Suppose the pilot or pilots of a plane become incapacitated for whatever reason. At the moment, there is nothing that can be done from the outside to stop it other than to shoot it down (am I right?).

Would it in theory be possible that planes have a 'safe landing' program built in that, at a signal from the ground takes over control? It then does everything necessary to land correctly and automatically at a nearby airport.

I realise I am making a lot of assumptions here. My main question is not about the logistics of what would have to happen on the ground, nor is it about whether a new generation of terrorists could hack into such a system for their own purposes. It is about how easily current autopilots could be adapted to have a 'land safely' option. What modifications would the manufacturers have to install within the plane itself that would make this possible?


  1. Is it true that currently, the only way to control a runaway aircraft is to shoot it down?

  2. In theory could an existing autopilot have a simple interface that is retrofitted such that a single signal from the ground sets it in 'return to base' mode and produces a safe landing?

  • 5
    $\begingroup$ The probability of being hijacked is very much lower than the proability of external remote control being hacked or malfunctioning. Therefore, what problem would it solve? $\endgroup$
    – Simon
    Oct 27, 2015 at 13:14
  • 1
    $\begingroup$ If the good-guys can remote-control the airplane, the bad-guys can to. Every system can be hacked. By trying to block one attack vector, you've added an additional vector. $\endgroup$
    – abelenky
    Oct 27, 2015 at 15:26
  • 1
    $\begingroup$ @abelenky Not only would this add another attack vector, hacking into a remote control system would likely be easier than physically hijacking a plane. $\endgroup$
    – FreeMan
    Oct 27, 2015 at 16:11
  • $\begingroup$ @Freeman has it right. Imagine a simple scenario where the runway is not safe for landing: covered with snow and ice, 50kts crosswind, and maintenance vehicles on the runway. If an attacker could issue the "normal" automatic-landing command, the landing would be worse than staying up in the air, and there'd be little the pilot could do about it. $\endgroup$
    – abelenky
    Oct 27, 2015 at 16:15
  • $\begingroup$ @Simon, I am not suggesting it as a solution. I am asking if it is possible in principle. Of course it could be a solution for a smaller plane with only one pilot who becomes unwell. $\endgroup$ Oct 27, 2015 at 16:49

2 Answers 2

  1. Yes for reasons of hacking prevention. There is no way to be absolutely sure any interface can't be hacked to crash the plane remotely into a suburb.

  2. Yes autopilots can follow a track of navigation marks to an airfield and can auto-land using the ILS.

But again there is no way to ensure only signals from people who have the best intentions with the plane can activate that mode.

  • 3
    $\begingroup$ Very, very few aircraft are set up so that the autopilot can extend flaps & lower the langing gear. Without a cooperative pilot in the cockpit to manipulate those controls, even Autoland can't land the plane successfully. $\endgroup$
    – Ralph J
    Oct 27, 2015 at 14:31
  • $\begingroup$ Valid, @RalphJ, however a no-flaps, no-gear landing at an airport with emergency personnel standing ready would likely result in fewer injuries than the same plane flowing into a building. Not that I'm supporting the idea... Also: running out of fuel will cause the plane to land successfully, if not safely. ;) $\endgroup$
    – FreeMan
    Oct 27, 2015 at 16:14
  • $\begingroup$ The DLR ATTAS does it already since 30 years, and there is no technical reason why this should not be possible on any airliner. $\endgroup$ Oct 27, 2015 at 17:14
  • $\begingroup$ @PeterKämpf "Could be" but isn't yet. It wouldn't be all that hard to design or retrofit equipment so the autopilot could tune the ILS, extend the flaps, arm the speedbrake and autobrakes, lower the landing gear, set the MCP altitude down to GS intercept altitude, engage the approach mode, and extend the thrust reversers after touchdown. But without that, you're looking, best case, at a coupled LNAV/VNAV approach to a high-speed (no flaps) gear-up crash landing on the runway. (Oh, with the engines still running, or burning, after the wreckage slides to a stop & pax start to evacuate.) $\endgroup$
    – Ralph J
    Oct 28, 2015 at 2:19
  • $\begingroup$ @RalphJ: The "couldn't be" is part of the question, so my comment refers to the possibility. You're right, in reality it isn't yet. $\endgroup$ Oct 29, 2015 at 10:57

In principle, no. But there are problems in practice.

  1. It is true that at present, the only way to stop a runaway aircraft is to shoot it down (unless there is some way of getting into the cockpit, like in movies).

  2. Modern aircraft (at-least the most recent ones) are programmed (via their fly-by-wire computers) to land automatically without any human interference (except for setting flaps etc., as already noted) and these can be programmed to do a landing on command. However, whether they can be used in practice is different altogether:

    • The autopilot landing is instrumented and is not possible in all the airports. The computer has to 'decide' where to land based on fuel situation, among other things.

    • What will happen if the runway contaminated? or one of the tires blow during touchdown? If the crosswind is more than what the autopilot is designed for? There are too many variables and in most cases it will do more harm than good.

    • And most importantly, what if the signal is sent by by someone intending to crash the aircraft? what if some one sends the signal (even accidentally) during takeoff?

Adding auto-land on remote command is more trouble than its worth.

  • 2
    $\begingroup$ Actually, it doesn't even need to have fly-by-wire. It 'just' needs an autopilot that can be coupled to the ILS receiver and that is certified for Cat IIIb operations. Lots of transport category jets before the age of fly-by-wire had operational autoland systems. $\endgroup$
    – reirab
    Oct 27, 2015 at 18:54

Not the answer you're looking for? Browse other questions tagged .