I think it would be useful to have a discussion on

(1) The options available to prevent decaying of flying skills that occur naturally over time

(2) What could be done to make unusual scenarios a more comfortable occurrence? Naturally a few accident examples don't define an industry, and what might happen to some pilots may never happen to other pilots, or even the same pilots on a different day, but for certain scenarios is there a void left open?

A few ideas that have been floated so far such as flying a whole sector manually may not be practical, and may not actually have much benefit.

Upset and recovery training or similar might be considered too rigid? Surprise actions in a sim training session might still not be able to realistically create the human factor element involved in a real flight? Sims might also not be accurately representing some scenarios.

Edit: Adding a few examples

[Edited 2015-12-02 06:17:06]

What you suggest is presently being performed.

Recurrent (simulator) training now includes a lot of the old school procedures like full stall and unusual attitude recovery. Also some of the LOFT/LOE legs in the simulator now include the loss of either autothrust or autopilot or both, and one (nasty) one includes loss of all FMGCs, where one must do an approach and landing without autopilot, autothrust and flight directors.

Also, pilots are encouraged, where conditions permit, to fly the aircraft on the line without some or all of this automation. (except the FMGCs, as that shuts down more than just the automation). I see it more and more often.

But I am fortunate. The the aircraft I fly, the B767, flies beautifully without autopilot and autothrust and is a pleasure to fly.

Quoting AIRWALK (Thread starter): Sims might also not be accurately representing some scenarios.

The problem you have in doing things in the simulator is any flight outside of the approved flight envelope will not be accurately replicated in the sim. The manufactures are only required to provide accurate representation of flight characteristics for normal flight.

You also have a problem with flight in RVSM airspace. You have to have an autopilot.

The best solution is encourage pilots to hand fly the aircraft as much as possible. Certainly having the autopilot do the work in a high traffic area is common sense but not every airport is like that. Hand flying the aircraft from top of descent to landing would be another option. I was fortunate enough to work for airlines that encouraged that. But, there are airlines who mandate when the autopilot will be engaged and disengaged.

Also, encourage quarterly sims to practice hand flying and concentrate on doing raw data approaches. If that was done, I think we'd see a rise in hand flying skills.

Quoting MMO (Reply 2): The problem you have in doing things in the simulator is any flight outside of the approved flight envelope will not be accurately replicated in the sim. The manufactures are only required to provide accurate representation of flight characteristics for normal flight.

They are working on this at this hour. I believe Flight Safety has a G550 or 650 simulator at this time that meets all of the requirements put forth in the Extended Envelope Training Advisory Circular. I believe deadline for this is the 1st quarter of 2019. Not perfect perhaps but a quantum leap from today's technology.

Great input so far,

Quoting longhauler (Reply 1): Recurrent (simulator) training now includes a lot of the old school procedures like full stall and unusual attitude recovery. Also some of the LOFT/LOE legs in the simulator now include the loss of either autothrust or autopilot or both, and one (nasty) one includes loss of all FMGCs, where one must do an approach and landing without autopilot, autothrust and flight directors.

Do you feel that the aspect of it being a just a simulator is enough to be a problem? Having worked with sims I know how advanced they are and how seriously the sessions are treated, but does the human aspect of replicating ones actions in a sim to a real life scenario get affected.

Quoting longhauler (Reply 1): Also, pilots are encouraged, where conditions permit, to fly the aircraft on the line without some or all of this automation.

Quoting MMO (Reply 2): The best solution is encourage pilots to hand fly the aircraft as much as possible

I agree that this is by far the best solution. It is important to be incredibly comfortable with the plane you fly, so comfortable that nothing really throws you out. Even if you find yourself in a situation you have not encountered before or trained for, if you comfortable flying it, it won't really matter, and it buys you that all important extra time.

