Asiana Airlines Flight 214: A Pilot’s Perspective | Hacker News Reader

Asiana Airlines Flight 214: A Pilot’s Perspective

slate.com

271 pointssheri170 comments

Off topic but of significant hacker interest:

Also in 1985...the crash of a Delta Air Lines L-1011 in Dallas that killed 137.

There's always been a lot of speculation over how IBM gave up an entire industry (the personal computer) to a seemingly insignificant start-up (Microsoft). One theory is that may not have happened if Don Estridge hadn't been killed in that crash. More about him here:

http://en.wikipedia.org/wiki/Philip_Don_Estridge

and here:

http://www.nytimes.com/1985/08/05/us/philip-estridge-dies-in...

Of particular note:

"His most revolutionary move was to make the computer's design specifications public, allowing thousands of young people to write programs for the machine."

We take it for granted now, but before Don Estridge and the IBM PC, "proprietary" was the rule and "open" was the exception. RIP

Now I understood the policy not to book more than 4 employees on same plane. RIP

depending on the size of the company, it could be no more than 1.

i was in a car accident with my ENTIRE team one time. minor, but scary.

to quote ... shit happens.

I'm a former Google employee, they had a similar policy. There were no more than x employees per team allowed on the same plane, above a certain seniority level, only one per plane.

The irony is that there's no such policy for buses, which in many ways are less safe than planes. One catastrophic bus crash on the way to Tahoe could conceivably wipe out most of search quality, or any other department.

Traveling by bus is surprisingly safe – measures in journeys and travel hours even safer than air travel:

http://www.numberwatch.co.uk/risks_of_travel.htm

Unless you take the notorious chinatown bus[0].

0. http://en.wikipedia.org/wiki/Chinatown_bus_lines#Safety

how many bus accidents in the US in the last 20 years did you hear of where every single person died?

http://en.wikipedia.org/wiki/In-N-Out_Burger#Second_generati...

Such an accident prompted changes in the time between large and small planes landing.

"One theory is that may not have happened if Don Estridge hadn't been killed in that crash."

Fate seemed to be sealed (or on it's way) much earlier than that (based on my memory, and supported by an obvious link below.)

http://inventors.about.com/od/computersoftware/a/Putting-Mic...

"Bill Gates then talked IBM into letting Microsoft retain the rights, to market MS-DOS separate from the IBM PC project, Gates and Microsoft proceeded to make a fortune from the licensing of MS-DOS"

That was of course way before Estridge got killed.

And Compaq came out in 1982...

http://en.wikipedia.org/wiki/Compaq

"In November 1982 Compaq announced their first product, the Compaq Portable, a portable IBM PC compatible personal computer."

This is a great article, with one exception:

There have been a handful involving regional affiliates, but the majors have been virtually accident-free

Now, I realize that even including those the stats are still fairly good, but it makes the "virtually accident-free" stats of majors a result of spinning off risk to affiliates, which is not something a consumer should care about. It might be a bit of a nit-pick (as it's still a handful), but moving things into different brands to avoid stigma is a common and, in my opinion, unhealthy practice.

Agreed. Those two flights that crashed were flown as Delta Connections and Continental Connections, and booking a flight on the "majors" to a smaller market could easily book you on one of those flights for a shorter leg. It's not like you can just fly majors only to be safer or anything.

Indeed... In fact, if you asked the passengers of Colgan Air Flight 3407 what airline they were flying, I suspect the vast majority would say 'Continental'

And the best comment on the article points out that this is even worse than it appears:

=========

"I'm sorry, but it's just wrong to say that. Yes, it may have technically been a ComAir flight that crashed in 2006 in Lexington, killing all 50+ people aboard (aside from the first officer), but those people booked a Delta flight, not a ComAir flight, so why make the distinction? And, yes, it may have been a Colgan flight that crashed in Buffalo in 2009, also killing 50 people, but they booked their flight on Continental. Major airlines ferret out their regional connections to these airlines all the time, and you don't know you're on one unless the flight attendant announces it (or you just happen to know about aviation), so why on Earth is this guy making a distinction?

I'm a fairly frequent flyer on USAir (enough for mid-level status) and I can't tell you how often I book a USAir flight but I'm actually flying Republic or Chautauqua. Yes, air travel is safe, but let's not ignore reality just to make a (false) point. There have been recent and very deadly crashes since 2001."

My experience is that the typical online flight booking websites show when a flight is serviced by an affiliate. That's typically a clue that the flight will be on a smaller regional jet.

Every regional flight I have booked through a major airlines has prominently shown both the affiliate airline and the type of aircraft.

I think the point remains though. The average consumer doesn't know the difference between a "major" airline and a "regional" one and shouldn't need to. Their safety record is just as applicable. Its not as if a consumer will see a regional affiliate and change their reservation because of it; its effectively the same experience.

You're right, but it's important that a smaller aircraft implies fewer potential casualties. The best thing would be to rate accidents by the log of the actual number of casualties, but that's not too intuitive.

Could it be that rather than a distinction between the big airlines and the small airlines, the author was attempting to make a distinction between flights with large amounts of passengers (100-200+) and flights of smaller size (50ish)?

I think the point is that there was a streak of 12 years with no fatal incidents of a large-size jet.

...in North America, which is just more fact-twisting by the author.

Lets not forget that 228 people perished when the flight crew of Air France 447 flew the plane straight into the ocean.

Well didn't really fly so much as fell out of the sky into the ocean. Absolutely terrifying.

If by "flew" you mean "stalled", then yes.

Well, it wasn't your standard controlled flight into terrain, but since one of the pilots was pulling back on his hand stick for the entire time.... (e.g. https://en.wikipedia.org/wiki/Air_France_Flight_447#Sullenbe...).

Of course, he could have said that instead.

Great and sensible article. Read this because it "pops the bubble" of some of the popular memes floating around, providing details on what it means to have low hours in a type, or whether Korean pilots have "perceived seniority" issues in the cockpit.

I note that checked bags cannot block passengers from escaping an aircraft. Perhaps someone will someday quantify the additional risk airlines forced on passengers when they started charging for every bag.

> I note that checked bags cannot block passengers from escaping an aircraft.

Perhaps not, but people standing in the aisle reaching to get them, dropped bags people trip over and block the aisle, etc do.

Checked baggage is not accessible from within the cabin.

Checked bags, as in baggage stowed in the luggage compartment and not what you carry on.

Patrick Smith wrote the "Ask the Pilot" column for Salon.com for 10 years, ending last year. His writing was always very insightful and very entertaining at the same time.

One of my favorites: http://www.askthepilot.com/essaysandstories/the-exploding-to...

It seems he came back to slate.com as a one-time thing, but you can read his archived work at Salon and at www.askthepilot.com. He seems to have a book as well now.

