TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Explanation of the Asiana SF crash by a former UAL captain

142 pointsby rikelmensalmost 12 years ago

16 comments

SilasXalmost 12 years ago
To save you from looking up the more obscure acronyms like I did (summaries where given are my own words):<p>RTO: Rejected take-off, when you have to abort a take-off (though it seems in the context of the post, the exercise described involved an edge case when you should go ahead with the take-off)<p>AGL: Above ground level<p>CAVOK: Ceiling and Visibility are OK: no clouds &lt; 5000 ft AGL, with &gt;10 km visibility<p>VOR approach: VHF omnidirectional radio range: navigating based on a radio signal, considered &quot;non-precision&quot;<p>IAF: Initial approach fix <a href="http://en.wikipedia.org/wiki/Initial_approach_fix" rel="nofollow">http:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Initial_approach_fix</a><p>LNAV&#x2F;VNAV: lateral&#x2F;vertical navigation <a href="http://en.wikipedia.org/wiki/LNAV" rel="nofollow">http:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;LNAV</a><p>FAF: Final approach fix<p>UNSAT: unsatisfactory rating (paperwork)<p>FCOM: Flight crew operating manual<p>CRM&#x2F;CLR: Cockpit resource management, a set of NASA-originating practices to limit and contain human error <a href="http://en.wikipedia.org/wiki/Crew_Resource_Management" rel="nofollow">http:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Crew_Resource_Management</a>
评论 #6024935 未加载
评论 #6025425 未加载
评论 #6024878 未加载
评论 #6024978 未加载
marzealmost 12 years ago
I find air disasters a strangely compelling topic, so I&#x27;ve read all 82 pages of comments on that site. Lots of very experienced pilots post wide range of views.<p>Here is a list of the likely factors that contributed to the crash, based on my reading of the comments at pprune.com:<p>poor piloting skills, specifically not monitoring speed during landing, a very basic skill<p>recent switchover of flight pilot from Airbus to Boeing (Boeing planes have throttles that physically move when the thottle setting is changed by automatic systems, Airbus doesn&#x27;t)<p>misunderstanding of details of automatic thrust control<p>inexperience of pilot conducting the training (new to being trainer, 1st flight apparently)<p>glide slope system on the runway was inoperable, otherwise would have possibly been used and the crash wouldn&#x27;t have occurred<p>possible crew social status issues, a known flight-deck issue in some cultures, where those of lower status are hesitant to correct those of high status, but it doesn&#x27;t seem likely a big factor here<p>FLCH &#x27;trap&#x27;, where if you use a certain type of autopilot setting, you might expect the thrust to keep the speed at or above a certain level with automatic throttle adjustments, but it doesn&#x27;t happen<p>To summarize, it appears the two pilots at the control thought the jet was in a mode where it automatic thrust control would maintain an approach speed, but when they increased angle of attack just before landing (basically gliding decent up until then), the speed fell off and the throttles did not move automatically, and they failed to note the falling speed until it was too late to avoid crashing. (once you command full throttle from idle, it takes five seconds for the engines to spin back up.)<p>If they had actually had the plane in the mode they thought it was in, they would have had a perfect landing of the most routine kind.
评论 #6024902 未加载
评论 #6025225 未加载
评论 #6024933 未加载
HarryHirschalmost 12 years ago
You encounter <i>a lot</i> of this behaviour in students when they show up in a research lab for the first time, having taken only lab classes and lectures before, and it isn&#x27;t at all confined to East Asians. It&#x27;s always a shock to see that your experiments don&#x27;t turn out as intended and to recognize that all the class exercises have some bearing in real life, so to speak.<p>Some people never make the transition to independent work, and those are handed their ass fairly quickly, at least in a working lab. If what the fellow writes in this article is true, KAL and friends have some work to do.
评论 #6024712 未加载
ghshephardalmost 12 years ago
An interesting thought experiment is speculate as to when the state of the art in automated flight control and supervision (AFC&amp;S) advances to the point at which human intervention in the flight process causes more harm than good.<p>I&#x27;m guessing it will be about the same time that human surgeons are significantly less effective than fully automated systems.<p>30 years? 50 years?<p>It&#x27;s probably already reached the case that for the non-corner case scenario (cross winds, flight emergency, exceptions) - that a fully automated system probably meets the performance of a human pilot, and the human pilot&#x27;s major contribution is the ability to exercise executive judgement in the event of exception cases.
