Once this 737 MAX stuff dies down, we'll probably stop seeing every pilot fart or hiccup reaching the front page. Pilots make mistakes, they always have and always will, we just haven't seen them reported as much. Commercial aircraft are still as safe as they ever were.
It's not a very big deal. In pilot training the common saying is "there are those that have forgotten the gear, and those that will", nobody is immune, you will forget it once at some point. Luckily it beeps at you loudly when you do :-)
This comment makes it to my Top 3 "I know what I'm talking about" list on Hacker News. Number 1 and 2 being the guy who won the Putnam, and the guy who burned the wood from a boat.
Gear warnings are the “this website is insecure, are you sure” of aviation. They go off for all kinds of dumb reasons. You learn to tune it out whether you want to or not.
Commercial aircraft are much safer than they ever were. In the 60s-80s there were typically just over 2000 fatalities per year. In the last decade typically under 1000 fatalities per year, despite the increase in passenger miles since the '80s. https://en.wikipedia.org/wiki/Aviation_accidents_and_inciden...
This story could make a good anecdote to remember when considering human factors, user interface, and checklist design.
> Investigators said that because the pilots flew the second circuit at 1500ft, the Electronic Centralised Aircraft Monitor (ECAM) had not reset on the second approach and it did not display a landing memo at 950ft.
“The absence of the landing memo should have prompted the flight crew to perform the items of the landing checklist as a ‘read-and-do’ checklist,” it said.
Like the lock icon in a browser’s address bar, it is not effective to rely on the user to notice the absence of an indicator.
I'd expect that the aviation industry be very much aware that relying on people noticing that something is NOT there is error-prone. But they did have several backup options, so even if the crew inadvertently managed to work around the first one, the second one (700ft warning) worked as expected.
Procedures are designed to do things always at the same time, triggered by something else. The human brain is great at making those connections.
For example, on takeoff the pilot monitoring watches the instruments and reports "positive rate" (means "we're climbing from the runway in a stable manner"), as pilot flying you always command "gear up" at that point. It is quite common in simulator practice to get an engine failure right after takeoff, and the most made mistake is forgetting to command gear up, because you're handling the emergency right at the moment you would normally do it.
Luckily (as described in the article) you would normally run a checklist after any of these events, and usually that makes you catch the mistake.
This was a situation where the automation normally prompted the pilots to check the gear. The automation failed to do this in a timely manner so that they had to go around.
The distraction argument didn't really make any sense at all to me. They were not distracted, they were flying the aircraft.
The UX moral here is that if you have a system that always prompts the user then you have to live with the fact that failing to prompt will almost always result in failure to act.
While not abnormal, performing a go-around with a visual circuit landing shortly following is a high stress situation. The argument is likely that the pilots were distracted/overloaded by the tasks they had to perform in short order, while maintaining situational awareness.
The history of flight is so interesting. Did you know that early flights were mostly postal runs? The mortality rate was something absurd. You pretty much got into the profession because you loved being a bird more than being alive.
One of the earliest telemetry systems was a beacon where you can only tell how far away you are from it. Like, you know you're X miles away, and you can see X increase or decrease. But you had no indication what direction it was in, or any of the miracles we take for granted today. So they would set up a path of these beacons, and you would fly from one to the next, and that's how you'd know where you were going. I want to say "at night", but honestly I am probably misremembering some of the fascinating details. But the point was, you were often navigating using primitive instruments that gave you very little data about how not to die within the next N minutes.
From the title, I thought the plane had accidentally landed with the gear up. But no, a master warning kicked in and the worst that happened was... they pulled up and went around. Woo.
but it really is woo. It's so fucking cool that humanity as a whole went from no flight to safe flight in one human lifespan. One old guy's worth of life! Modern civilization is only a few thousand generations old, and we're going from ground to air to air-but-safe in the blink of a slice of a microsecond of human evolution, relatively speaking.
I wish I'll be around to see it happen for space travel. SpaceX is coming tantalizingly close, yet space-but-safe is a different matter entirely. I think it will take a few generations for us to work out those problems.
Note that "most aircraft" maybe applies if you count globally and include all the ones sitting still in a hangar.
Larger aircraft never fly on NDB. Everything with a flight management system (which is all airliners currently in service in developed countries) will use GPS + DME/DME + inertial systems as backup and auto-tunes VOR also as a backup. The computer will not use NDB to update the position, even if GPS fails.
We do fly procedures (holdings for example often in Europe) based on NDB, but we do it using the GPS position of the NDB, not the actual radio signal.