That's everyone's favorite "Ask the Pilot".

He also has a new book on Amazon that is basically the book version of his Ask The Pilot column.

I actually just read it the other day and was pretty interesting (and a quick read)

He now blogs semi-frequently on boston.com:

http://www.boston.com/community/blogs/askthepilot/

"I note that checked bags cannot block passengers from escaping an aircraft. Perhaps someone will someday quantify the additional risk airlines forced on passengers when they started charging for every bag."

I don't think this is as big a problem as one makes it out to be and here is why. There are certain things I need when I travel that I can't afford to go without in the case my checked bags don't make it to my destination at the same time as me. Things such as prescription medicine, contacts and my retainer (though not nearly as important as the previous two but still important none the less) are things I always carry on because they are things I can't easily replace if my bags are lost. Then I bring the entertainment, usually a laptop and/or a tablet and books/magazines. All of that stuff would need to be in at least one carry on.

Most, if not all, of the airlines I have flown have a limit to how many bags you can carry on. So it isn't like charging for check bags is going to have people start bringing 10 carry ons, it would have them bring the max if they wanted which as far as I know is 2 for most airlines.

But it still definitely incents passengers to carry on more, larger and fuller bags.

I think everyone else is misunderstanding your point. You're saying someone should quantify the additional dangers and costs to society associated with an increase in carry-on luggage which happened at the same time airlines started charging for checked bags.

So now that checked bags cost money, people bring more and larger carry-on luggage. This additional carry-on luggage:

1. increases boarding times. 2. increases injuries from passengers and flight attendants lifing heavy bags into overhead compartments. 3. increases the amount of crowding on a cabin from under seat carry-ons thus increasing evacuation times.

You cannot have all the bags checked, as the checked luggage is treated like dirt. I usually travel with $6000 photo equipment, I would not check it even if they pay me.

Custom cut the foam inside http://www.pelican.com/ cases to your bodies/lenses and your photo gear will be fine in anything short of a plane crash. And, in that scenario, odds are it will come through a lot better than anything you pack in overhead.

For your own peace of mind, please never put expensive gear in checked luggage. It is very likely to get stolen.

It's not practical (or even possible) to carry on all of your gear all of the time. Our field engineering team(s) have been travelling for 5+ years with > $100K worth of checked electrical gear to Brazil, London, Singapore, Germany, Hong Kong, and Malaysia.

A couple times gear was delayed, and so the FE team had to wait a day or so for it to arrive, but nothing has ever been stolen.

Nice thing about a Pelican Case - no straps to get tangled up in any baggage sorting gear, and large enough that it doesn't usually get stuck in the corner somewhere.

From a probability perspective, much more likely to get mugged coming out of the airport, or getting hijacked using your equipment in the field - this has happened to our Field Engineers multiple times, but no losses in the airplanes (yet).

Anecdote != data.

You put $100k of equipment in checked luggage, then you probably have some very specific insurance that covers you. If you don't, you are lucky. Most people most of the time do not have anything stolen.

If a baggage handler steals a $3000 camera out of someone's luggage, the airline will totally deny liability.

Here's an article interviewing a former TSA handler who was arrested with more than 80 stolen cameras and other assorted electronics in his home at the time of arrest. http://rt.com/usa/tsa-stealing-from-travelers-358/

> But consider for a moment the year 1985, one of the darkest ever for commercial air travel. By the end of that year, 27 crashes had resulted in the deaths of almost 2,400 people. These included the Air India bombing over the North Atlantic, with 329 casualties, and, two months later, the crash of Japan Airlines Flight 123 outside Tokyo, with 520 dead. (These, the second- and fifth-deadliest incidents in aviation history, happened 49 days apart.) Also in 1985 were the Arrow Air disaster in Newfoundland that killed more than 240 U.S. servicemen, the infamous British Airtours 737 fire, and the crash of a Delta Air Lines L-1011 in Dallas that killed 137.

Wow, the number of fatalities that year is incomprehensible by today's standards. Yes, this accident has made big news because so few accidents have happened in this decade...but I'd say that today's news cycle has served to amplify just about everything, no matter what their statistical, relative, or absolute significance.

So it's ok to forgive Pilots' incompetence just because statistics say it doesn't happen a lot?

Is it OK to forgive the dozens of incompetent car drivers who kill people on a daily basis? And if not, how would you feel if each fatal accident was given such national news treatment (if there were enough bandwidth to do so, which, theoretically, there is)

My comment is not about the culpability of the pilots, but the long recognized psychological phenomenon of overestimating risk probabilities due to increased "hearing of" incidents, rare or not, the cause being irrelevant.

Because each driver is not responsible for 300 lives in their cars. That's why there is so much attention to plane crashes. It doesn't kill one person at a time.

Besides, it's incredible such an accident happen when the weather is perfect, there is no wind, it's not at night, the plane is working just fine, and it's just a couple of pilots who don't know what they are doing because they don't know how to land a plan manually since they are used to pressing buttons the whole time and relyong on autopilot. That's what the news is about.

Just like the AF447 crash between Rio and Paris: Pilots incompetence at its best.

Did you read this article? And if so, do you have any more well thought out arguments against that author's point? His whole point is that this type of landing is not uncommon and there is little reason at this point that it had anything to do with the pilot's experience. It's still possible, yes, but you're making statements that indicate you understand neither the article nor the comments to which your are responding. People are explaining why there is a media frenzy around the accident, not why it's okay. Trolling, at its best.

Yes, I have read the article, and this final statement is ignorance at its best: "Whatever happened on final approach into SFO, I highly doubt that it was anything related to the culture of Korean air safety in 2013. Plane crashes are increasingly rare the world over. But they will continue to happen from time to time, and no airline or country is 100 percent immune. "

There were clear indications from the communication with the air tower that the pilots did not communicate any emergency situation until 1.5 seconds before the crash, and that the co-pilot was apparently silent in the cockpit. Check the Los Angeles times article about that. It's clearly linked to cultural aspects, all over again.

I do not make this accusation lightly - but this strikes me as a particularly racist thing to say. Your argument boils down to "because the co-pilot was silent during the crash this points to deficiencies in Korean culture".

Which is a wild leap no matter which way you cut it.

But I mean, congrats I guess. You were looking for a conclusion and unsurprisingly you found it.

Where did i imply racism? Culture has not only to do with nationality, its also part of education, training, and skills when interacting with others. This kind of accidents can happen in all cockpits where there is no open critique and communication.

That's exactly what we had in the 80s, a deficient culture in the cockpit, a bunch of hot shot pilots who 'didn't want to be told what to do' and would ignore co-pilot advice. I'm not surprised another country has these same hurdles to overcome.

