If a pilot decides to crash an airliner, a la Germanwings or Egyptair, could he be stopped?

In those cases the offending pilot waited until he was alone in the cockpit to act. Thus many airlines have made it mandatory for there to be two crew members in the cockpit at all times, even if it's just a flight attendant to stand in there and wait for a pilot to return.

Although that might give passengers some piece of mind I am doubting that it would do much good. It seems to me that, even if the other pilot is in his seat and strapped in, if a pilot decides to crash the plane he could kill the engines and put the plane in an unrecoverable attitude very quickly.

Is this a correct assumption? Would Airbus envelope protections prevent this? Can the plane be placed in direct mode to override the protections? What about a 777? Does the yoke feedback system allow one pilot to overpower the other? The engines could easily be throttled back, but could they be quickly shut down requiring a restart? What could the other pilot do to stop his actions?

  • $\begingroup$ I'm mostly asking what the other pilot could do and if FBW aircraft make the scenario more difficult. If it seems too much of an opinion question focus on the specifics mentioned in the last paragraph. Can engines be killed and could a plane be put in an unrecoverable attitude quickly $\endgroup$ – TomMcW Dec 30 '16 at 19:40
  • $\begingroup$ As it's currently worded, this seems like a whole lot of questions rather than a single question. Maybe it would be better to focus on a single scenario? Whether there's a way to stop a pilot from crashing an airliner seems like it will be very specific to the particular scenario. It's also worth noting that shutting down the engines of an airliner is far from ensuring a crash. From cruise altitude, most airliners can glide around 100 miles with no engine power. $\endgroup$ – reirab Dec 30 '16 at 20:02
  • $\begingroup$ @reirab Yeah, but dead engines would make unusual attitude recovery more difficult. And engines need some fairly specific speed and altitude parameters for restart. Maybe I can break it down. I'll have to do it later, though. Asked the question during lunch. Unless you feel like fixing it $\endgroup$ – TomMcW Dec 30 '16 at 20:06
  • $\begingroup$ It might make it harder depending on what the unusual attitude in question is. For others, it might not matter at all. And some attitudes might even be made worse by having the engines produce thrust (e.g. a steep dive.) Once you have time, this seems like it would be a better question if it described a particular set of actions of the suicidal pilot and asked if/how those could be countered. Setting up a particular "threat model" of what you're trying to counter is usually the first step to determining an effective counter in the security industry. $\endgroup$ – reirab Dec 30 '16 at 20:17

COULD you do that? Sure. The relevant question, though, is if the tradeoffs involved would be worth it.

How could you do that? Essentially, a trusted 3rd party could have the ability to disable all controls that the demented pilot has access to and assume all control over the aircraft. If you're willing for somebody in the network operations center to have the ability to "turn off" every switch & control in the cockpit and remotely pilot the airliner to its landing, then you're set. The pilot is now every bit as much a passenger in the aircraft as everyone sitting behind him is, and short of taking the crash ax and causing a depressurization, he can't affect things that much. (And even if he did that, the airplane still won't crash.)

The problem with that plan is obviously that if the remote control function gets taken over by hackers or a demented individual at the operations center, the pilots are helpless to intervene. After all, for the demented PILOT to be rendered harmless, you have to remove ALL ability to override the remote control, all ability to pull circuit breakers, all ability to affect everything.

So, where is the greatest threat? A demented pilot that overcomes the other pilot (by force, by locking him out of the cockpit, whatever), or problems at the operations center or in the remote control process? The track record of airline pilots is unmatched by nearly anything else for reliability and resilience, and we aren't nearly as vulnerable to being hacked as any remote control datalink would be!

The general understanding is that the best solution is to keep the pilots in control of the aircraft, and have measures in place so that if one of them loses it, the other pilot can keep things going. Things like having a flight attendant in the cockpit to let the other pilot back in (would have changed the course of the Germanwings flight) can go a long way. Or if there is a struggle for control, a pilot could summon the flight attendants over the PA to get help restraining the madman.

Perfect solutions? No. But better than allowing the aircraft to be taken over by somebody outside!

After all, while "pilot goes nuts" is a vivid headline, the extent of the actual risk is quite limited. Millions & millions of commercial flights happen without incident every year, and introducing a slew of new risks in the form of remote control that can't be over-ridden, seems like a severe over-reaction to what is in fact an incredibly remote scenario. There are better ways to deal with the extremely rare & unlikely "crazy pilot" possibility than that.

  • 1
    $\begingroup$ We've had a question or two about the remote control thing. Clearly that is exactly the wrong thing to do. Even without the spectre of hacking you would then have to trust, not only the pilots, but the persons on the ground that could do an override. And their life would not be at risk. Bad bad bad! $\endgroup$ – TomMcW Dec 30 '16 at 19:24
  • 1
    $\begingroup$ And still, the guy on the ground would have no more ability to recover from dead engines and an unrecoverable attitude than the other pilot would. $\endgroup$ – TomMcW Dec 30 '16 at 19:33

Not the answer you're looking for? Browse other questions tagged or ask your own question.