[jervisfm]

It goes without saying that my qualifications in aviation matters are only as an aviation enthusiast and for the full cause of the accident, we'll have to wait for the NTSB report (which has access to vital black box data and cockpit recordings).

However, I did give this accident some thought as well given the circumstances of the accident and came the pretty much the same final conclusion as the OP.

An airline pilot I once spoke to told me once that all the big jetliners use automated ILS with autopilot to automate the landings. The system is pretty accurate and reliable and can deliver landings as good as a well-trained and experienced pilot can.

When I saw the rather steep final approach the aircraft was making, I had suspected that the pilots were doing a manual landing. Now that we have information that the ILS was temporarily inoperational this appears to be what transpired and the pilots manually flew the final approach.

The precise reasons for why the manual approach ended up in this accident are still to be determined. One thing that I will point out though is that pilots for modern jetliners do relatively little actual manual flying of the plane. Modern airplanes have advanced FMS (Flight Management System) that allow for automatic flying from lift-off until landing with pilots overseeing the system. Given this lack of manual flying, there is a greater potential for human error when a pilot is required to actually do manual flying.

[cjrp]

Pilots will actually very rarely use the autoland functionality (apart from when weather, e.g. low visibility, requires it, and also to ensure that they're current with the procedure). They'll use the autopilot for the initial part of the approach, but then when visual with the runway will often disconnect it in order to fly the actual landing manually (still following the indications on the flight director, which are provided by the localiser/glidescope signals).

[simpleigh]

Relevant link: http://www.freakonomics.com/2011/12/12/cockpit-confidential-...

"The touchdown had been a fully manual one, as the vast majority of touchdowns are."

[jevinskie]

Why is this? Can they do a better (in terms of passenger comfort) job landing manually?

[cjrp]

I couldn't say in terms of comfort, but they need to be able to perform a manual landing (for example if certain equipment on board is not functioning correctly, or the airport is not equipped for autolanding) so it makes sense that they would take every opportunity to keep their skills current. There's probably also a sense of pride in wanting to carry out a smooth landing by hand - especially given that most of the climb/cruise/descent will likely have been controlled by the autopilot.

[lisper]

No, autoland is pretty good. But there are two factors that cause pilots to land manually most of the time: 1) pilots think they are better at handling last-minute unexpected contingencies like wind sheer or bird strikes (and they may well be right about that) and 2) pilots don't want to give the Powers that Be any additional data to suggest that they aren't actually needed to fly a plane at all. The fact of the matter is that there are no technological barriers to making aircraft completely autonomous, but pilots want the world to remain in denial about this for as long as possible for the sake of their job security. (FWIW, I'm an instrument-rated private pilot.)

[sneak]

> The fact of the matter is that there are no technological barriers to making aircraft completely autonomous, but pilots want the world to remain in denial about this for as long as possible for the sake of their job security.

Two words: anomalous conditions.

As we both know, pilots aren't needed in the vast majority of commercial flights. The plane is perfectly capable of taking off, navigating, avoiding collisions, and landing with minimal, mostly unskilled supervision.

We don't need pilots on planes... until something unexpected goes wrong, and we do.

https://en.wikipedia.org/wiki/US_Airways_Flight_1549

https://en.wikipedia.org/wiki/Gimli_Glider

https://en.wikipedia.org/wiki/United_Airlines_Flight_232

[lisper]

That's true, but 1) anomalous conditions are very rare and 2) as often as not the pilot is the anomalous condition. Pilot error is the most common cause of aviation accidents.

EDIT: Notice that you had to go back 30 years to find your third example of a human pilot saving the day. And automation technology has improved a lot since then.

[snom380]

Most of the time when a pilot saves the day he just disconnects the autopilot and writes an incident report. There's been at least four inflight upsets due to autopilot or sensor malfunction in the past few years that I can recall. Some of these caused serious passenger injury. Latest one I heard about here: http://www.aibn.no/Luftfart/Rapporter/13-18 Another one: http://en.m.wikipedia.org/wiki/Qantas_Flight_72

[CamperBob2]

Devil's-advocate: fine, OK, we still need pilots, for the time being. But do they actually need to be on the planes?

In the case of two recent crashes (AF447 and this one, assuming pilot error turns out to be the cause as seems likely), it's safe to say that a computer would've easily done a better job than the pilots.

Would be interesting to hear Capt. Sullenberger's take on the question.

[bdonlan]

If you're going to use remote pilots to deal with anomalous conditions, you'd better have an emergency electrical supply that'll keep your remote control running for however long it'll take to glide at the plane's maximum range. You'll also need a communication system that'll work over the ocean, in poor weather conditions, while the plane is having difficulty maintaining a stable attitude (good luck maintaining the alignment of a satellite dish when the autopilot disconnects due to a sudden upset).

On the other hand, if your onboard autopilot is in control of everything, you run the risk of having autopilot bugs causing an accident - indeed, a software issue in the airbus flight protections (which, unlike the autopilot, are always active unless a very severe system failure occurs, and can override the pilot's inputs) has caused an incident that resulted in passenger injuries: http://en.wikipedia.org/wiki/Qantas_Flight_72

While none of this is impossible, it's certain to be expensive, heavy, have its fair share of bugs at the start, and by quite difficult to get past the regulators. Not to mention passengers probably won't be very happy being on an unmanned plane, even if the system really was perfect. So it's easier in the end to just put pilots on the plane.

[snom380]

Assuming we could make such a system secure and unhackable as well as unjammable, why would we want to? To save some money? I would rather pay a little more to have the pilot sit in the same airplane and have his life on the line as well.