But Korean Air at least, and Korean pilots have already faced that hurdle (if they aren't already the poster children for culture issues among pilots), and if it comes down to a cultural issue again I'll be more than a little surprised.

Can't find the link I originally saw that went into this in more detail, but the basic subject is covered here: http://blogs.wsj.com/middleseat/2008/12/04/malcolm-gladwell-...

More on the business side here: http://www.nytimes.com/2002/03/26/business/new-standards-mea...

Prepare to be surprised:

http://www.pprune.org/rumours-news/518568-asiana-flight-cras...

That's exactly what I have been talking about and I was voted down to hell for claiming that out loud.

In my opinion, you were getting voted down to hell because you made very inflammatory remarks that had nothing to do with the parent comment. Somebody remarked on how the relative lack of airline disasters in recent history probably amplified the media frenzy this time around, and you said "So it's ok to forgive Pilots' incompetence just because statistics say it doesn't happen a lot?"

Don't play the martyr. If you had posted that link as an entirely separate comment and suggested that it was a major factor in the crash, no one would have down voted you, I'm sure. But because of that first comment that was, IMO, a huge jump in both tone and topic from its parent, people probably perceived your subsequent comments as being irrational and motivated by emotion. It was nothing to do with what you were talking about, and everything to do with how you were talking about it.

Sure, but that's not the part I had a problem with. Korean Air, after all, had an incident in the 90s where the captain overruled the strenuous objections of his copilot, resulting in a fatal crash.

The main issue I had with the post was that he's claiming this to be the case purely on the notion that the copilot did not speak during the crash. That's severely reaching, and smells a lot like fitting facts to a predetermined conclusion rather than the other way around.

Here's an extract from a pilot's post who was instructing in Korea. Don't take my word for it, take his:

"The Koreans are very very bright and smart so I was puzzled by their inability to fly an airplane well. They would show up on Day 1 of training (an hour before the scheduled briefing time, in a 3-piece suit, and shined shoes) with the entire contents of the FCOM and Flight Manual totally memorized. But, putting that information to actual use was many times impossible. Crosswind landings are also an unsolvable puzzle for most of them. I never did figure it out completely, but I think I did uncover a few clues. Here is my best guess. First off, their educational system emphasizes ROTE memorization from the first day of school as little kids. As you know, that is the lowest form of learning and they act like robots. They are also taught to NEVER challenge authority and in spite of the flight training heavily emphasizing CRM/CLR, it still exists either on the surface or very subtly. You just cant change 3000 years of culture.

The other thing that I think plays an important role is the fact that there is virtually NO civil aircraft flying in Korea. Its actually illegal to own a Cessna-152 and just go learn to fly. Ultra-lights and Powered Hang Gliders are Ok. I guess they dont trust the people to not start WW III by flying 35 miles north of Inchon into North Korea. But, they dont get the kids who grew up flying (and thinking for themselves) and hanging around airports. They do recruit some kids from college and send then to the US or Australia and get them their tickets. Generally, I had better experience with them than with the ex-Military pilots. This was a surprise to me as I spent years as a Naval Aviator flying fighters after getting my private in light airplanes. I would get experienced F-4, F-5, F-15, and F-16 pilots who were actually terrible pilots if they had to hand fly the airplane. What a shock!"

The whole post : http://www.pprune.org/rumours-news/518568-asiana-flight-cras...

People, please read the whole linked post before passing judgement. It sounds highly believable, from someone who was teaching pilots in Korea very recently (2003-2008) and makes the claims of a fantastic turnaround success story sound mostly like a very ineffective re-education campaign that has failed in the face of an entrenched educational culture.

Because each driver is not responsible for 300 lives in their cars. That's why there is so much attention to plane crashes. It doesn't kill one person at a time.

Right.. and this crash only killed 2 people. Perhaps one if the reports of the girl being hit by a rescue vehicle are believed; so what's your point? The real story is just how better aircraft safety has come over the years. This same accident years ago could very well have killed everyone aboard.

Just like the AF447 crash between Rio and Paris: Pilots incompetence at its best.

...and a week after this accident everyone was blaming the weird weather that happens at the equator for the crash. The point is that everyone is speculating, even you, that this this accident is the pilot's fault, but more often than not it's found to be a combination of mechanical failure and the pilot not being able to respond to it quickly enough -- which just so happens to be the official cause of AF447 crashing in 2009.

It's obvious they have been very lucky on the landing. It could have turned out in an explosion and kill half of the passengers. It doesn't excuse anything.

It certainly doesn't excuse your profound and distressing ignorance and belligerence. The facts are that you are not a pilot, were not there, have not received an official statement from the parties involved and are merely voicing an opinion based on what little information the mainstream press has been able to glean in the time since the accident.

Air disaster investigations typically take a while. Most reach conclusions that are extremely non obvious, nor intuitive. I recommend, for the sake of my own sanity and the rest of the people you inflict your opinion on, that you read the following books:

http://www.amazon.co.uk/Air-Disaster-Vol-Macarthur-Job/dp/18...

http://www.amazon.co.uk/books/dp/1875671196

There is a third volume available as well if you can stick it out, but perhaps you may gain some much needed perspective.

Doctors make mistakes. Programemrs make mistakes. Pilots make mistakes.

In fact, if you can actually get an airline pilot in a relaxed situation where he's certain you won't rat him out, you'll hear lots of stories about stupid mistakes. The fact that this inevitable human error almost never results in fatalities is a testament to the level of redundancy in air traffic procedures.

As a case in point, I talked to a pilot in one of Europe's major low-price carriers. They had an incident where both pilots literally fell asleep in the cockpit and were intercepted by German jets. Apart from the jet fighter escorts, the passengers never knew anything was off.

"plane is working just fine"

This you know how?

See https://news.ycombinator.com/item?id=6013520 It certainly does not look like plane failure.

Pilots suggest plane slowed despite presence of auto-throttle: http://www.cnn.com/2013/07/10/us/asiana-airlines-crash/

My point is, it is premature to rule out aircraft problems.

This is about how we deal with the news, not how an airline should discipline pilots who screw up.

I don't believe that is even close to what he was saying.

This is orthogonal to the original argument, but yes, we should forgive pilot error, because mistakes are not necessarily a demonstration of incompetence.

Life, and all subcomponents of life, is one giant ball of probabilities. Training, preparation, and planning all go towards decreasing the probability of undesirable events, but they do not rule them out entirely.

This includes plane crashes, car crashes, or even scraping your knee.

Professional football players don't play perfectly every game. Olympic athletes still make mistakes despite years of daily training. And you still have major bugs in your codebase - would you describe yourself as incompetent?

Mistakes happen. Accidents happen. So long as we can confidently determine that it is accidental rather than systemic (i.e., incompetence), we learn from it and keep going.

But nevermind all that. Let's hang a sign around their necks, parade them in the streets, let people throw things at them, and then execute them publicly.