评论 #6025180 未加载
评论 #6024827 未加载
评论 #6024854 未加载
nikcubalmost 12 years ago
<i>Original forum post was deleted. A mirror:</i><p>Low-down on Korean pilots, (From a friend).<p>After I retired from UAL as a Standards Captain on the -400, I got a job as a simulator instructor working for Alteon (a Boeing subsidiary) at Asiana. When I first got there, I was shocked and surprised by the lack of basic piloting skills shown by most of the pilots. It is not a normal situation with normal progression from new hire, right seat, left seat taking a decade or two. One big difference is that ex-Military pilots are given super-seniority and progress to the left seat much faster. Compared to the US, they also upgrade fairly rapidly because of the phenomenal growth by all Asian air carriers. By the way, after about six months at Asiana, I was moved over to KAL and found them to be identical. The only difference was the color of the uniforms and airplanes. I worked in Korea for 5 long years and although I found most of the people to be very pleasant, it is a minefield of a work environment ... for them and for us expats.<p>One of the first things I learned was that the pilots kept a web-site and reported on every training session. I dont think this was officially sanctioned by the company, but after one or two simulator periods, a database was building on me (and everyone else) that told them exactly how I ran the sessions, what to expect on checks, and what to look out for. For example; I used to open an aft cargo door at 100 knots to get them to initiate an RTO and I would brief them on it during the briefing. This was on the B-737 NG and many of the captains were coming off the 777 or B744 and they were used to the Master Caution System being inhibited at 80 kts. Well, for the first few days after I started that, EVERYONE rejected the takeoff. Then, all of a sudden they all got it; and continued the takeoff (in accordance with their manuals). The word had gotten out. I figured it was an overall PLUS for the training program.<p>We expat instructors were forced upon them after the amount of fatal accidents (most of the them totally avoidable) over a decade began to be noticed by the outside world. They were basically given an ultimatum by the FAA, Transport Canada, and the EU to totally rebuild and rethink their training program or face being banned from the skies all over the world. They hired Boeing and Airbus to staff the training centers. KAL has one center and Asiana has another. When I was there (2003-2008) we had about 60 expats conducting training KAL and about 40 at Asiana. Most instructors were from the USA, Canada, Australia, or New Zealand with a few stuffed in from Europe and Asia. Boeing also operated training centers in Singapore and China so they did hire some instructors from there.<p>This solution has only been partially successful but still faces ingrained resistance from the Koreans. I lost track of the number of highly qualified instructors I worked with who were fired because they tried to enforce normal standards of performance. By normal standards, I would include being able to master basic tasks like successfully shoot a visual approach with 10 kt crosswind and the weather CAVOK. I am not kidding when I tell you that requiring them to shoot a visual approach struck fear in their hearts ... with good reason. Like this Asiana crew, it didnt compute that you needed to be a 1000 AGL at 3 miles and your sink rate should be 600-800 Ft&#x2F;Min. But, after 5 years, they finally nailed me. I still had to sign my name to their training and sometimes if I just couldnt pass someone on a check, I had no choice but to fail them. I usually busted about 3-5 crews a year and the resistance against me built. I finally failed an extremely incompetent crew and it turned out he was the a high-ranking captain who was the Chief Line Check pilot on the fleet I was teaching on. I found out on my next monthly trip home that KAL was not going to renew my Visa. The crew I failed was given another check and continued a fly while talking about how unfair Captain so-and-so was.<p>Any of you Boeing glass-cockpit guys will know what I mean when I describe these events. I gave them a VOR approach with an 15 mile arc from the IAF. By the way, KAL dictated the profiles for all sessions and we just administered them. He requested two turns in holding at the IAF to get set up for the approach. When he finally got his nerve up, he requested Radar Vectors to final. He could have just said he was ready for the approach and I would have cleared him to the IAF and then Cleared for the approach and he could have selected Exit Hold and been on his way. He was already in LNAV&#x2F;VNAV PATH. So, I gave him vectors to final with a 30 degree intercept. Of course, he failed to Extend the FAF and he couldnt understand why it would not intercept the LNAV magenta line when he punched LNAV and VNAV. He made three approaches and missed approaches before he figured out that his active waypoint was Hold at XYZ. Every time he punched LNAV, it would try to go back to the IAF ... just like it was supposed to do. Since it was a check, I was not allowed (by their own rules) to offer him any help. That was just one of about half dozen major errors I documented in his UNSAT paperwork. He also failed to put in ANY aileron on takeoff with a 30-knot direct crosswind (again, the weather was dictated by KAL).