reducing "Crew-caused"
approach and landing
accidents 

Pilot-in-charge Monitored Approach

Perceived risks of PMA in visual transition.
Many pilots believe the PicMA concept has potential or demonstrated risks, and these should be examined. The principal problems generally perceived to exist are as follows.
 
"The final approach path will be less accurately flown". In many airlines, PicMA is associated with visibility conditions in which the operator specifically requires the Captain to make the final decision and land the aircraft. In these circumstances the aircraft will be flown by the First Officer, who is normally the less qualified and generally less experienced of the two pilots. This is thought to raise the risk of an accident due to a poorly flown instrument approach down to DH. In fact this concern is baseless and irrational, for the following reasons. 
 
Manual approaches: First Officers who hold an Instrument Rating must have demonstrated their ability to fly a manual Instrument Approach to DH and execute a go-around. If this concern is valid, it raises several linked issues.  
  • If Captains generally do not have confidence that their First Officers are able to fly the aircraft to this standard, then the entire safety basis of the operation is called into question.
  • If the individual Captain on a specific flight is not satisfied that his First Officer is capable of flying to the required accuracy, then serious consideration needs to be given to the overall wisdom of conducting an approach in minimum conditions.
  • In either case, it's doubtful whether First Officers who are unable to fly to the required standard could effectively carry out the role of Pilot Monitoring to a Captain flying the approach.  
"The aircraft could be 'out of trim' on handover.  This is closely related to the previous concern. If the aircraft is out of trim following a manual approach, then there is clearly an issue of training involved as qualified pilots should not arrive at DH on an instrument approach with the aircraft significantly out of trim. Likewise, after an autopilot approach, the trim situation on disconnect is identical regardless of which pilot was programming the autopilot. 
 
Automatic approaches: if the approach is being conducted using the autopilot, the same considerations apply. With traditional PF/PM procedures the First Officer is monitoring the performance of the Captain's setting up and monitoring of the autopilot. However, the autopilot's actual performance is unaffected by which individual is actually programming it.
 
The issue then becomes one of competence at setting the systems up prior to final approach, which does not affect the accuracy of the final approach itself. The autopilot will of course fly an inappropriate flight path just as accurately as an appropriate one. 
 
"There's no crosscheck of the land/go-around decision." This concern actually illustrates a popular misconception, which is that a "decision" can be "cross-checked". A cross-check involves two or more people evaluating or verifying information and confirming that they are in agreement, before any subsequent action is taken. For example, a cross-check might be that both pilots must confirm that the speed is correct before the PM moves the flap lever. 
 
But operating minima are based on requiring that "the pilot" (which means only the pilot who will make the landing) "is satisfied that the aircraft's position and rate of change of position" are such that he can safely complete the landing. This is a "Yes or No" decision that has to be made at a specified point followed immediately by action to go around if necessary. This is not compatible with being "cross-checked".
 
With traditional procedures, the PNF might hold an opinion to the effect that the PF could safely make a landing, but he cannot enforce it. He is not the pilot flying, who must actually do it. Crosschecking implies that disagreement is possible, that the decision could be wrong and therefore must be reversed.
 
It is certainly the case that most airlines would expect - indeed require - the F/O to call for a go-around from below DH if they detect a deviation from a safe flight path, such as an increased rate of descent or divergence from the glide path. But that is based on having maintained instrument information, to supplement the Captain's visual information - it is not a "cross-check" of the decision based on the visual cues. Similarly it is hard to envisage a situation where the F/O's "cross-check" resulted in a Captain's decision to go around being reversed.      
The situation is similar to that on takeoff with an engine failure just before V1, when the pilot makes a decision to continue or reject the takeoff. Just as the overall takeoff performance assumes a correct decision will be made at V1 when calculating runway length versus maximum takeoff weight, the DH/RVR determination assumes a correct decision will be made at DH. 
 