Perhaps more to the point, no single error should ever destroy an aircraft. As the parent points out, people make mistakes. This is not news, and good systems (including aircraft) are designed to allow for that. So, even if pilot error was a contributing factor, I find it very hard to believe that nothing else went wrong. There were at least two people in the cockpit as well as controllers on the ground. No one was worried until a few seconds before the crash.

Agree that a single, simple error has little chance to down a modern transport aircraft. It's almost always a chain of errors, each necessary but only jointly sufficient.

I realize that not all the facts are in, and the NTSB full process will take quite some time, but I believe that once all the facts are in, you will find this to be a stunning display of lack of basic airmanship skills.

Why would no one on the ground warn the flight? The altitude transmitted to ATC radar is in hundred foot increments and the high and fast on initial won't trigger an alert, once they pull the power off and try to capture the appropriate vertical profile, ATC radar has no information that anything is wrong, then when they start blowing through the vertical profile, there's not enough resolution to trigger a low altitude alert.

More fundamentally perhaps, it's not ATC's job to fly the plane. ATC keeps planes separated; pilots fly them. On a clear-and-a-million day, and an 11,000 foot runway, the crew couldn't manage a visual approach. There was more than one error, but I think the NTSB findings on probable cause will overhwelminginly find the failures to be in the forward portion of the aircraft, immediately behind the yokes. (It is hard to fathom.)

Thankfully the NTSB and other accident investigators do not usually come out with a single easy cause in a newsbite size. They will examine lots of contributory factors, and they will come up with suggestions that fix those - typically dozens or more. In aviation they talk of "swiss cheese" theory - slices of cheese with holes in them. The idea is to have many in such a way as holes don't line up.

As you say, a failure like this is not "pilot error" but rather a failure of the aviation system and that is what needs to be corrected for all aviation. The "pilot error" is a symptom of the problem, not the cause.

For everybody talking about pilot error, please note that none of us have full info. Just as another datapoint read this http://www.dailymail.co.uk/news/article-2357585/Crash-landin...

Same type of plane( yes different type of engines) , same type of issues while landing. I am not suggesting that sfo crash was due to engine failure. I am just suggesting that we all should wait for the investigation to conclude.

While we don't yet know the details, have enough information to say that 1) the approach was unstabilized, 2) the engines provided thrust when they initiated a go-around (too late), 3) no emergency was declared before the accident happened unlike the BA accident. The only thing in common with these two accidents is that they are the same model of airplane and that both landed short of the runway. We don't know whether it was pilot error, but it looks very unlikely to be the same problem, especially since Boeing reviewed all 777 aircraft after that accident. A third option of course is that the root cause was not found in the earlier investigation and that there is a different problem affecting the type, but so far pilot error seems to be the most likely cause.

It's not the same problem, as the engines responsed to thrust. The pilots applied thrust 8 seconds before impact, and the engines revved up to about 50% in that time.

That's exactly what I'm saying :)

Do we know if that is a reasonable rate of acceleration for a 777 engine?

Highly knowledgeable people have commented on it and not said it was anything unusual, so, yes, I'd say it's reasonable. Jet engines don't change their output on a dime.

>Highly knowledgeable people have commented on it and not said it was anything unusual, so, yes, I'd say it's reasonable.

You know, there's not a single thing about your comment that's useful.

>Jet engines don't change their output on a dime.

Yeah, I know. That's the reason for the question.

What's not useful about that? Experts commenting on their field of expertise don't count for anything, or what?

This specific item has been addressed by people who know this stuff. They didn't indicate it was abnormal. How is that not useful?

>Highly knowledgeable people

Who? You're some guy on the Internet. I should take it on faith that you can tell who is knowledgeable and who is a blowhard?

> have commented on it and not said it was anything unusual, so, yes, I'd say it's reasonable.

Again, instead of providing a fact I have to trust the judgment of a person I do not know.

> Jet engines don't change their output on a dime.

It should have been obvious from my question that I know that much about turbines. Do you have any information specific to 777 engines or something comparable?

>Experts commenting on their field of expertise don't count for anything, or what?

They might. It depends who they are. I don't always believe everything I see on television or the Internet. Sometimes I like to know their reasoning.

Why the hell are you posting on an internet forum if you're not interested in what "some guy on the internet" has to say?

It seems that regulations call for 6-7 secs from idle to full thrust.

1. http://www.pprune.org/archive/index.php/t-242588.html

Thank you, that's the kind of answer I was looking for.

>Boeing reviewed all 777 aircraft after that accident.

Boeing, the same company who months ago reviewed their 787 batteries and found no reason to ground their planes?

It's hardly "another datapoint" when it's armchair speculation by a tabloid journalist from an infamous scandal rag. Admittedly it's actually not that badly written, but for context his other article published that day features nightmarish visions of San Francisco under 25 feet of water...

The initial conclusions from the US National Transportation Safety Board is that the engines were functioning normally; i.e. a totally different issue while landing. I doubt a Seoul-San Francisco flight in mid-July was likely to have experienced the same unusually cold temperatures that resulted in the BA038 fuel droplets freezing (over Siberia in January) either

> Yes, this accident has made big news because so few accidents have happened in this decade...

And there is video footage of it happening. Never underestimate the how much excitement there is in news rooms where there is actually video of the news that isn't just people talking.

Yes the news cycle amplifies everything but when they have video the volume goes to 11 (or beyond)!

So true. There was another plane crash that killed ten people in Alaska over the weekend...even though it was more deadly it hasn't gotten the same coverage, in some measure because there's no video of the accident.

I think the issue is that major airline crashes are so relatable since many people fly on a major carrier every so often. "It could've been me" plays. On the other hand, very, very few people take air taxis in Alaska.

There are still quite a lot of bigger accidents, but since they happen in Africa and Asia, they're not that big news in the West.

For example, 2012: * April 20 – Bhoja Air Flight 213, a Boeing 737, crashes near Chaklala airbase, Rawalpindi, Punjab, Pakistan reportedly due to bad weather, killing all of the 127 passengers and crew on board. * May 9 – In the Mount Salak Sukhoi Superjet 100 crash, a Sukhoi Superjet 100 crashes into Mount Salak, Indonesia on an exhibition flight, killing all 45 passengers and crew on board. * June 3 – Dana Air Flight 992, a McDonnell Douglas MD-83 carrying 146 passengers and 7 crew members crashes in a suburb of Lagos, Nigeria on approach to Murtala Muhammed International Airport, killing all on board and an estimated 6 more people on the ground.

Of course, air safety has been greatly improved, but it's not like these larger accidents are unheard of.

> Early theories as to why a plane crashed almost always turn out to be wrong or incomplete.