I helped ferry an open cockpit biplane from Colorado to the Chicago area with more or less basic VFR equipment - compass and maps. (We did have a handheld GPS... which did not seem to stay connected) Eight hours of hand flying was a wild change of pace from the normal equipment we were use to. Started 'chasing' the compass when my co-pilot wanted a course modification... and I had missed our heading while already in turn. Embarrassing. It had been a long time since I actually tried navigating with an actual compass, and had to pay attention to mentally adjusting for compass errors in acceleration and turns.
With a plane and engine like you were flying, if you were getting low on fuel and flying over a highway with little traffic, could you have successfully refueled at a roadside reststop?
In this video from Alaska they land at a roadside gas station and refuel in a fuel emergency. In an in-flight emergency you can deviate from any rule... still poor planning by the pilot.
Not OP, but my guess is that their fuel requirements are different than automotive gasoline. Doesn't mean it won't work, but might not work as expected...
I haven't seen an airplane with NDB or ADF ever, and I've been regularly flying for about 5 years. They are super rare to find on charts, at least in the US. VOR is still common but most people use GPS.
NDB in the US is almost unheard of anymore. I don’t know a single pilot that willingly flies an NDB approach. I am not sure there are any NDE approaches still published in the US. And VOR in the US is being slowly phased out as well. As a side note, HIWAS is being ended in January. I don’t know of any GA aircraft that are manufactured in the past 20 years that had NDB as standard equipment and DME in modern aircraft is pretty rare (true DME, not the GPS substitute.)
I wouldn’t say “most aircraft” rely on NDB — at least not most IFR-certified aircraft actually flying. It’s pointless to use it. ILS and RNAV are by far the most common. NDB approaches aren’t required anymore on an instrument checkride.
When I look at the early history of flight, I find it so fascinatingly absurd that inventors simply jumped from high cliffs on the *promise" that their invention would work. No one else had mastered human flights. Designs similar to theirs had failed in the past.
Some people get a bit nuts when bitten by the "personal flight" bug, even to this day; people with no great knowledge of aerodynamics or mechanical engineering are always hopping in some homebuilt ultralight and killing themselves. People lose their rationality, chasing a dream.
Franz Reichelt warrants a mention at this juncture. In 1912 he was so invested in the success of his parachute invention that he refused the pleas of his friends to test it with a dummy and simply jumped off the Eiffel tower, to his death. Rather shockingly, Wikipedia has video footage of this event:
And the wright brothers were not the first to fly. They were the first powered flight. People had been experimenting with kites and gliders for a thousand years. It wasnt a "jump off the cliff" moment more than an incremental breakthrough after many years of struggle.
There are many disputers. One thing they all have in common is a complete lack of evidence. Ader, for example, pops up 17 years later saying he flew. But he's got no airplane, no witnesses, no photo, no followup - just a drawing. Even worse, he claimed an altitude of 3 inches - about what you'd get from going over a bump or a gust of wind.
And does anyone really believe someone would toil away for months/years, build an airplane, make one flight with it, not tell anyone, dismantle and trash the airplane, and forget about it?
Nope. I'd be flying all the time showing off my creation.
1. years of notebooks where they logged their ideas, progress, data, designs, etc. The notebooks clearly betray their understanding of what they're doing
2. a clear photograph of their first flight (a contender for the greatest photograph in history)
3. witnesses
4. a stream of prototypes leading up to the first flight
5. a stream of improved machines after the first flight
6. many demonstration flights
7. the original machine was preserved and sits in the Smithsonian for anyone to examine
8. exacting replicas have been built, which fly and exhibit flight characteristics that match what the Wrights described
The other claimants have literally zero of any of these.
There were large concrete arrows painted yellow every 15-20 miles and towers with flashing lights for night time. This was the 1920s so there weren't so many lights in rural areas.
The arrow shown in the wikipedia article looks a little rough to access, but the one at 37.0648086,-113.5952549 is an easy walk from the back of a suburban neighborhood. And I mean, what else are you going to do to break up a drive through southern Utah?
I was flying a Cessna 172 on a beautiful day when I got cleared for the visual approach from 15-20 miles out. I got so focused flying an ILS approach for practice that I
- didn’t remind approach to hand me off to tower
- didn’t talk to the tower at all
- landed at a class c airport without authorization
- radioed approach from the taxiway that I was holding and then realized how many things I fucked up.
It’s so easy to forget stuff and get tunnel vision and “remind ATC” isn’t on the checklist.
