certainly any software/autonomous/pilotless aircraft would have made the wrong decision and placed the aircraft and its payload in a high-risk condition. This is because software cannot parse what the belly senses, nor is there is a way that a mere abundance of "data" could replace experience and render a "correct" decision.

The "audit" item I brought into the dialog as only lately has this added notion entered into the discourse on human factors accidents when SMS began to take hold. I like SMS and think it is far better approach to risks and occurrences than blame/discipline/enforcement models. But the difficulty with the audit process is that it is a vertical, bureaucratic model (vice lateral process), which requires satisfaction of superiors in the up-and-down structure of command-and-control, without permitting variability in standards, (lest one open oneself to liabilities, blame and ouch).

I think that is a potential failing of the audit process should it not be designed or handled well, and can cause precisely what such process is intended to avoid/prevent.

Last edited by FDMII; 5th Jan 2016 at 15:51.
FDMII is offline