I just find it fascinating how obsessed people seem with getting rid of pilots. You don't often hear the same discussion about train or bus drivers, or doctors for that matter, even though we are just as close to having automated robotic surgery and expert systems have been around for a long time.

[revelation]

I guess latency is why they are on the plane.

[kenkam]

Another big reason for piloted planes is, seriously now, who here would get on a plane knowing that it was being flown without a pilot.

If it's hard for most of us from varying technological backgrounds to say yes than it's going to be almost impossible to convince the more mature passengers.

This is one of the reasons why all planes look the same (they have the same configuration, low wing with engine below wing) even tho Airbus and Boeing keep churning out concepts. Would you get on a plane that looked even slightly funky?

[bradleyland]

> ...seriously now, who here would get on a plane knowing that it was being flown without a pilot.

Given that the most common cause of aircraft accidents (by quite some margin) is human error, I'll take my chances with the computer.

[snom380]

I don't find it likely that pilots hand-fly approaches to make a point. Isn't it more likely that this is codified in airline operation instructions?

As for completely autonomous, do you seriously think we have the technology available today to make an autopilot handle any conceivable situation without a human being present and ready to take over? (non-IR PP here)

[lisper]

It isn't "just to make a point". It's also to stay proficient, or to make the job less tedious. But whether to hand-fly an approach is always the pilot's discretion (as far as I know -- I'm a just a private pilot so I could be wrong about that).

And no, automation can't handle "every conceivable situation", but neither can humans. Furthermore, humans screw up more often than autopilots. Pilot error is currently the single biggest contributor to the overall accident rate.

Mind you, I'm not advocating fully autonomous aircraft. I like having a human in the loop, but that's in part because I am the human in the loop. It's far from clear that human pilots are a net win for safety.

[snom380]

Neither can humans, but humans can handle a lot more situations. We can take into account a lot of information and come up with creative solutions, while current computers will need a set of humans to come up with and program potential scenarios beforehand. I think we'll have computer level AI that will perform better in accidents eventually (AI is improving and our brains are not) but it a long way into the future, and not an imminent threat to pilot jobs as some would believe.

In the meantime, the solution to the human factor isn't to eliminate humans, but to improve training (which is already happening after AF447).

I like having human pilots primarily because I'm a programmer and I know how difficult it is to design robust computer systems. There's been one runway overrun and one serious in flight incident with passenger injury due to software design faults. Now try to design a system that makes sense of audio, video and smell in addition to the existing sensors, and not have it fail in some spectacular unforeseen way..

[aghull]

I can't imagine anyone stepping onto an autonomous passenger aircraft for a long time, not until long after cars and smaller commercial aircraft are regularly driven using computers only. And even that seems like a stretch. I don't see how pilots use or non-use of some automated systems is going to influence popular opinion on that one way or another. But maybe there's a perception among pilots that it will.

[lisper]

The fact of the matter is that most commercial jetliners today are fully automated. It is only takeoff and landing that are still done by hand. No one ever hand-flies a jetliner during cruise except in dire emergencies. In fact, at high altitudes, where the plane is flying in the "coffin corner" of the flight envelope (http://en.wikipedia.org/wiki/Coffin_corner_(aviation)) hand flying can be extremely dangerous.

[damian2000]

Getting in a Google self driving car is one thing, where you still have the ability to take over control or apply brakes yourself. But how many passengers would be willing to get on a fully self piloting aircraft? I think not too many, at least at first.

[shin_lao]

They are bound by regulations to do a certain number of landings every year.

[will_asouka]

I'm a widebody captain, maybe I can add some insight.

The handling of the aircraft is a small part of the job. Autoflight systems have been capable of flying and landing since the 1960s. It is a mature technology although it does need to be monitored by humans as equipment failures do occur. I am confident a system could be developed with sufficient fail safes to remove the requirement for monitoring, and a fully automatically flown aircraft could be built tomorrow although at a higher cost than the current monitored setup.

Even if an unmonitored autoflown aircraft was developed, judgement decisions about the operation of the aircraft still need to be taken, and these are best done by a human. Examples of decisions I make:

1. The best fuel consumption maybe achieved at a higher flight level, but I elect to stay low and burn into contingency because I've heard reports of turbulence at the higher level. I judge that I probably won't need the contingency fuel later based on previous experience of the route and destination. So I give my passengers a smooth ride, which helps repeat custom for the airline.

2. I experience radio crackle whilst in cloud and elect to change my routing as I suspect we are near a developing thunderstorm that may soon become active. I have to balance the extra fuel burn of a non-optimal route versus the risks of being caught in a thunderstorm which could destroy the aircraft.

I'm not an AI expert, but it seems to me these are value judgements that are best made by a human. So the fully autoflown aircraft will still need a human in the loop at some point.

It is possible to remotely site this human, and maybe make efficiency savings by a single human controlling many aircraft but this would require a bulletproof, near realtime datalink. Again, I'm not an expert in the field, but I would suggest this is, if not technically impossible, then a lot more expensive than the cost of employing human pilots. I imagine this datalink would also be of interest to people with ill intent- no fun when you get the airborne equivalent of a rickroll.

TL;DR: A human is required in the loop, not to handle the aircraft but to make decisions. It's cheaper to have them on-board than the cost of a datalink. It's cheaper to have them trained to take manual control than the cost of an autopilot sufficiently fail safe that it can be left unmonitored.

[dopamean]

My father is a commercial airline pilot. He flew Boeing aircraft for 15 years and for the last 10 has been on a variety of Airbuses. He tells me that pilots very, very rarely use the auto land features and that he has never used it himself.

[gcv]

So the price of automation is eternal vigilance? Reminds me of the old argument that overuse of calculators means people have lost the ability to do mental arithmetic.