I switched to tower and told them I was holding and they said, in an annoyed voice, to taxi to parking.
I was with my CFI so we called the tower from the classroom and filed a NASA report.
Lucky for me this was in Asheville (KAVL) and they train controllers there too and it's all, basically, in one room so everyone knew we were coming in and there was only one other A/C sitting at the departure end holding short waiting on us. Super lucky. It sucks when you are so used to being told everything to do and having to argue with ATC about bogus turn instructions at 300' AGL that then when they forget it's still your fault.
We went back and listened to the ATC recordings and the regional carrier that landed before we took off asked the tower if they were cleared to land so I think we weren't the only pilots missing some calls.
My short final check is mixture, prop, gear down and locked, cleared to land. I have kind of a rhyme about it that helps. I will sometimes ask even if they cleared to land a while ago in case they forgot and put someone in front of you which has happened.
“Let”... he was just as distracted as I was. Not a single plane in the sky and smooth afternoon air for a change. Tuning in the ILS for me and being my eyes out the window safety pilot. Complacency got us.
On the flip side the chief was a hardass and would make me call the tower and get clearance to land before beginning my descent while running the pattern. Other schools / instructors always made a call on short final to remind the tower we were still out there but the chief would fly a 10 mile downwind before he would descend and turn base without clearance. Now I think he really did know something
Commercial aircraft likely have quite the list of checks to do when on base / final approaches; but we learn BOUMFISH / PUFLR:
B - Brakes: Checked / Working
O - Oil: Temps/Pressures Green
U - Undercarriage: Down / Green
M - Mixture, Master, Mags: Rich, On, Both
F - Fuel Pump: On
I - Instruments: Baro, DG, Radio's set
S - Switches: All in
H - Harnesses: Secure
--
P - Pitch: Full fine
U - Undercarriage: Down / Green
F - Flaps: Set
L - Landing Clearance: Received
R - Runway: Clear
Most pilots in the US learn a GUMPS check:
G – Gas (Fuel on the proper tank, fuel pump on as required, positive fuel pressure)
U – Undercarriage (landing gear down)
M – Mixture (fuel mixture set)
P – Propeller (prop set)
S – Seat belts and Switches
Between flying the aircraft, communicating with ATC, and ensuring all the procedures are correctly followed (including monitoring each other) - there is a lot of stuff going on in the cockpit; and things are happening really quickly; you can sometimes get "behind" the aircraft - especially when unexpected things crop up, kudos to the pilots for recognising an unstable approach, going around, and then responding appropriately to the alarms in the cockpit and not being afraid to go around again.
I'd say it takes around 3-5 seconds for each checklist item; calling out, verifying it's in the required state, and continuing to the next one - you're looking at about 30 seconds on average; which is less time than what the pilots would have between going around and coming back for another attempt. Not to mention they're already likely following a missed approach procedure etc etc, so already pretty distracted, and then given "unexpected" (to them) directions by ATC (different circuit direction).
The list on an airliner is actually shorter than for a small 172. And we don't "read and do" the checklist, we do it first and then read the list to verify. So the checklist also goes much faster.
Divert to an airport with as long a runway as possible and fly a "short field" approach. That uses maximum flaps and aims at minimum touchdown speed. You can land and stop most small planes in under 200m that way (if the brakes work)
Most commercial airports have runways of well over 2000m, So you just land there and let it roll out.
Large planes typically have multiple hydraulic systems, and within those systems there are "fuses" and accumulators to protect the vital parts (flight controls, brakes) in case there is a leak in another part.
The incident also suggests to me that there isn't an easy way to check that the airplane is configured properly for landing. A pilot should be able to tell at a glance that it is right.
On a modern type like the 777 there is an easy way to do it. You can call up the checklist on the screen and it marks everything that's "done" as green. So essentially you press the button, and all should be green.
One thing they learned with airplanes is to set up the gauges so that when things are good, the needles all point in the same direction, regardless of what they were measuring. This made it easy for the pilot to scan the instruments. Later they added green, yellow, and red indications so an out-of-normal indication showed yellow, and really bad showed red.
Something car manufacturers still haven't cottoned on to.
>Something car manufacturers still haven't cottoned on to.
For racing applications it's pretty common to have a multicolor LEDs as backlight for your gauges so your whatever gauge turns red when the reading is bad.
For normal cars the goal is totally different. They want to give you the minimum information needed to operate the vehicle and only alert when something is critically wrong. This is why proper oil pressure gauges went the way of the dodo in most vehicles.
