What it would do is provide certainty for the bereaved and confidence in the public. Having to deal with your loved ones simply vanishing must be an incredible mind fuck. No idea how they died, did they suffer, are they still alive, it is mind boggling to contemplate. Also, how do we know that the 777 does not have some fatal flaw. There was another cockpit fire in a 777 that was not very publicized. I think the hull was written off. But because it happened on the runway, everyone got out. If it happened in the air, I doubt they could have made it back. I think having the AC send a burst of data if transponders are switched off will be a good step. As things get more advanced maybe trigger a return to closest known airport, but you are right, other problems can be caused by automation. I suppose the pilotless plane would be the solution.
That's exactly my point, the public isn't confident because of one passenger aircraft in over 100 years of aviation. Perhaps the media, instead of pandering to the public, should be the voice of reason. I realize the media never has been the voice of reason, but the point remains. The experts (meaning the airlines, pilots and maintenance not the investigators) keep saying all these ideas for "always on" systems are bad ideas. The public should start listening to them. One incident like this in 100 years. I challenge the "public" that have a problem trusting aviation to show me a better record in their professions.
The 777 has been flying for 19 years, it's one of the best selling airliners of all time. If it had some "fatal flaw", we would've seen it already. I'm not saying this can't be a mechanical problem, and if it does turn out to be mechanical I'm not saying there won't be a service bulletin published demanding mandatory retrofits to address the issue, but in that case it would be an extremely rare occurrence given that we haven't seen it yet. 777's aren't going to start suddenly falling out of the sky because we don't fix a potential problem that brought down 1 airliner in 19 years.
Automated flight is a bad idea. Pilotless aircraft are a bad idea. Well, I should clarify, the technology isn't there yet. Good idea in another 50-100 years maybe, bad idea right now. Pilots solve more problems than they cause even though the number 1 cause of crashes is pilot error. Again, pushing for "pilotless aircraft" or even autonomous systems is just as bad a knee-jerk reaction. Why are you proposing such sweeping changes to prevent the recurrence of an incident that's happened once in over 100 years? Windshield wipers are just over 100 years old. Your proposals are akin to implementing a worldwide ban on windshield wipers after a single car crash is found to result from a fault in the wiper controller. What about the danger of not being able to see? Ignore it, because they cause an accident, let's get rid of them. That's the type of knee-jerk, sweeping changes you and others are proposing. It's ludicrous.
Here's what I propose. Look at the failures of automated systems in airlines. Look at things like Turkish Airlines Flight 1951, or Air France Flight 296. In the case of THY1951, a radio altimeter malfunction during autoland caused the aircraft to attempt to land to early. Complete disaster was averted by the pilots. While they should have recognized the problem sooner, it was a reliance on automated systems (the ones that would be 100% responsible in pilotless aircraft) that made them complacent. In AFR296, the alpha protection of the autopilot prevented the pilots from attempting to pull out of a dive. On one hand, alpha protection is designed to prevent stalls and stalls at 30 feet above ground are likely fatal if you aren't configured for landing, on the other hand, alpha protection is based on math and has a margin of error which is always on the safe side (meaning you almost always will have some alpha left before stall). It's quite possible they could have nosed up gently while waiting for engines to spool, cleared the trees during acceleration and recovered the aircraft. Also, let's not forget Air France 447, where the autopilot encountered an abnormal condition and turned itself off, because autopilots aren't able to accurately factor in all probabilities and locate the cause. Oddly enough, the crew correctly deduced the cause to be a stall but improper recovery was prevented by an inexperienced 3rd pilot, and then they moved on to look at other causes since the stall recovery failed thereby indicated the issue wasn't a stall. If the captain and primary first officer had been at the controls, the plane likely would've survived. A well-trained crew with proper crew resource management would've saved that aircraft, and I can assure you that Air France has taken drastic steps to avoid a similar problem again, as have many airlines.
Then there's the big factor to consider: the hairs on the back of your neck. On more than one occasion I've been doing an approach in bad weather using autoland, felt uncomfortable for whatever reason and observed my first officer looking antsy as well. I've expressed my concern that something didn't feel right, they have sometimes agreed, and without any known reason or cause, we initiated a go-around maneuver and came back again on a stabilized, manual approach. We have no idea if we were right or not since we landed safely and there were no errors reported in the landing systems both on the ground and in the aircraft, but who knows if things would've been fine had we continued with the autoland. Instinct and gut feeling is trained because instincts in flying in bad weather when you can't see out the window is so bad. We naturally want to do bad things. So we train and train and train, to the point that when you do feel uncomfortable, there's usually a reason even if you can't put a finger on it. An autoland system will never have that. It will never say, "Why do the controls feel sluggish?" It might say, "Verify control functionality - verified and within limits", where the limit is some arbitrary value set by some engineer. Sometimes you expect the aircraft to be sluggish, sometimes you don't. It's a feel thing, not a math thing. Sometimes that turns out to be nothing, something you find malfunctions that could've brought down the plane if you hadn't done something, and do it long before the computers register any kind of fault of malfunction.
Lastly, I'd suggest a review of Pacific Western Flight 314. Due to several errors on the part of both Air Traffic Control and the pilot, the aircraft touched down while there was a plow on the runway. The pilots noticed the plow and initiated an emergency go around. During the procedure the aircraft was left in a dirty state with clamshell thrust reverses not completely locked on one side when they returned to the air. As a result, the aircraft yawed violently in one direction and the throttles where automatically being pulled back by the force of the air against the mechanical linkages. The pilots fought so hard to maintain control they broke the bones in their hands and feet by pushing so hard on the controls. Ultimately, only 7 people survived, but this shows you the grit and determination a human pilot can have. What would a computer do? Up the amperage and burn out motors fighting for control? No, it would try to always fly within the preset parameters. Which in the case of PW314 likely would've meant barreling right through the plow. The pilots tried something. Ultimately we don't know how the outcome would differ, perhaps their maneuver resulted in a greater loss of life than would have occurred if they had just followed through with landing and crashed into the plow, but the point I'm making isn't whether they did the right thing or not, it's how hard they fought. In any given incident, pilots make snap decisions. Could you perhaps program the perfect set of steps to check and determine the cause? Possibly. But that would take longer than a human eliminating possible causes on instinct and experience and then exceeding the limitations of the human body to attempt to resolve the problem.
I believe I'll see the advent of pilotless aircraft in my lifetime for non-military use. I may or may not see it for commercial passenger aviation, that'll be the last category to go. But I do know we aren't there yet. Not by a mile. And planes will still be made to crash by humans even once that comes to pass. If someone really wants to down an airliner, they'll be able to find a way.
