RAT, I agree 100% with your sentiments although my background is rather different - 1960s high school > cadet programme flight school > advanced airliner jet right seat with 220 hours. At the risk of being seen as flogging a hobby-horse to death, on another website a few months ago this was written.......
"Automation and "de-skilling".
A problem of increasing concern in the airline industry is the risk of pilots losing manual flying skills. Historically, much training emphasis has been on strict adherence to SOPs that involve the use of autoflight systems, to maximise the economic benefits of having these systems fitted as well as the safety benefits in situations such as very low visibility.
But as noted by the FAA in 2013, "Autoflight systems are useful tools for pilots and have improved safety and workload management, and thus enabled more precise operations. However, continuous use of autoflight systems could lead to degradation of the pilot’s ability to quickly recover the aircraft from an undesired state."
To counter this, the FAA now encourages operators to incorporate emphasis of manual flight operations into line operations. This attitude is reflected in the attitude of manufacturers as well: both Boeing and Airbus are revising some aspects of their conversion training programmes.
However, many airline managements have now become convinced by earlier arguments that automation could only bring benefits, believing that "automation has made flying safer, so much so that pilots need to be instructed to use it as much as possible to achieve leaner, better, more efficient flight and save on wear and tear to engines and so forth."
This has created what many operations experts now see as a dangerous over-reliance on the autoflight systems, particularly among less experienced pilots, who are sometimes referred to as "children of the magenta line" (magenta being the intended flight path as shown on a Boeing electronic map display). Perhaps subconsciously, the overall chain of control (if not command) is being perceived as one where autoflight systems should be allowed to control the aircraft to the maximum possible extent, and by implication have the highest authority over the flight path.
The new advice to "incorporate emphasis of manual flight operations into line operations" therefore presents airline managements with a dilemma: how to have manual flight in line operations simultaneously with maximising the safety benefits of automation? The use of PicMA as an SOP may well be helpful in this respect.
The PROPER hierarchy
Operations using PicMA as the SOP can clearly be shown to reinforce the proper chain of command and control of the aircraft, in which automation is a subordinate to both crew members. In normal operations, the Captain (P1) supervises the First Officer (P2). The P2 is responsible for setting up the autoflight systems to achieve a satisfactory approach for the P1, and for supervising the autoflight systems to properly achieve it.
This chain recognises that autoflight systems normally do a great job of flying accurately, but have no intelligence and cannot handle unforeseen circumstances. As clearly pointed out in the FAA's PARC/CAST Automation report "The safety and effectiveness of the civil aviation system rely on the risk mitigation done by well trained and qualified pilots (and other humans) on a regular basis."
If this hierarchy is recognised, then it becomes easier to treat the autoflight systems as the useful but subservient tools they are intended to be, and find a way to safely practice manual flying in routine operations.
Good and bad days.
Every flight is unique in its details, but has common characteristics with thousands of others. The extremely rare approaches that end in CFIT accidents have a combination of features that meant that an ordinary crew error that would have been survivable on almost all other occasions, had proved lethal.
Consider the "normal", every day, safe operation. The aircraft exists in a "bubble" or "regime" which has a certain amount of tolerance of crew mistakes. This tolerance is reduced by threats in three broad domains - the aircraft itself, the world (and in this case specifically the approach) it is flying in, and the pilots themselves. As risk factors in these domains build up over a spectrum from negligible to severe risk, the flight's regime becomes less and less tolerant of errors.
For example, "aircraft" factors include its complexity, existing MEL items, and maintenance standards that cause defects to arise in flight. The "world" includes terrain, the actual weather encountered compared to that forecast and then reported to the crew, ATC quality, and traffic density. And in terms of "people", the crew members' training and actual skills, experience, culture and language, personalities, individual concerns, fatigue and alertness.
When risk factors or threats coincide, vulnerability to errors is multiplied. This is of course why initial pilot training, when errors will be frequent, is conducted in a very error-tolerant regime in simple aircraft, over flat terrain, in good weather with low traffic density etc.
The level of safety expected by the public should be provided by operators meeting regulations which set required margins to provide a reasonable degree of tolerance for crew errors even after some threats arise, like the examples shown.