<p>This Asiana SFO accident makes me sick and while I am surprised there are not more, I expect that there will be many more of the same type accidents in the future unless some drastic steps are taken. They are already required to hire a certain percentage of expats to try to ingrain more flying expertise in them, but more likely, they will eventually be fired too. One of the best trainees I ever had was a Korean&#x2F;American (he grew up and went to school in the USA) who flew C-141s in the USAF. When he got out, he moved back to Korea and got hired by KAL. I met him when I gave him some training and a check on the B-737 and of course, he breezed through the training. I give him annual PCs for a few years and he was always a good pilot. Then, he got involved with trying to start a pilots union and when they tired to enforce some sort of duty rigs on international flights, he was fired after being arrested and JAILED!<p>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&#x2F;CLR, it still exists either on the surface or very subtly. You just cant change 3000 years of culture.<p>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!<p>Finally, I&#x27;ll get off my box and talk about the total flight hours they claim. I do accept that there are a few talented and free-thinking pilots that I met and trained in Korea. Some are still in contact and I consider them friends. They were a joy! But, they were few and far between and certainly not the norm.<p>Actually, this is a worldwide problem involving automation and the auto-flight concept. Take one of these new first officers that got his ratings in the US or Australia and came to KAL or Asiana with 225 flight hours. After takeoff, in accordance with their SOP, he calls for the autopilot to be engaged after takeoff. How much actual flight time is that? Hardly one minute. Then he might fly for hours on the autopilot and finally disengage it (MAYBE?) below 800 ft after the gear was down, flaps extended and on airspeed (autothrottle). Then he might bring it in to land. Again, how much real flight time or real experience did he get. Minutes! Of course, on the 777 or 747, its the same only they get more inflated logbooks.<p>So, when I hear that a 10,000 hour Korean captain was vectored in for a 17-mile final and cleared for a visual approach in CAVOK weather, it raises the hair on the back of my neck.
评论 #6027729 未加载
评论 #6027331 未加载
评论 #6025713 未加载
评论 #6027687 未加载
lawnchair_larryalmost 12 years ago
For anyone as confused as me, the post that this linked to was deleted from the conversation.
评论 #6025285 未加载
评论 #6026193 未加载
kaptainalmost 12 years ago
One interesting aspect that the poster brought up was attributing some of the deficiencies to culture. Cultures which emphasize rote memorization in their education are usually maligned by western cultures because they result in individuals who lack the ability to improvise and adapt. Does anyone have any examples that go the other way (i.e. cultural aspects of western education that result in an individual being poorly suited for an occupation...even to the point of causing fatalities like being an airplane pilot)?
评论 #6025052 未加载
评论 #6025112 未加载
评论 #6024995 未加载
brettnakalmost 12 years ago
Since a lot of this has been coming up I thought many people might like to know the basics of how the auto-thrust system is used. Some of this information could be slightly inaccurate so if a real pilot wants to correct anything here, that would be great. The auto-pilot setting for autothrust (<i>A&#x2F;T</i>) has a lot of modes. You don’t specifically set the mode so much as a combination of settings can affect the mode that it’s in. The modes are the following (this is fairly generic and probably not exactly how the 777 system works):<p><i>A&#x2F;T</i> by itself will simply hold a specified airspeed by adjusting the Thrust Levers.<p><i>A&#x2F;T</i> + <i>FL&#x2F;CH</i> + <i>V&#x2F;S</i> + <i>Alititude Hold</i>: (Auto-Thrust + Flight Level Change + Vertical Speed + Altitude Hold) climb or descend to the set altitude by the set vertical speed at the specified airspeed. The thrust levers are maniplated by the system. Generally once the altitude is attained, V&#x2F;S and FL&#x2F;CH will turn off.<p><i>A&#x2F;T</i> + “V&#x2F;S”: The same as the above but with no set altitude.<p><i>A&#x2F;T</i> + <i>Thrust Hold</i>: I looked at the 777 main control panel and I couldn’t find the setting for Thrust Hold, however it seems to be referred to in various things on the internet and I just don’t know what to look for. This setting is normally used for takeoff (maybe landing?) where the pilot sets an airspeed and a thrust (via the thrust levers) and the airspeed is regulated by pitching the aircraft up or down.<p><i>A&#x2F;T</i> + <i>Thrust Hold</i> + <i>FL&#x2F;CH</i> + <i>Altitude Hold</i>: This setting is similar to the one above, but without setting a vertical speed. The indicated airspeed is controlled via pitching the aircraft up or down until the specified altitude is reached. I believe once the altitude is reached FL&#x2F;CH mode is switched off as well as Thrust Hold.<p>This is my best guess as to what the 5 modes the NTSB says the 777 has. If any pilots could correct anything I’m missing or that is simply wrong, that would be great.<p>Switching between modes is really not intuitive because you’re fiddling with settings and it’s hard to know what button press is actually going to tell the system to do something. If you’re interested in how confusing it can actually be, I’d recommend installing the x-plane trial and trying to mess with these settings.<p>If you want to give it a go, the beginner’s tutorial for x-plane &amp; the 777 has you do a visual landing on 28L @ KSFO<p><a href="http://wiki.x-plane.com/777_beginners_tutorial#Landing_the_777_at_KSFO" rel="nofollow">http:&#x2F;&#x2F;wiki.x-plane.com&#x2F;777_beginners_tutorial#Landing_the_7...</a>
arjnalmost 12 years ago
I&#x27;m curious about what the other pilot was doing. According to news articles, the second pilot was more experienced and qualified for that aircraft. Shouldn&#x27;t he&#x2F;she have been monitoring the approach ?
