Airplane cabin oxygen masks...more "feel good" than help?

[Whack-a-Mole]

@Xema

But why bother? Padeye has pointed out that this situation is extremely rare, and no more dangerous than the plane descending out of control.

Forgot to add…

I would think a plane flying on autopilot is more dangerous than one plummeting to the ground (at least for other planes flying…not counting whoever is on the ground under the plane). The chances of another plane being struck as a 737 hurtled towards the ground seem pretty slim. There is only a few moments where the two planes would share the same altitude. However, if the jet descended to (say) 12,000 feet and started flying merrily along and if, as you say, it has no radar to see ahead of itself to avoid obstacles, then it may well be at the altitude of another plane that thought it was flying in clear airspace. Big surprise for that guy when the 737 rolls over him from behind. (Still seems like it’d take major bad luck for that to occur though.)

Basically you have more chance for mishap when a plane is at an altitude it is not supposed to be at as other pilots fly about thinking they are in the clear…more time means mroe chances for disaster. Plummeting to the ground it is only at any given altitude for a moment. Flying about it stays at the “wrong” altitude for who knows how long.

[Dog80]

@Whack-a-Mole

Commercial airliners (read multi-engine jetliners) do not have full blown radar that shows what is ahead of the plane? Really? Smaller planes I can understand but I thought radar was standard equipment on the larger jets.

IIRC neither civilian airports have radars.

[Bookkeeper]

@Whack-a-Mole

Commercial airliners (read multi-engine jetliners) do not have full blown radar that shows what is ahead of the plane? Really? Smaller planes I can understand but I thought radar was standard equipment on the larger jets.

Airliners have radar but it is weather radar, designed to “see” rain and clouds. The radar used to find other aircraft uses different frequencies and techniques.

[Xema]

As Bookkeeper has noted, radar on civilian aircraft is nearly always of the weather-spotting variety. Some versions of this are suitable for higher-end single-engine aircraft, but none will show primary returns from other planes.

Aircraft avoid hitting each other in two basic ways. Using VFR (Visual Flight Rules), it’s “see and avoid” - you’re supposed to look out the window and not run into anyone. Using IFR (Instrument Flight Rules), you follow a clearance issued by ground-based air traffic controllers; such clearances are designed to ensure safe separation, so you don’t have to look out the window (except when flying in airspace where VFR traffic is also allowed) and thus can safely fly in clouds. [Preceding explanation somewhat oversimplified.]

@Dog80

IIRC neither civilian airports have radars.

Towered fields usually do. Many aircraft that use such airports (especially the busier ones) have transponders, which both enhance the image on the screen and cause an identifying code to be displayed near it.

[Xema]

@Whack-a-Mole

Basically you have more chance for mishap when a plane is at an altitude it is not supposed to be at as other pilots fly about thinking they are in the clear…

If they are being tracked by ATC (air traffic control), the other pilots can be warned.

But in either case, the chance of a problem is really remote, given how few such “unpiloted” descents can be expected.

[Broomstick]

@Whack-a-Mole

You are correct that I am not a pilot. Still, it seems to me much of this can be manageable. Couldn’t the autopilot be programmed to pay attention to its own radar? That is, if it “sees” something out in front (such as Broomstick’s single engine plane) it would know enough to modify its descent to avoid whatever it is? Isn’t that what a human pilot would essentially do?

Yes, but there are still some important differences between machines and humans. As a general rule, machines deal best with unvarying routine. Humans deal much better than machines with the unexpected. This is why we have mechanical autopilots - to deal with routine flying - and humans to deal with emergencies.

Human pilots are much better at spotting and avoiding other airborne obstacles with unpredictable behavior such as other aircraft and flocks of birds than machines are.

I mean, if cabin pressure is lost the pilot presumably takes the plane ASAP to lower altitude and reports what they are doing to ground control. I doubt they wait a few hours for everyone to be contacted. The autopilot would essentially do the same thing and while it is not as savvy a flier as a human today’s computers are good enough to tell it to avoid proximity with other “things” on the radar (the military uses terrain following autopilots).

Which is just spiffy for the military - however, even unmanned drones still have human minders sitting behind a video screen watching what’s going on.

The other thing is that human pilots are well aware that there may be other fliers out there who haven’t gotten a warning of their unusual location - a machine wouldn’t know or care, and could quite easily plow into someone or something else.

While it may be a much more significant addition to the avaiation system in airports and planes enabling ground control to send commands to an autopilot would seem very useful.

It would also be a very good way to hijack an airplane and turn it into a missile, wouldn’t it?

That way ground control could steer the plane appropriately and avoid many issues you are talking about.

There’s a time delay issue to worry about, interference with signal transmission, and the fact that the remote operator is divorced from the crisis in a way the people in the cockpit are not. It’s not easy to fly an airplane by remote control - how much are you willing to pay to maintain an appropriate level of trained personnel able to do this in the real world? Nor can you just plunk an airplane pilot in front of a remote control set up - it’s a different skill set, as anyone who have flown both remote control and in-person aircraft can tell you. I think it would be cost prohibitive.

Of course, such things as the plane in Greece are rare occurrences so that alone may not be sufficient reason to bear the cost and effort of such a system (unless they wanted to be able to seize control of a plane (say one that is hijacked) so terrorists could not do whatever it is they may want to do with the plane.

In which case, how do you secure the remote controls so terrorists do not take them over?

