Submitted via IRC for Fnord666
SFO: The typo that almost crashed a plane
In 2017, a commercial airliner lined up for takeoff at San Francisco International Airport on runway 01 Left, the main departure route.
The pilot accidentally punched 10 Left — a much longer SFO runway — into the cockpit computer, causing the plane to incorrectly calculate the appropriate thrust and wing flap settings.
The pilot’s simple reversing of the number caused the plane to nearly run out of runway, lifting off with only 400 feet left of asphalt, according to a Federal Aviation Administration report obtained by The Chronicle through the Freedom of Information Act.
It wasn’t the only such close call at SFO. The March 2018 FAA safety report found 25 cases from 2014 to 2017 in which airplanes from several carriers took off with less than 1,000 feet of runway remaining. The FAA believes some of those cases probably were a result of “transposition error” and said no other major airport in the United States has had a similar problem.
Aviation experts say airliners need to lift off the ground with enough runway left to abort a takeoff — 400 feet isn’t nearly enough and 1,000 feet is too close.
“Wow, that is practically the end of the runway!” retired pilot Ross Aimer, an aviation consultant familiar with SFO, said of the 2017 incident. “They were lucky they didn’t take out some of the instrument landing equipment erected at the end of that runway.”
The runway 01 error revelations are the latest issue at the airport involving its runways, taxiways and tarmac. The airport closed its busiest runway, 28L, on Sept. 7 for 20 days of repairs, leading to more than 1,000 flight delays and hundreds of cancellations. The closure was not related to the runway number issue but resulted from deteriorating concrete.
Runway 28L was also closed overnight in July 2017 for construction, contributing to a near-catastrophic botched landing. An Air Canada Airbus A320 mistook a crowded taxiway for its runway and came within 14 feet of crashing into four fully loaded planes before pulling up and narrowly averting what could have been the worst aviation disaster in history.
The aborted landing prompted a National Transportation Safety Board investigation and a Government Accountability Office report published last month saying the FAA needs to do a better job collecting and analyzing data on ground incidents. Reported runway incursions across the country nearly doubled, from 954 in fiscal year 2011 to 1,804 in 2018, according to the report.
The SFO close call also led to a three-day FAA safety visit to SFO in late February 2018. At the time, SFO had experienced four wrong-surface events involving two or more carriers during the previous year, according to the FAA report.
The agency determined that the runway 10-01 confusion was “high risk” and issued a memo in September 2018 to pilot unions and other groups to alert flight crews and airlines of the issue.
“We have not received any reports about this kind of incident occurring at SFO since 2017,” FAA spokesman Ian Gregor said. Reporting such an incident is voluntary, so it’s unclear whether the confusion remains.
SFO spokesman Doug Yakel said he believes the issue has been fixed.
Read the rest of the article for even more incidents that may give second thoughts about flying into San Francisco.
(Score: 2) by hemocyanin on Friday September 20 2019, @05:32PM (6 children)
Why not runway 00, 10, 20, etc.?
(Score: 2) by JoeMerchant on Friday September 20 2019, @05:34PM (2 children)
If they rename runway 10L to 09L, does that mean they will have to tear it up and lay new surface on the correct heading?
🌻🌻 [google.com]
(Score: 0) by Anonymous Coward on Friday September 20 2019, @06:17PM (1 child)
It's 1 degree. It's not that important.
(Score: 1, Informative) by Anonymous Coward on Friday September 20 2019, @06:26PM
(Score: 5, Informative) by Anonymous Coward on Friday September 20 2019, @05:40PM (2 children)
The runways are numbered according to the direction they face in degrees. Runway 01 faces ten degrees east, runway 18 faces due south, etc. Due north is numbered 36, not 00. If there are multiple runways facing the same way, they'll be marked left and right (and occasionally center). This is done so the pilots know which way to go without having to look runways up on a map, which would certainly cause more errors.
Because virtually all runways can be used in both directions, an airport with runway 10, for example, will normally also have a runway 28.
(Score: 3, Interesting) by All Your Lawn Are Belong To Us on Friday September 20 2019, @10:45PM
More or less. They can also be shifted by one (10 degrees) if you have enough of them, especially more than 3 like O'Hare which has a 9L, 9R, 10L, 10C, 10R (and the reverse also), which are all on a 093 final heading except for 10R which is 094 but still a couple degrees short of a 10 designation.
The problem with SFO redesignating 01/19 to 02/20 is that then you have that pair which can be messed up. It's not that bad a problem since it's the same runway, but since the FMC's all calculate initial waypoints based on the departure runway one still has problems. (For that matter, I would have thought that FMC waypoint screwups for certain departures would be more of a problem than FADEC power calcs but maybe not). Maybe changing 10/28 to 09/27 (11/29 still gives a 01 and 11 confusion) but 10 is an approach of 104 and that's a bit of a stretch.
Maybe just increased pilot training and awareness will have to do. Or... they could spend a load of money buying land and extending runways. :P :)
This sig for rent.
(Score: 2) by pipedwho on Saturday September 21 2019, @01:27AM
Sounds interesting. With that in mind runway 10 could become runway 46 and the rule of thumb stands.
(Score: 0) by Anonymous Coward on Friday September 20 2019, @05:35PM (2 children)
Instead of 10-01, how about 10-02?
(Score: 1, Insightful) by Anonymous Coward on Friday September 20 2019, @07:12PM (1 child)
How about using GPS to confirm that the plane is on the runway that the pilot specified? The error could be announced as soon as the engines were at high/takeoff power and the plane rolled a little bit.
(Score: 2) by RS3 on Saturday September 21 2019, @01:58PM
Yes, GPS, a magnetic compass, a gyroscope, and some kind of ground-to-plane proximity communications telling the flight computer what runway the plane is on. Then offer that info to the pilot, flag any inconsistencies, and allow the pilot to make final decision.