In the Air France crash over the Atlantic, the early theories all centered around the loss of airspeed indicators--a mechanical problem. When the flight recorders were finally found (an astounding feat of salvage BTW), it turned out that the airspeed indicators did fail...but then they came back online, and the pilots flew a perfectly good aircraft into the ocean anyway.

So as we look at this SFO crash, the lessons from Air France are mixed. On the one hand it's possible that the early theories of pilot error will be wrong, and it will be some subtle or previously unknown software or hardware error. On the other hand we know that modern highly trained pilots can make enough wrong decisions in a row to crash a sound airplane.

Reading the account of that crash was scary. One pilot diving, one pulling up. and as a result the plane crashed.

For those who want to read the story: http://www.telegraph.co.uk/technology/9231855/Air-France-Fli...

Seems like the "I have control" "you have control" thing went out the window pretty quickly. Very chilling account to read.

I've always considered that a cautionary tale on interface design.

The linked yokes in an Airbus to average their inputs without force feedback, unless one yoke or the other is explicitly disabled. The experienced pilot had no idea that his co-pilot was stalling the plane the entire time.

Yeah. Seems it would make more sense to mechanically move the other inputs to match.

The behavior of the co-pilot was so incomprehensibly baffling in this crash. Pulling back on the stick is like an economist mixing up supply and demand curves. I dont think anyone has ever been able to figure out why the co-pilot never thought to stop pulling back.

The depressing part of that transcript is immediately after the co-pilot tells the pilot he's been pulling back on the stick the whole time - the pilot instantly knows it's all over and that he's got a couple seconds to live, and that it was such a stupid error that got them there. It's horrifying.

> I dont think anyone has ever been able to figure out why the co-pilot never thought to stop pulling back.

Under the normal fly-by-wire control law, constant back stick wouldn't stall the aircraft. The computers would stop pitching the aircraft up before it reached the stall angle of attack. But, the system had dropped to Alternate law which is a reversionary measure, in this case caused by the pitot system failing. Under Alternate law, high angle of attack protection is lost and you can pitch up into a stall.

That said, in the AF447 case, the suspicion is that the F/O never even appreciated they were in a stall, irrespective of Alternate law being active.

It's still a mystery why they couldn't comprehend that they were in a stall though. The captain did the math when the co-pilot told him he'd be pulling back the whole time: descending rapidly plus nose up = they were in a stall. He recognized this instantly.

The only explanation is that sometimes the brain just goes haywire during incredibly intense moments. Like pedal misapplications in cars, when people frantically try to brake but jam the accelerator instead, it never seeming to click in their heads that they're speeding up, ie pressing the wrong pedal.

Single page:

http://www.slate.com/articles/news_and_politics/transport/20...

Just like Patrick Smith thinks it's rash to indite the pilots of Flight 214 before we have all the data, I'd ask Pat to do the same for the victims, ahem, passengers.

I don't think the emergency exit was some walk down the isle where a few people stopped to reach under their seats and grab their carry-on to save the inconvenience. From the pictures I saw of the cabin, the seats were torn loose, piled up on top of each other, and there was no 'center isle' left. However, it does seem that the condition of the cabin after the plane came to rest was highly variable depending on where you were sitting.

For all we know, passengers that were seen carrying bags outside the aircraft could have had to throw those bags down the chute to clear the egress. Or maybe they were in a state of panic and shock and were running on autopilot (if only the plane had landed that way).

I do think Patrick is on to something, that the increased level of automation makes it more likely that when the automation isn't there to help that accidents will happen. What I still don't get is flying 30 knots too slow on approach should have been ringing alarms all over the place, including in the tower, and yet no one noticed?

I agree - I think a little more than the instrument landing system was offline during that period of time. Surely ATC had them on radar and could have easily seen them coming in way too slow to properly land.

That's not ATC's job. A pilot should be able to land the plane himself/herself in visual conditions.

I'd be rather surprised if the radar could be useful in this capacity. You'd have to build a large database of all aircraft approach speeds, taking into account all the various factors (different weights and configurations) that could change. You'd also have to take into account wind speed and direction, since radar will measure ground speed, but airspeed is what matters. And not just wind, but wind at their precise altitude, which your ground weather station is not measuring. (Wind speed and direction can vary enormously over altitude, especially close to the ground, due to friction.) Finally, you'd have to catch it and communicate it in time to do something useful, which seems tough in this case, since the airspeed didn't get low until fairly shortly before the crash.

I wonder if a simple and effective thing to do here would be to simply display a velocity vector, either in a HUD on the windscreen itself, or on some kind of video display in the cockpit. Just a symbol on the screen that shows exactly which way the airplane is headed at any given moment. When it's on a runway, that's where you'll touch down if nothing changes. When it's on a seawall before the runway, that's where you're going to crash unless you fix things. Maybe I'm underestimating the difficulty, but it doesn't seem like it would be too hard to construct one based on GPS and a forward-looking camera.

Edit: thinking more about that last suggestion, more recurrent training for visual approaches would probably be a better idea. A normal human has a built-in velocity vector already. You look out the window and find the point on the ground that isn't moving relative to your view. That's the point you're headed toward. This is Aviation 101 stuff, literally one of the first things a pilot ever learns about landing an airplane. High and low comes later: the first thing you pay attention to is whether that non-moving point is, in fact, on the runway where you intend to touch down. (Or, because of the flare, slightly before where you intend to touch down.)

ATC would not know the stall speeds for every model and configuration of aircraft that land every day. That's not their job.

People also don't respond rationally when in shock. When SQ006 crashed there were reports of passengers sitting in their seats frozen instead of evacuating.

On normal flights you spend hours or days before of being in the mindset to keep your valuables, medicine, money and identification on you (or within sight) at all times. Life will be miserable when you don't have those. A few minutes after landing they would have to collect them, and make sure to not to leave any behind before facing immigration etc.

It is easy to see how in a state of shock, the instinct to what you have been doing and were concerned about kicks in. The cabin crew will shout at passengers to leave their stuff in an effort to overcome this. It should be noted that on this flight there are reports that some of the escape slides opened up inside the plane trapping the flight crew, hence there being fewer available at the beginning of the evacuation.

Shock is a very strange condition. No matter how cool, calm, rational and effective we all think we will be, it isn't a predictor of what we will do.

This piece by James Fallows is better than the OP and touches on your comment wondering why low airspeed was not noticed:

http://www.theatlantic.com/national/archive/2013/07/professi...

One supposition is that the normal system for correcting low airspeed was temporarily disabled. Key lines from the article I linked:

"The 777 can catch you out with with what is known as the FLCH trap. When you are above the glide slope and need to get down in a hurry Flight Level Change (FLCH) is a useful mode to use."

[...]

