There have been several meaningless statements about poor CRM or lack of airmanship, or that crew should have ‘just flown the aircraft’; none provided explanatory detail, any reasoning, or offered a practical solution.
What part of CRM failed, how, why – if we knew, then solutions could be found? Similarly, for airmanship or flying ability, what critical aspects would have saved this crew?
One view of safety is that humans generate safety. Thus establishing what other crews have done in similar circumstances may help with an understanding of this accident and provide points of learning.
It is reported that at least one crew knew of the RA abnormality. Additionally, it’s hard to imaging that in the long history of the 737 that other crews have not encountered this fault – so how did they manage.
How did the crews detect the RA fault;- by viewing the display, an unusual alert (gear), or AT mode change?
Even discounting seeing the display, the alert and AT mode should still have occurred providing a range of cues.
An EGPWS gear alert is given when RA <500ft, which suggest that the fault occurred at the time of the alert, i.e. not earlier when the RA operating / display logic became valid.
The AT mode change requires a combination of flap and RA <27ft, thus in the accident the AT mode change may not have been coincident with the gear alert, i.e. it required a flap selection. The crew may have considered the gear alert as an isolated problem – possibly cured by lowering the gear. In this instance then the flap selection started the low speed event and the AT mode change; was this the same for other crews.
In comparison to an ideal operation the accident approach profile appears higher than normal (joining GS from above) and possibly fast/tight on an energy schedule; if so, the flap selection may have been later than ideal, which reduced the time for detecting the AT mode change, time which could be further compressed by the pressures of a training flight.
In other operations, did the crews have extra time, by either earlier flap selection or not suffering similar pressures of operation; - time in which to scan the flight deck and check the display. Scanning – situation awareness, and time management are aspects of CRM.
During the accident approach, it would be expected that the AT would be at idle, enabling speed reduction on the GS and deceleration with further flap selection. Did previous crews have a similar expectation, if not why not? Checking and control of expectation - CRM, discipline - Airmanship.
A problem for the accident crew was the drift from normal operations – small deviations in height (above GS) and higher speed (energy), compressed time, and in a training environment.
A Training Captain requires skill to judge how far to let a training situation drift (skill / judgment; qualities of airmanship).
All crews require similar skills of judging how far a normal operation can be allowed to drift towards the boundary of safe operation, but as our judgments can be biased, distorted, and fallible, we need to introduce constraints – boundaries or check points.
The stabilized flight path concept has constraints and check points, but these depend on discipline to adhere to the procedures – doing what we say we will do (airmanship).
As for the skills of flying, many of these come from practice and exposure to similar situations, which are the basis of experience. Without sufficient experience, such as during training, then the boundaries of normal operations must be strictly respected, even tightened to ensure that if the unexpected happens there is time to adjust and remain within the bounds of a safe operation.
I suggest that many of the safety points from this accident come from identifying the things we (industry and individual) do routinely to maintain safe operations, and from those rare occasions where we drift towards the edge of safety. We have to look to ourselves before looking too closely at others, and share best practices of normal operation and lessons learnt from the non-normal operations.