It is a major accomplishment of modern commercial aviation that now, a gear-down error + missed approach qualifies as a noteworthy safety catch. Aviation safety has come so far in 50 years.
I have read that the frequency of major safety incidents is so rare now that detection and diagnosis of errors has had to go farther and farther "upstream" towards monitoring and detecting the potential contributors to eventual safety problems. There just are almost no more accidents to study each year. A non-stabilized approaches is itself now cause for study.
Truly admirable.
Early, setting the thesis of the video, "...these [loss of control] accidents always follow a pattern. The pattern is: an event occurs, the event causes distraction, distraction causes loss of control."
While not exactly loss of control, it certainly sounds similar and seems to follow that central thesis.
> The captain elected to remain at Flaps 3, which investigators described as permissible and safe but not Jetstar’s standard configuration for a visual circuit.
This surprised me (I’m a naive non-pilot). I guess I assumed that things like this would be decided by manufacturers and regulatory agencies, with (as happened here) pilots making on the spot decisions. I had not realized different companies might make different safety protocols. Is that common?
Yes, so common that every airline has it. They're called SOPs, Standard Operating Procedures, which is how the company wants you to fly the aircraft. They're based on manufacturer recommendations / procedures, but do differ per airline.
SOPs are very detailed, down to the level of the things each pilot says and how the other responds. As an example: "80 knots" is said by the pilot flying when reaching that speed on takeoff, the other pilot verifies and says "checked", or "abort" if the speed on their side is not the same. You don't say "hey Bob, mine is different", you say exactly what the SOP prescribes.
It makes it possible to fly with every other pilot from the same company. Usually pilots don't know each other before the start of the trip. It's safe because everyone does everything in exactly the same way.
There is, but it's smarter than that, because you normally start to use the first states of flaps (and slats, at the front of the wing) before putting the gear down.
The terrain warning system is aware of where airports are and will also call "too low, gear" if you're approaching an airport without the gear down.
Yes it says in the article they got a master warning. That's a big red light in the center of their view accompanied by an alarm sound.
The terrain warning system comes on after that if you still continue and will keep making the "too low, gear" announcement continuously on a loop unless there is a higher priority problem (engine on fire)
No, that would kill pilots. If you put the gear down, the drag increases greatly so if you do that during approach it increases the rate of descend significantly.
If it does that while the crew is distracted (why else would they forget the gear) you're putting them in an unstable situation, close to the ground, descending rapidly.
The current warning is fine. After what this crew got, if you still continue, it will literally shout "too low, gear!" at you in a continuous loop.
No, I'm pretty sure lowering the landing gear can change the aerodynamic properties of an aircraft pretty significantly. You wouldn't ever want that to happen unexpectedly.
You'd be right. Next time you're flying, pay attention when the gear goes down. You'll hear an increase in wind noise, coupled with more vibration. The airplane will also kinda "sag" from the increased drag.
It would be interesting to deploy the nose gear only when the main gear is in contact with the ground. You'd get better aerodynamic properties that way, because you'd never fly with the nose gear down.
But in that case if the nose gear fails to deploy you're likely past the point where you can throttle up and try again. If you try to lower the gear while you're still in the air and the gear doesn't go down, you can circle back and try again.
It's not supposed to fail. Anything could fail, and yeah that sucks. If it fails on the first try, it will probably fail on the second try.
Landing with gear up happens sometimes, and pilots are supposed to do a decent job with it. Large passenger jets have done this and been put back into service after repairs.
Failure can be detected. Throttle up, and any other needed adjustments, can be done immediately by the computer. Getting up off the runway again can be automatic.
Not exactly true. Many airplanes have manual gear release systems if the mechanical systems fail [1]. These systems don't always work, true, and they may still have to land gear-up true, but there is a manual fail-over system and it does have to be used occasionally. If the rear wheels are already down, though, it's too late to try this.
Aviation is not the type of industry where "failures can be handled automatically by the computer" is an acceptable answer. There's almost always manual systems in place for when the computers fail, which they do.
Qantas https://forums.jetphotos.com/forum/general-discussion-forums...
El Al https://www.timesofisrael.com/el-al-pilots-nearly-try-landin...
IAF https://indianexpress.com/article/india/chandigarh/when-an-a...
Jetstar, 2012 (!) http://www.digitaljournal.com/article/323319
Once this 737 MAX stuff dies down, we'll probably stop seeing every pilot fart or hiccup reaching the front page. Pilots make mistakes, they always have and always will, we just haven't seen them reported as much. Commercial aircraft are still as safe as they ever were.