"The 777 has autothrottle wake up, i.e. when the aircraft approaches a stall the power comes on automatically to almost full power. This gives pilots great confidence; however, autothrottle wake up is inhibited in FLCH. "

Of course, time will tell, and we're all out of our depth in trying to understand what happened.

FlightAware has radar data chronicling the crash: https://flightaware.com/live/flight/AAR214/history/20130706/...

The NY Times made a nice graph with it: http://www.nytimes.com/interactive/2013/07/07/us/asiana214-u...

That graph doesn't seem very meaningful. It compares the descent with the a perfectly smooth curve, but there's no indication of how much variation you can expect in normal landings. It would be much more interesting to see that descent and the ten before it overlaid, for example.

The FlightAware data isn't particularly accurate, and the update frequency nowhere near fast enough to get good information (vs. FDR).

For the less sensationalist view of the crash and analysis, see the AVHerald information [1]

[1] http://avherald.com/h?article=464ef64f&opt=0

I have to say I think the conclusion that this isn't related to the Korean Airlines mishap series is mistaken. KAL Cargo Flight 8509 [1] was the last KAL crash, when cultural reasons turned a instrument malfunction into a fatal crash, because a junior FO wouldn't speak up when a Senior Captain(Colonel RoKAF) over-banked the 747, and crashed.

One thing that surprises me is that the same personality trait that crashed 8509, would make them a lousy flight lead as well, not to mention a lousy Squadron CO or Wing CO. If the pilots in the squadron aren't comfortable raising safety issues with their CO, then sooner or later, a crash will happen.

It's not just a foreign issue, poor safety culture lead to the VFMAT-101 crash in San Diego, where 4 people on the ground died[2]. The CO of VR-1 (Fleet Logistics Squadron/VIP transport) got canned for poor command climate [3], which caused a few safety issues, then fudging her NATOPS flight proficiency paperwork. A more "balanced" CO would have realized that flying the checkride, staying fully current would have been easier, and certainly better for safety climate.

IMO(Former pilot): Based on reports, the Pilot Not Flying(PNF/Experienced FO) didn't inform the veteran Captain that he was significantly slow by 500 ft, which could be considered the "waveoff window"[3] for landing a 777 on a runway, even when on a visual approach. Others have pointed out the regulations requiring a go-around for a visual approach are vaguely worded as "when required" without specifying the conditions. Some pilots take that the regulations literally, and believe that a visual approach doesn't require a decision altitude, call-outs or a stabilized approach. It does require to pilots to be safe, and is often required by company Operations Manuals.

[1] http://en.wikipedia.org/wiki/Korean_Air_Cargo_Flight_8509

[2] https://en.wikipedia.org/wiki/2008_San_Diego_F/A-18_crash

[3] http://www.navytimes.com/article/20120916/NEWS/209160317/UAV...

To some extent these chain of command cultural issues (or whatever you want to call them) are unavoidable and show up in most airlines. The worst aviation disaster in history could largely be chalked up to this problem. Two 747's collided in Tenerife because the co-pilot of one was too timid to tell his pilot, who was practically famous within the company, that he was wrong and should wait to take off.

Not disagreeing that KLM Capt. van Zanten was wrong.

The point is that they are not supposed to adversely effect safety and cockpit environment. It takes real leadership to encourage co-pilots/subborinates to raise operational issues in an appropriate way.

Some reports indicate KLM Capt. van Zanten wasn't as overbearing as the official report said, but he still would have been somewhat intimidating due to his status, and the fact is that he had opportunity to take 15 seconds to confirm with the tower, and/or Pan Am 747, after the co-pilot pulled the throttles back. He didn't confirm, and a crash occurred.

One very senior test pilot I know, commented that if a crew chief DID NOT ground abort them (usually before taxiing) because of a safety issue, or at least discuss the problem, and let them fly anyway, then the crew chief would be history. It would be considered dereliction of duty, and heading towards Article 15 Non-Judicial Punishment.

Needless to say, it is not that uncommon for this senior pilot to ground abort, and the crew chiefs do their job.

I believe it's up to the airline to have a go-around policy for unstabilized approaches. Westjet, for example, requires the approach to be stabilized by 1000 feet. An unstabilized approach is a major causal factor in the majority of landing accidents (according to IATA research). I know which airline I would rather fly on.

Yes, Its up to the airlines. 500ft to be stablized is relatively safe. 1000ft is a little safer margin, but not having a stabilized visual approach procedure is definitively risky at best, and sometimes fatal.

http://en.wikipedia.org/wiki/Southwest_Airlines_Flight_1455 is an example of an experienced crew making unstable approach and landed high, then overran the runway and crashed. 500 ft for Southwest pilots to be "in the slot" or go around.

I agree 500ft is fine too. The important thing is to have some SOP for a stabilized approach, and for it to be monitored by the company. I suspect Asiana isn't doing this, but I guess we'll find out.

> It’s not yet clear which Asiana pilot was physically at the controls, the captain or first officer

That's not true http://www.npr.org/blogs/thetwo-way/2013/07/08/200086858/asi... We already know that the captain in training was flying. We also already know that a cockpit with 4 experienced pilots let the airspeed fall below vref for 30 seconds before doing anything.

Really good article!

Probably the best take-away is that the media is sensationalizing this, primarily because there hasn't been a crash by a major airline in over 10 years.

that and to leave your damn carry-ons behind in an emergency since you are slowing the process of getting everyone off the plane.

Anyone who has had to wait while someone goes "against the flow" to retrieve their bag from the overhead (because the bin above their seat was full when they sat down) knows this pain. In an emergency, that delay could kill someone.

When Boeing designed the 787, the composite construction allowed for larger windows. I wonder if it allows for larger emergency exit openings too.

None with fatalities, however in 09 US Airways Flight 1549 ditched in the Hudson to much fanfare. There have also been accidents with fatalities involving feeder airlines to the majors within the US. Outside the US is another story entirely, but we tend to ignore most of the accidents unless the numbers are large.

The interest I have in stories like this are, how many are preventable by having systems that can override a pilot and is there and preference to doing such?

Look at your bug database. Multiply that by the complexity of not only an airliner, but the huge and complex system in which it operates. Add weight for consequences in human fatalities.

Now, would you prefer that system overriding the pilot, or the other way around? Humans can think outside the box. Software is the box.

"Look at your bug database. Multiply that by the complexity of not only an airliner, but the huge and complex system in which it operates."

And take the n-root of the total based on the number of quality checks every minimal part of that software has been through before even being tried on a real plane, not considering the ones before it went into service.

Seriously, you really think the software for an airliner is written in Ruby+Javascript and the developers use a cloud-based bug tracking app? I work on systems used in public trasportation that are NOT safety-critical and the number of tests, checks and counterchecks they have to go through before being approved for service is soul crushing.

