reducing "Crew-caused"
approach and landing
accidents 

Pilot-in-charge Monitored Approach

"Not a problem we have" ?

The Guam accident and others caused an industry focus on "culture" and CRM training to improve monitoring effectiveness. But the NTSB had also specifically pointed to the investigation and subsequent adoption of monitored approach procedures as a priority. What then happened (or rather did not happen) to this recommendation gives an interesting indication of the reluctance in many quarters to even contemplate such changes.

From its subsequent actions, there seems to have been little enthusiasm to follow the recommendation in whatever office of the FAA it reached. Although the report specifically referenced the NTSB Special Studies of US accidents, perhaps people in that FAA office, like Gladwell subsequently, took the view that the big problem was that of "Korean culture". As the FAA could not affect that, the recommendation could hardly be relevant to US operators. 

The report was adopted in January 2000, but no significant action appears to have been taken until 2004. Then, the FAA commissioned research psychologists specialising in human factors at George Mason University to make a survey of pilot opinion on the strengths and weaknesses of the monitored approach procedure, and on ways to improve it.
 
A random opinion survey is possibly not the best way to determine the most effective use of a safety-critical procedure and the maximum degree to which it can be safely used. However, such a survey on the subject of monitored approach procedures would by its nature be less likely to find a positive rationale than a different type of enquiry, perhaps into why operators had developed it, and the extent to which it had proved effective.
 
The GMU survey.
 
The researchers discovered that although monitored approach procedures are used by a wide variety of operators, it was hard to obtain definitive information on the subject.  Procedural details such as specific callouts and applicable conditions varied from operator to operator, with their "proprietary nature" amongst other factors contributing to a lack of firm information, so the researchers considered that post-accident reports might provide the best source of information.
 
To meet the brief for a survey of pilot opinion, the researchers created an internet questionnaire. Although generating some useful data, this unfortunately lacked technical clarity, possibly because of the detail variations discovered. The small number of respondents (205) were by definition self-selected; less than one third (66) had significant experience of using the procedure ("dozens of times"), while more than a third (72) had never in fact done so and could not speak from experience.
 
The survey did not reach the large numbers of pilots with individual experience of performing many thousands of monitored approaches.  Hardly surprisingly, the resulting data showed multiple areas of confusion and misunderstanding of the subject.
 
The report did not make any connection between its basic subject (the "monitored approach" procedure) and the 1976 and 1994 NTSB Special Studies. Although both of these refer to procedures where the First Officer is PF for the approach prior to the Captain making the landing, they do not specifically use the term "monitored approach", and neither is cited as a reference in the study.
 
Consequently, the researchers concluded that they had merely obtained "an initial glimpse... upon which further research can be built", cautioning that the questionnaire had many limitations and had indicated multiple areas needing further investigation, including empirical simulator-based research.
 
However, it seemed that this was enough to satisfy the FAA office concerned, which took no further action until May 2009. Then, more than 9 years after the NTSB's initial recommendation, the FAA finally responded to it.
 
FAA response to NTSB.
 
The response was signed by Lynne Osmus, who had recently been appointed as Acting FAA Adminstrator during the change of Administrators when President GW Bush handed over to President Barack Obama. By a quirk of timing, Ms. Osmus, a Security and Dangerous Goods expert without significant flight deck experience, was then replaced as Adminstrator with a highly experienced former airline captain, just a few days later.
 
The FAA letter sent by Ms. Osmus as Acting Administrator in response to AAR Rec. A-00-10 was largely derived from the survey. It showed considerable misunderstanding of the subject and contained several factual errors. Being based on subjective opinions from the survey, it turned detailed and sometimes company-specific aspects into generalisations which can best be described as partial and in many respects were downright misleading - for example it strongly implied that the procedure did not have any application other than in Category 2 auto-approaches.
 
Finally, Ms. Osmus' response said that the FAA had adopted alternative methods to improve approach safety.  It did not mention that the FAA itself had several years earlier published the joint FAA/ICAO/FSF training guide recommending operators to consider using the procedure "for all approaches in IMC and at night".
 
The NTSB finally commented that one of its recommendations was that the FAA "require operators to modify their procedures accordingly" [i.e. adopt monitored approach procedures to the maximum safe extent]. Although of course the FAA had not done this, the NTSB said it "has generally found that carriers are increasingly adopting and using the monitored approach technique". It then closed the recommendation on the basis that the FAA had taken "acceptable alternate action".

Recurring factors... 

In an accident report, the NTSB describes how a wide-body jet crashed short of the runway, killing all aboard.  The crew were tired after a long duty period, on the wrong side of their normal circadian rhythm.  The Captain was the PF, making a night approach and  expecting to become visual well above their IFR minima. The crew had not received all information relevant to their flight and as the aircraft descended it became apparent that not only was the weather worse than expected, but the instrument vertical guidance the Captain intended using was not available.

The Captain modified the type of approach he needed to carry out, but did not fully brief the First Officer (PM) on what would be required for the non-precision approach to which they were now committed. 

The PM First Officer attempted to acquire visual references, and did not make all the required call-outs approaching minimums. The PF Captain also attempted to acquire visual references but because the weather was significantly worse than anticipated, neither pilot was able to do so prior to MDA.  At a high rate of descent and below the required profile, the aircraft passed below the MDA.  3 seconds after both pilots had seen the runway lights, the aircraft was destroyed on impact with the ground.  

An informed reader might think these paragraphs to be describing the Guam accident in 1997. According to the "ethnic theory" of high cultural PDI, "the culture the pilot came from was the single most important variable in determining whether the plane crashed". Hence appropriate preventative action for the future should include getting the crews of this airline to behave more like US crews. 

.... but a different accident.

But this report was not into KAL at Guam in 1997.  It was published in August 2014, and relates to a US airline which already had a commendable CRM training program.  The CVR does not seem to indicate problems due to a high PDI culture inhibiting the First Officer - rather the reverse.

The pilots seem quite relaxed together, with the First Officer "chuckling" at a comment made by the Captain shortly before impact. The F/O was described as a "a top-notch....  professional aviator who followed procedures"; as "efficient, did her job, was on time, was someone you could depend on, and used the procedures as trained". Colleagues had no concerns about the F/O's CRM skills, and "believed that [the F/O] would speak up to a captain if necessary. 

In this accident "only" the pilots died, because it was a cargo flight, and hence it may not generate the public attention of a major passenger aircraft accident. But it illustrates that generalisations about the pilot's national culture as the primary cause of accidents can be, at the very least, extraordinarily misleading.  In fact, many factors of this accident, the Guam one and many in between, are identical to those in the accidents listed in the NTSB's Special Study in 1976, nearly forty years earlier.