Quoting BravoOne (Reply 3):

Good info, thanks

Additional commentary:

Due to recent FAA and pending EASA regulatory changes regarding Expanded Envelope and Upset Prevention and Recovery Training for airline operators, Boeing Flight Services must begin a coordinated effort to prepare for delivery of such training, including training beyond first indication of stall. These regulatory changes are being made because of a Congressional Mandate (Public Law 111-216 tasking the FAA) and the publication of ICAO Document 10011 (Manual on Airplane Upset Prevention and Recovery Training).
Simulator handling characteristics will require mandated improvements in simulator fidelity to support realistic demonstration of all required training, including full aerodynamic stall and recovery. Certain other modeling upgrades may be required to support adverse weather or bounced landing requirements. Some training objectives (such as gusty crosswinds and unusual attitude training) may already be supported by existing devices/data packages.
Notwithstanding simulator technical capabilities, it is Boeing’s position that all flight training requirements regarding Extended Envelope and Upset Prevention and Recovery Training should derive from the aforementioned ICAO Document 10011, in which ICAO recommends full stall demonstration to emphasize awareness that impending stalls can have different indications, and to identify the type-specific characteristics of an impending stall. Full stall demonstrations must also show that the published recovery procedures can be used for any developing or developed stall.
PART 121 REQUIREMENTS (§ 121.423(c) and Part 121 Appendices E and F). All part 121 air carriers, including those who train under an Advanced Qualification Program (AQP), are required to conduct stall prevention training, and beginning March 12, 2019, all part 121 air carriers must conduct instructor-guided, hands-on experience of recovery from full stall and stick pusher activation, if equipped. The requirement for part 121 pilots to receive stall recovery training is statutorily mandated in Public Law 111-216, Section 208. The FAA does not have the authority to exempt any part 121 air carrier from this requirement.

Someone in the air asia thread suggested flight simulator dog fight sessions between A320's and 737's.

Quoting AIRWALK (Reply 4): Having worked with sims I know how advanced they are and how seriously the sessions are treated, but does the human aspect of replicating ones actions in a sim to a real life scenario get affected.

That of course, is the big question.

To their credit, with the advent of AQP, simulator sessions are no longer the same every 6 or 8 months. The scripts change depending on how well pilots are flying and looking at flight data to see trends and errors. Also, the LOFT (Line Oriented Flight Training) and LOE (Line Oriented Evaluation) sessions are played out in real time ... gate to gate, with errors and threats introduced to replicate real life. For example, during a engine failure at high altitude, ATC may call with a re-route, or a F/A may call up to change cabin temperature ... real life stuff that must be dealt with.

That is as realistic as they can get. But still not "real life". The rest is up to the pilot him/herself. It is good practice to remind oneself, no matter how sophisticated the aircraft is, with the automation shut down, it is just an airplane ... that must be flown the same way we learned when we started.

So good steps on the simulator side,

Do you (or anyone) think there is merit in having airlines perform flight sessions in a light aircraft with no automation as a supplement throughout the year? Logistics aside such as cost, would there be any tangible benefit to this, considering some pilots may not actively engage in private flying themselves, or is the discrepancy between airline flying too large for the benefits to translate to the flight deck? If they weren't pressured sessions that are graded (to not add stress and workload to the pilots) but just relaxed sessions in which pilots can reconnect with the barest form of flying, I can imagine there might be some benefit and such sessions could actually be enjoyable.

I don't see any cost versus benefits case for this. Delta is putting a number of their instructors through an airplane upset recovery program down in Arizona with emphasis on instructors with no previous military time in their logbooks. Sounds like an excellent start combined with the enhanced simulator data packages that will be coming out in the next couple of years. At least it's start...

Quoting AIRWALK (Reply 8): Do you (or anyone) think there is merit in having airlines perform flight sessions in a light aircraft with no automation as a supplement throughout the year? Logistics aside such as cost, would there be any tangible benefit to this, considering some pilots may not actively engage in private flying themselves, or is the discrepancy between airline flying too large for the benefits to translate to the flight deck?

I think the benefits would be limited at best given the substantial differences in flying qualities between light and transport category aircraft (I have considerable time in both) and the subject of negative training and lack of transferability are real.

Quoting BravoOne (Reply 9): Delta is putting a number of their instructors through an airplane upset recovery program down in Arizona with emphasis on instructors with no previous military time in their logbooks.

Quoting BravoOne (Reply 9): Sounds like an excellent start combined with the enhanced simulator data packages that will be coming out in the next couple of years. At least it's start...

I will be interested in their results. I have military high performance aircraft in my logbook as well, but I (and I suspect many in the industry) will be watching Delta's results here carefully. I certainly give them credit for thinking ahead and exploring this option. In the end I suspect that better simulation models will ensure this is all done in type-specific simulators, which would certainly be my preference when the models reach maturity.