(Score: 2) by Thexalon on Friday September 20 2019, @05:56PM (3 children)
The reason aviation is ridiculously safe worldwide is that there has been a longstanding global effort to make it that way. Every accident and near-accident gets investigated, equipment is constantly tested and inspected, and there are agencies from multiple governments constantly on the lookout for problems. And there's also constant redundancy and margins for error built into everything. It's not perfect, because no system is perfect, but it's much much safer than other ways of getting around because of that effort.
And that is why a problem like this typo gets caught and addressed before someone gets hurt. Compare that to how much damage drunk or sleepy drivers do every single day.
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 4, Funny) by nobu_the_bard on Friday September 20 2019, @06:01PM
https://xkcd.com/2030/ [xkcd.com]
(Score: 1, Interesting) by Anonymous Coward on Friday September 20 2019, @07:06PM (1 child)
All sounds good, until this article the other day in NY Times -- according to the author* the regulatory agencies in a number of countries, including Indonesia and Ethiopia, are more than a little corrupt and investigations may be halted or otherwise influenced by money. Available here, https://www.nytimes.com/2019/09/18/magazine/boeing-737-max-crashes.html [nytimes.com] or https://archive.is/bRvbM#selection-1017.15-1019.8 [archive.is]
* the author, William Langewiesche, is a former pilot himself, as well as son of the author of "Stick and Rudder"--a real classic pilot training book. At least in my mind, this gives this story a lot of credibility, compared to 737Max stories written by more ordinary journalists.
(Score: 2) by Thexalon on Friday September 20 2019, @09:06PM
I'm sure that there are aviation agencies that can be bribed. I wouldn't be surprised in the least if Boeing tried to bribe their way out of the problems in the 737 Max. And you know what? It didn't work for very long. The entire fleet of 737 Max planes has been grounded for some time, and at most that bribery cost 157 lives (the second 737 Max crash). Compare that to, say, the bad gas tanks on the Ford Pinto that cost 180 lives.
Part of the reason is that every plane crash is news, while nearly all car crashes aren't.
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 2, Insightful) by fustakrakich on Friday September 20 2019, @06:04PM (4 children)
It's risky behavior. It's like doing intersection takeoffs. All that runway behind you does no good.
The best practice is to use maximum thrust at take off, regardless of the available space. Get as high as you can as fast as you can, and your chances of hitting anything are greatly reduced.
La politica e i criminali sono la stessa cosa..
(Score: 1, Interesting) by Anonymous Coward on Friday September 20 2019, @06:20PM (1 child)
Higher thrust puts more load on the engine. There is a higher chance of something breaking in the engine itself. This is not purely to save money but also helps with noise regulations.
(Score: 3, Informative) by fustakrakich on Friday September 20 2019, @06:28PM
No, it just means you check and replace them a bit more often. Reduced take off thrust is all about the bean counters. Noise regs don't cover the take off itself.
La politica e i criminali sono la stessa cosa..
(Score: 4, Interesting) by legont on Friday September 20 2019, @06:25PM
While I don't think intersections are that big an issue, in general you are absolutely right. Taking off with full power is the best tactics for many reasons. My favorite one - I want my engine to blow up while I am still on the ground and running it at full power while it is still not very warm increases the probability.
Modern airline engines though have every gram of weight and efficiency squeezed out of them so they have much less serviceable time under full thrust. Hence airlines force pilots to use the least power setting possible. This is a typical trade-off of money for safety.
Note that no airline by itself can change the practice because it will lose competition. The only way is for the government to enforce rules on everybody.
"Wealth is the relentless enemy of understanding" - John Kenneth Galbraith.
(Score: 0) by Anonymous Coward on Friday September 20 2019, @06:39PM
The more conservative airlines set 90% as a maximum and treat going to 100% throttle as a loggable incident.
"Low-fare" airlines on the other hand, have already adopted your suggestion.
(Score: 4, Insightful) by mhajicek on Friday September 20 2019, @08:02PM (3 children)
I think there's a sort of "uncanny valley" of partial automation, where much of a task is automated but it still relies on human inputs with insufficient error correction which can throw the whole process off. In this case, the plane should either figure out which runway it's on and adjust accordingly, or warn the pilot if the setting or runway is wrong.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 1, Insightful) by Anonymous Coward on Friday September 20 2019, @08:15PM
See also: Tesla Autopilot.
(Score: 4, Insightful) by Thexalon on Friday September 20 2019, @10:22PM (1 child)
"Well, I don’t think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before, and it has always been due to human error."
- HAL 9000 computer
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 2) by RS3 on Saturday September 21 2019, @02:01PM
And human errors went into the design, construction, programming, and mission objective of HAL 9000.
(Score: 0) by Anonymous Coward on Friday September 20 2019, @11:44PM
-nt
(Score: 3, Informative) by choose another one on Saturday September 21 2019, @10:49AM
I'm glad these "experts" are merely talking bollocks on the internet and not actually sitting at the pointy end when I'm in the back.
Proof:
Anyone with even a cursory knowledge of flight procedures knows that V1 is the takeoff abort speed - and yes, you should hit that with enough runway left to abort
Above V1, there is no abort, you are committed to fly, or crash, that's it - runway remaining is now academic and is not a factor in the calculations
VR is rotate speed - when you lift the nose up
V2 is the actual "lift off the ground" speed (actually on a twin it is the one-engine-failed lift off speed)
V2 is (normally, not absolutely sure if always) greater than V1, and therefore achieved later, with less runway remaining
Therefore, lifting off with not enough runway left to stop is entirely normal