It's questionable whether any airline would in fact want to specifically write into its procedures a requirement that the First Officer should cross-check a Captain's decision at DH whether to to land or not land. By definition that would involve a delay, and opens the question of what happens in the event of a disagreement. 
 
However, the safety of the flight will always remain the Captain's ultimate responsibility, and it the situation is different when ANY decision is to be made by a subordinate, e.g. the First Officer. Then, the Captain will almost always want to "cross-check", and must have the capability to to do so.
 
So this concern is really linked to the question of "First Officers Sectors", and is discussed at more length there. 
 
"There could be confusion about who's flying at DH". The idea of exchanging control at low altitude engenders fear in some people that "confusion" will occur. In all critical aircraft operations, clear procedures and callouts must be laid down and appropriately trained, and in this respect a PicMA procedure is no different to any other. 
 
The simplest resolution of this concern is to minimise the number of words used at DH.  What is required at DH is that a DECISION be announced.  
The callout at DH should be "DECIDE": this is a clear indication to the Pilot that the time has arrived to announce his decision. Calls of "minimums" could be taken as simply "advisory" information.   
This should require an immediate response of either
    "Go-around" - a clear command to the P2 to commence the go-around; or
    "Land" - a clear announcement of intention, that can be stated as meaning "I have control" in this situation, and causing the P2 to relinquish control.     
Because a PicMA generally involves fewer exchanges of information source for each pilot, experience indicates that, when the procedures are compared on a like-for-like basis, there is in fact less potential confusion and fewer callouts are needed. However, again analogous to engine failure on takeoff, it must be routinely rehearsed to maintain proficiency.
 
Of course, unlike an engine failure on takeoff which cannot be simulated in normal line operation, a PicMA can be practiced in complete safety on any flight where an instrument approach procedure is available, regardless of actual weather. 
 
If PicMA is used as the SOP for all approaches, (while still allowing early transition to visual in appropriate conditions, this concern competely disappears. 
 
"There is more risk of mishandling during go-around." This has again been raised as a potential drawback of PMA but again there is no rational reason why this should be the case, and no evidence that control handover makes go-arounds less safe. Rather the reverse is true.
  • As the approach success rate tends to be higher in minimum conditions, there are fewer go-arounds to start with.
  • The go-around is flown by the P2 who is fully acclimatised to instrument flying. With traditional PF/PM procedures the PF has to make a double "instruments - external - instruments" multiple transition. This is in itself results in a serious risk of disorientation etc.
  • The P2's basic mental model is one of a plan to fly to a go-around, so when a go-around is necessary it is much less subject to either "plan continuation bias" or "startle factor" which has been a serious cause of accidents. (The 2013 French study of this subject is very valuable.)
"Rejected landings will be harder to control". The situation on the case of a rejected landing, i.e. a go-around initiated from below DH following an initial decision to land, is no different between the two procedures. The pilot who had been intending to land has now changed his mind because of some factor, e.g. changed environmental conditions of visibility, wind, runway incursion etc. He calls for a go-around which he flies himself with the P2 following normal PNF duties. 
 
"Sharing the handling means reduced job satisfaction."  A fairly common comment is that there are negative psychological aspects to PMA which are significant drawbacks. In fact these appear to be more emotionally based than rational in that they often refer to it being "unsatisfying" or "less enjoyable" to fly an approach but then hand over the landing. While this can be a real emotion it is not a sufficient reason to ignore the safety aspects which have been demonstrated earlier. In fact it appears that for many First Officers there are distinct psychological benefits to being entrusted with a critical approach flying task even if it does not terminate in a landing, as competence and consequently self-confidence rises rapidly.    
 
Data to support these worries.  
Although concern over these perceived disadvantages of the PicMA concept have frequently been expressed, there is little if any data that indicates that they are justified, or that these perceived risks outweigh the known advantages in overcoming the human limitations involved in manual landings of an aircraft in poor visibility.