评论 #6024850 未加载
Alohaalmost 12 years ago
I&#x27;m going to ask a dumb question.<p>I have a commercial drivers license, I&#x27;ve driven semi tractors, large straight trucks, and a wide variety of passenger vehicles.<p>Is it safe to presume that flying one type of widebody airliner is the same as flying any other kind of widebody airliner?<p>I can drive a Pete, Freightliner or International and they all largely drive in the same broad generalities.<p>Just as I can drive any full size car, and they all drive the same.<p>Would two similar classes or aircraft be the same?
评论 #6024931 未加载
评论 #6024873 未加载
评论 #6025165 未加载
评论 #6024880 未加载
joshuakalmost 12 years ago
This just made me instantly search for the Malcolm Gladwell Korean air story, which of course I&#x27;m not the first to think of.<p><a href="http://www.theatlanticwire.com/national/2013/07/malcolm-gladwells-cockpit-culture-theory-everywhere-after-asiana-crash/67058/" rel="nofollow">http:&#x2F;&#x2F;www.theatlanticwire.com&#x2F;national&#x2F;2013&#x2F;07&#x2F;malcolm-glad...</a>
lsh123almost 12 years ago
<a href="https://www.youtube.com/watch?v=h3kREPMzMLk" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=h3kREPMzMLk</a>
ronilanalmost 12 years ago
&gt; &quot;The X are very very bright and smart so I was puzzled by their inability to fly an airplane well&quot;<p>Turns out to be true for most X as long as: 1. The speaker does not belong to group X 2. Speaker has one data point to support said inability.<p>Humans error. Humans that program machines error. Machines error.<p>Buckle your seatbelt.
评论 #6024872 未加载
评论 #6029262 未加载
jontroalmost 12 years ago
Looks like the pilots were blinded: <a href="https://news.ycombinator.com/item?id=6025618" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=6025618</a>
tomgirl1almost 12 years ago
What NTSB has now said is that Auto-throttle has not 2 but 5 modes. Im not a pilot so all I knew of was 2. That may be a combination of different auto-pilot modes combined with auto-throttle modes.<p>But what seems clear is that the moment that the pilot turned off auto-pilot, the flight was doomed to crash as the auto-throttle was not in the right mode for landing. As in, a mode that would actually use thrust vs elevators to maintain airspeed.<p>NTSB also revealed that course corrections were being done not at the latest moment of 500 feet, but at 4000 feet 2.5 minutes before landing the plane was off course, in which case supervision, and correction would have been appropriate in a training scenario.<p>I hate to draw conclusions like this, but in today&#x27;s modern aircraft pilots only fly about 20-30 mins of a flight, takeoff and landing at most with autopilot managing most of the rest under normal conditions.<p>The inability to hand fly an airliner, is a disqualifier to be a professional pilot.<p>NTSB revealed that several modes on autopilot and autothrottle were being cycled in the last 2.5 minutes of the Asiana descent perhaps indicating a reluctance to hand-fly the plane and an over-reliance on automation.<p>Any commercial pilot that cant hand-fly a 777 on a perfectly clear day at SFO should never see a commercial cockpit again.
评论 #6024851 未加载
评论 #6024774 未加载
maxcanalmost 12 years ago
This post was on &#x2F;r&#x2F;flying and was thoroughly debunked to be a hoax.
评论 #6026967 未加载