However, in the real world, these assumptions cannot always be met. Threats become bigger until on rare occasions they overwhelm the crew's capacity to deal with them, and the crew become very vulnerable to their own errors.
Although obviously many other aspects are involved, some airlines generally operate in a much more error-tolerant regime overall than others. For example, US domestic operations are conducted with no issues of language to affect pilots' understanding either of their aircraft (training and manuals etc), or the environment (ATC communications), and with underlying high standards of training, regulation, and air traffic control.
It does not require major threats in all factors to remove error tolerance and bring about an accident: for example the Asiana B777 SFO accident did not contain major "world" threats in terms of weather or terrain. The aircraft was fully serviceable, but as in other events, the problem appeared to be crew unfamiliarity with aspects of its automation's complexity and the aircraft's basic flying qualities, factors which have been partly responsible for the current re-emphasis on manual flying.
"Automation is safer and cheaper, so how we can allow manual flying ?"
Where managements need to deal with the dilemma of "autoflight is safer - but manual flying is still needed", the use of PicMA procedures as standard could provide a solution. Manual flying for pilot practice and familiarity provides a long term safety benefit, but it needs to be done in benign conditions, when the notional reduction in safety levels of not using the automatics can be balanced against the greater error tolerance of having fewer threats.
Even so, this may not be acceptable to many managements, unless additional safety measures are put in place. The demonstrated greater effectiveness of monitoring using PicMA is exactly the sort of additional precaution needed. For example, during a normal sector in benign conditions the Captain would have good visual cues, etc., and plenty of "spare capacity".
Under these circumstances the First Officer could safely concentrate on flying a MANUAL instrument approach, in the full knowledge that the Captain will be taking control himself either approaching DH/DA minima, or at a higher altitude and with full visual reference, in the event that the F/O experiences difficulties.
Automation and Captain's discretion.
Manoeuvring the aircraft, especially hand flying, is one of the most enjoyable experiences imaginable - and that is why so many pilots, especially Captains, are reluctant to accept the concepts of PicMA. But the simple fact is that passengers are not paying for pilots to enjoy themselves. They have the right to expect the safest possible flight.
Commercial operations should always be based on the assumption that there will be problems, and require crews to prepare themselves by planning an approach to give the maximum possible safety margins so they can deal with such foreseeable problems. But then, once it becomes clear that those problems have not materialised, for example on a fine day with little ATC congestion and a fully serviceable aircraft, Captains can and should use their discretion as aircraft commander to allow the extra safety margins, that have now been proven to exist, to be used for other purposes to increase the operation's long term safety. This could include
• continuing to fly a PicMA down to the appropriate minima without using all the elements of available automation.
• after confirmation of being stabilised "in the slot" with full visual reference and a landing clearance, the Captain (pilot) assumes control well above minima to practice a primarily visual final, while the First Officer (Copilot) maintains full instrument monitoring down to touchdown.
Practice for both pilots.
While this appears to refer only to the First Officer getting significant manual instrument flight practice, the reverse would also of course apply during a First Officer's sector, to allow the Captain manual flying practice. As we've seen in the section on "benefits in all cultures" this also gives an incentive to Captains to allow First Officers practice at takeoffs and landings.
These practices help maintain the essential basic handling skills that many pilots, manufacturers and regulatory authorities are concerned are now being eroded by an overemphasis on the use of all available automatics on all possible occasions. Emphasising that this is a permitted deviation from the default SOP, but is always at the Captain's discretion, leads to an increased alertness of all crew members to the need for vigilance.
"Never land off your own approach?!?"
Making PicMA the default SOP does not mean that in airliners, the same pilot should never fly both approach and landing - that would be absurd. It is absolutely essential that all pilots be able to conduct an approach and then land off it. There will always be the need for visual approaches; for training; to develop skills to handle abnormal configurations safely; and to cover rare cases like incapacitation, among others.
Test pilots in particular need the entire flight envelope, and check pilots and trainers must see it demonstrated to proficiency. Pilots get great satisfaction from taking an aircraft through the whole cycle from takeoff to landing, and any pilot who does not is probably in the wrong job and will not last long. But dividing these tasks is simply the sensible thing to do for maximum passenger safety in normal commercial operation."
Awaiting incoming.....