I'm pretty sure that the total count of in the bug database of any airline software currently in service basically amounts to zero.

"Seriously, you really think the software for an airliner is written in Ruby+Javascript and the developers use a cloud-based bug tracking app?"

No. The point was to reflect on the nature of software and complexity.

I doubt if bug counts are near zero.

EDIT: And undiscovered bugs are definitely not zero.

> the number of tests, checks and counterchecks they have to go through before being approved for service is soul crushing.

> I'm pretty sure that the total count of in the bug database of any airline software currently in service basically amounts to zero.

The military, NASA, and a bunch of other critical organizations really wish that what you were saying was true. http://en.wikipedia.org/wiki/List_of_software_bugs

That's pretty much the point. The safety required makes writing a complex, life critical system like this and getting it certified pretty much impossible with our current software methodologies.

There's been quite a few software bugs in certified aviation software. One in the control system for ABS brakes caused a runway overrun, another in the Airbus fly by wire system caused an in flight upset with multiple head and spinal injuries.

Besides which, I'd really hate to be the guy scoping out the software that decides when to ditch an aircraft in the Hudson river. Avionics software is considered reliable because it doesn't try to do too much.

Air France isn't major enough for you?

( https://en.wikipedia.org/wiki/Air_France_Flight_447 )

"in North America".

Best to read the whole sentence before dishing out inaccurate sarcasm.

AF447 crashed 2009 on its way from Rio to Paris. Co-Pilot error, wiped out all 228 onboard.

Armchair experts tallyho!

The article's statement is more qualified. "the first multiple-fatality crash involving a major airline in North America since November 2001"

well sure, but why even make this geo distinction? it was a Korean airliner on an trans-continental flight. hence the multi-national news this is making (vs. the even more recent crash in Alaska that killed more passengers but is indeed a local event).

and this was not exactly a minor crash, it broke the spines of several passengers for chrissakes. the evacuation was top notch, even if the rescuers rolled over one of the victims. and a lot of luck, if that plane had turned on its head during the stumbling phase, the evacuation wouldn't have gone so fast.

I feel like if you are following your "first" with that many qualifiers, it starts losing meaning.

The first multiple-fatality crash involving a major airline in North America with an airliner carrying over 100 passengers and at least five kosher meals, not arriving on a Tuesday.

And the best part is, in the huge amount of speculation around that crash, I never heard anything that came even close to what turned out to be the real cause of the accident. Which is part of the reason I think people should really stop speculating about this crash.

Because they spun out a lot of traffic to regional airlines. There's a good Frontline special about the regionals. Well worth watching.

http://www.pbs.org/wgbh/pages/frontline/flyingcheap/

When I compare the coverage of this with the endless coverage of the safe landing of the US Airways plane in the Hudson in 2009, this coverage does not seem sensationalized or over the top. In fact, this strikes me as some of the least sensationalized coverage of a fatal plane accident that I can recall.

I saw another opinion on Philip Greenspun's blog: My visual approach, and Asiana’s http://blogs.law.harvard.edu/philg/2013/07/07/my-visual-appr...

He is an amateur pilot.

Perhaps the most favorable indication of the culture of commercial air flight is that the attitude exhibited by this author has been exceedingly rare.

First of all I haven't read or seen any speculation that wasn't qualified as being mere speculation. Second, this is the first instance of reading someone minimizing the accident because of the excellent safety record.

Is it too obvious to say that the safety record comes from not minimizing accidents as flukes? The rigorous/tedious checklists, innovations like removing oxygen from near empty fuel tanks so they're less likely to explode, and flight attendant training for quick exits come from experience and the enormous amounts of attention directed at every past failure.

Such advances come from a combination of the expertise of people with domain knowledge and people outside of the domain having high standards. In 1985, described as an example of a terrible year for air travel, one still could have put the safety record in perspective with other hazards. Surely it was far more dangerous to travel by car. It is easy to imagine pilots and engineers of 1985 wishing that outsiders who knew less than them would just shut up and let them do their jobs without unnecessary distractions. It's even easy to imagine arguments about the meddling not only being costly, but even a dangerous distraction.

Some hysteria, and definitely a lot of lawsuits, aided in the prioritization of problems for people who had the expertise to come up with solutions to minimize fatalities in the future. Outside, unwelcome pressure also disrupts the status quo which, by definition, is part of every failure.

I imagine anyone who has domain specific expertise has at least at some time been incredibly annoyed to have to deal with and attempt to manage the reactions of nonexperts while they're in the midst of solving a problem immediately at hand. And yet, such a reaction is cowardly, and enormously unproductive in the long term.

I'm not sure if I am beating a dead horse over an obvious point, or not making the point very well at all, but I think it is really important to avoid the pitfalls of only talking to experts, or peers, or people within your established hierarchy of influence when attempting to address failures. This article, made me recall students of nuclear physics on Reddit talking about Fukishima, who couldn't decide whether it was irrelevant because it was an older reactor design, or whether it was irrelevant because it occurred after an earthquake and tsunami that were more severe than the models forecast as possible, or whether nuclear disasters in general are irrelevant because their understanding of worst case was the not-that-bad-all-things-considered case of what was only achieved through heroics of people at Chernobyl.

What made the point for me personally was reading about failures at NASA. I highly recommend the long "Columbia's Last Flight" by William Langewiesche in The Atlantic[1] as it had a great influence on me when it came out in 2003, and helped me learn to value input from people without specialty knowledge. Even when special expertise defines who is most likely to develop fixes, the big picture view sometimes only afforded to outsiders is important, or sometimes just enough disruption so different voices within the system are heard.

In the case of NASA, things were humming along for the entire edifice because it had worked before when they squelched concerns by small groups within the enormous operation. As a result, when it failed, it was the system itself that failed, and it would have been reckless to trust that system to determine what went wrong. Perhaps even more famously, with Challenger, Richard Feynman was the gadfly in that investigation, and later had plenty to say about the ability of entire cultures to encourage mistakes specifically as a result of their desire to build in-group consensus.

So, I think the air travel industry deserves an enormous amount of respect for its excellent safety record, and that the shortage of voices trying minimize this failure suggests that they will accurately determine what went wrong. The value of this column is as an example to the rest of us, not in the air flight industry, of what not to do, and what attitudes stand in the way of improvement and finding solutions.

[1] http://www.theatlantic.com/magazine/archive/2003/11/columbia...

Very thoughtful comment, and not too obvious; thanks.

I don't see this as trying to quiet dissent nearly so much as trying to inject some understanding into the fast-and-loose news coverage that seems to always follow big disasters like this. He's basically arguing that people wait for the investigation results, rather than getting caught up in wild speculation - that it would be very bad if people soaked up all the misinformation that the news was spitting.

Otherwise, I agree with your comment, and that it is great that the airline industry has focused so intently on removing every single failure mode they spot.

"Even when special expertise defines who is most likely to develop fixes, the big picture view sometimes only afforded to outsiders is important, or sometimes just enough disruption so different voices within the system are heard."

Yes! This is not unlike having a child solve the problem of the 18-wheeler stuck underneath a too-short overpass. Have outsiders review the data, the situation, suggest possible causes and solutions. This extends to so many venues, IMHO, beyond accidents like the Challenger/Feynman, Chernobyl, Fukishima and Asiana Flight 214. Some of the most successful and impressive businesspeople I have know are those that discuss their work with outsiders openly, and really listen to what the outsiders have to say.

I'm not familiar with the 18-wheeler anecdote you mentioned. Could you provide a link or details?

I was curios about it too. Straight googling was fruitless but guessing at the solution (Something within the common functionality of a truck and which would be "obvious" in hindsight and thus work for a riddle.) The google string that worked brought this up: http://rawksoup.com/neighborhood/pg/blog/read/1987/brain-tea...

There's a widely circulated story about everyone being stumped about how dislodge a truck wedged underneath an overpass, then a child suggesting that they let air out of the tires. It may be apocryphal, but it used to be referenced pretty often.

the article hardly says anything... (more than dont believe what media provides, which again is nothing new)

Yeah, its really an article full of air. Sounds right from a pilot's perspective.

So you would have preferred the pilot author make different uninformed, speculative claims about the cause of the crash? That's the whole point: no one reporting on or writing about this has all the facts yet

> no one reporting on or writing about this has all the facts yet

yep, that is implicit in the lack of an official and conclusive report. But I have spent 10 minutes reading the article waiting for something interesting, and he just says nothing apart from wait for the official info, dont speculate. which is the default mindset of many. but yes, there might be a whole lot of other people who needs to read this carefully ;)

The point of the article is to wait until the official report is released before jumping up and down on an innocent party. As has happened in many other situations, the media circus surrounding an unfortunate event can cause people to prejudge, and sometimes they don't stick around to learn the truth.

Case in point: I was reading something recently (can't remember what) where the author pointed out that many people still believe Richard Jewell [1] to be the Atlanta bomber.

[1]: http://en.wikipedia.org/wiki/Richard_Jewell

I'm finding it a little annoying how people seem to like to defend every aspect of the flight - the pilots were fine, the plane was fine, the runway is fine (two of those I believe, BTW), as if that makes it better. The plane crashed and two people died, something was _definitely_ wrong.

The distinction between regional carriers and the majors made in the article stinks. If one chooses lowest cost flights, it's impossible to not end up on a regional carrier on one leg of your flight.

Can someone explain to me why humans land these planes?

These airports are fixed. We can easily set up transmitters around them to allow the plane to triangulate its position to less than an inch. Even excluding the tragedy of the loss of human life, you would think the lost of a multi-billion dollar airplane even once every decade would push people to automate the most dangerous part of flying (landing) to machines.

Because machines aren't perfect either. Airports already have those transmitters and can autoland, but pilots always have to be ready to take over, and sometimes have to. There are just too many things that can go wrong technically to not have trained, proficient pilots in the cockpit.

I didn't argue not having pilots in the cockpit. I'm arguing against pilots being in manual control of the plane when landing when the plane is capable of landing itself. If it isn't, fine, have a pilot take over, but that shouldn't be the case 99% of the time.

If pilots only hand flew 1% of landings, they wouldn't be sufficiently practised to do a good job. The simulator just can't recreate real world conditions accurately enough to be relied on totally.

Hence, pilots must fly a certain number of manual and auto landings in a quarter to stay legal.

Exactly. A contributing cause of AF447 was a lack of high altitude flying experience and training on emergencies there, since the training regime is traditionally focused on the more accident-prone takeoff and landing phases.

The glide slope transmitter for the ILS approach on runway 28L at SFO was out of service since June 1. That said, I believe there is an alternate, GPS-based system that could have been available as an alternative for automated approach.

I'd suppose that given a pilot with low experience on that aircraft and relatively good conditions, this was considered a good opportunity to get a visual approach and landing under his belt. That presumes a lot, of course.

Autoland is often used (mandatory in very low visibility) but requires 2 humans to monitor in case of equipment failure. To develop autoland with sufficient redundancies to require no monitoring, and install accompanying ground equipment at every commercially used airfield is way more expensive than training humans to land aircraft. Airframe loss is an insurable risk at market price.

Presently, the numbers favour accepting human error.

I wonder how the failure rate for autoland compares to that of human-controlled landings (specifically on the Boeing 777). I'm fine with having the human as a backup, but if autoland on a particular plane has a lower failure rate, I prefer to trust my life to autoland.

A 777 is about $250m. But yes, the total cost to the airline of the accident will be in the billions.

Whatever happened on final approach into SFO, I highly suspect that it was in fact related to the culture of Korean air safety in 2013:

http://www.pprune.org/rumours-news/518568-asiana-flight-cras...

More dribble and speculation

Say what? This was a detailed and well-written post by a United check captain who actually worked in Korea. Also the info has been confirmed by another captain who worked there. I'd downvote you if this site would let me.

I can't believe I read this article on HN. People died due to some human error (if it's not the weather it must be some human error somewhere) and the public wants to know why. Media is serving that demand, it could be over-speculating but it doesn't mean that a polit should feel obliged to start defending the engineers on the expense of first categorizing it as a non-catastrophe and then blame the passengers.

To me, listing all the past tragedies to make the point that this one doesn't deserve the public attention is at the very least inconsiderate if not offensive.

As the article said, stop speculating. You don't know it was caused by human error. Wait for the data to be analyzed.

Yeah but he seems to take the liberty to speculate that passengers were being selfish in lugging their carry-ons down the aisle while evacuating and endangering others' lives. Unless anybody actually complained about this being an issue in the evacuation process, he's not sticking to the facts either. So please, let's stop with the "let's not speculate" garbage. Everybody will do it. He's just trying to cover for his industry and the people in his profession.

I agree - there was an egregious failure here and if it was due to somehting that could be fixed than it should be. Nobody is asking for a lynching or anything.

And I've no doubt that people whose job it is are already sorting through the available data to work out exactly what did go wrong and whether it's something that can be fixed. But breathless speculation based on the fragments of information available to the media isn't going to make that go any faster.

Yeah, let's give the benefit of doubt to the pilots and only scold the passengers. Typical apologist bullshit.

I wish the coward downvoters would actually point out why I'm wrong with my summarization of this article.

Well, I haven't earned my downvote wings yet, but I'd wager you did yourself in with "typical apologist bullshit". They're not silencing your opinion, they're disapproving of your tone.

It's not cowardly to downvote your comment without providing you an explanation.

Reply on news.ycombinator.com