Think about it - in order to turn a civilian passenger jet into a poor man’s cruise missile at present the Bad Guys must infiltrate a flight and sacrifice a skilled operative for every airplane… but with remote overrides they need only seize one, crash it, seize another, crash it, seize another… For as long as the Bad Guys can hold out at remote control central they’ll be able to keep their “skilled operative” alive and destroy one plane after another.

I imagine the thought of the plane’s controls being seized by ground control is one pilots would definitely not like though.

No, I would not like that - and maybe you shouldn’t like it, either.

[Broomstick]

@Xema

Civilian aircraft almost never have radars that depict other aircraft. They do have systems such (such as TCAS) that can give proximity warnings about some other aircraft (basically, those equipped with transponders).

Basically, if I recall correctly, TCAS picks up the transponder signals of other aircraft. Which presumes said aircraft have a functional transponder, that it is turned on, and that it is transmitting accurate information.

I once had a transponder reporting my Cessna 150 as being at 30,000 feet - which is a physical impossibility. And would severely confuse any TCAS picking it up.

Lack of transponder and transponder malfuction happen far more frequently than depressurization accidents resulting in all cabin crew incapacitated. In most depressurization incidents they crew are able to respond appropriately and do so, eliminating the needs for elaborate electronic backups. In fact, it is so rare that all crew are incapacitated that I am sure the investigators are looking very closely at the oxygen system for the pilots.

[Broomstick]

@Xema

Towered fields usually do.

Actually… many don’t.

Of those that don’t, some have DBRITE displays, which is essentially radar piped in from a remote location that does have radar, but it’s not as reliable and they many restrictions on what they can use it for.

The rest are just a bunch of guys in a tower with lots of windows and some binoculars. On cloudy days, maybe they open the windows and listen hard.

In the Chicago area O’Hare and Midway both have radar, but Waukegan doesn’t, and doesn’t even have a DBRITE. If I recall, Palwaukee has a DBRITE but I was last in their tower something like 9 years ago so I might be remembering wrong, or things might have changed. Gary Regional - which does accomodate scheduled passenger service - has just the DBRITE, not a real radar.

Even in the US, there are scheduled passenger flights to airports that have no tower at all, and no on-site “ground control”.

Many aircraft that use such airports (especially the busier ones) have transponders, which both enhance the image on the screen and cause an identifying code to be displayed near it.

Even a lot of aircraft that never see a towered airport have transponders and they even get used. Not only do they open up such nice features at “flight following” to the weekend hobbiest, they make it far less likely you’ll be run over by a 747.

[Whack-a-Mole]

@Broomstick

There’s a time delay issue to worry about, interference with signal transmission, and the fact that the remote operator is divorced from the crisis in a way the people in the cockpit are not. It’s not easy to fly an airplane by remote control - how much are you willing to pay to maintain an appropriate level of trained personnel able to do this in the real world? Nor can you just plunk an airplane pilot in front of a remote control set up - it’s a different skill set, as anyone who have flown both remote control and in-person aircraft can tell you. I think it would be cost prohibitive.

My idea was not to have someone remotely flying the plane. Just a way for ATC to issue a broad command to the autopilot of a plane with (presumably) unconscious pilots onboard to go to XX,XXX altitude at XXX speed going in XX direction. This would be to allow the plane to be set on a heading that avoids obstacles (mountains for instance) and other planes as one would think ATC that issued such commands would be aware of what was a safe route for the plane. At the least they could direct the unpiloted plane to the middle of nowhere so when it ran out of fuel it wouldn’t hurt anyone on the ground when it crashed.

If you do not like the idea of ATC being able to seize controls of an airplane (I think many of your issues could be addressed and is not as grim as you make it out to be but grant it would take some thinking about) the pilots, once awake and aware, could disable the autpilot and resume control on their own.

[gotpasswords]

Remote programming of autopilot scares the bejeezus out of me.

Second in my mind is the idea that terrorists could issue a global “descend to elevation 000” command and crash planes into the ground, or even issue a command to “go to coordinates x,y” where x,y is the location of a bridge, or building.

First in my mind is the idea that the thing could (and eventually will!) fail. What if a power spike causes the remote-control system to engage unexpectedly? What if its default values are all zeros?

In ground-based systems, we can throw enough hardware and programming at the idea so if one computer fails, the other takes over, or if a command seems unreasonable (eg: descend to 0 feet) the programming will ignore it or say “Sorry bub, I’m not gonna do that!” Problem is that this involves heavy and bulky stuff - airplanes are not good places to try adding heavy and bulky.

[Padeye]

Well elementary chaos theory does teach us that robot (airplanes) wil rise up against us. Nnnnng glaven!

I don’t think we’re quite ready for ground based remote control today but what makes you think that programming overrides for unreasonable commands will add weight or bulk? All the lines of code I have written in my life weight zero ounces. Backup compters? You know they make computers nowadays that can fit inside a suburban home, usually on a small child’s lap.

As a programmer I know better than to start writing code while I’m discussing the feasability of the project. All your what-ifs are very reasonable concerns but this is why we won’t put freshmen computer science students on the project but experienced engineers who know how to build a fault tolerant system. if you want something more specific we don’t use 0’ ASL as a default altitude. Good programs are written by defining the problem, then the rules then making a flowchart of decisions. Bad programs… are generally the ones I spend my working day fixing.

No one is saying that progress is always smooth. Fly by wire systems have potentially added a lot of safety to airplanes but we’ve already seen one Airbus crash because the system overrode what the pilot tried to do with the flight controls.

[Whack-a-Mole]

@gotpasswords

Remote programming of autopilot scares the bejeezus out of me.

Second in my mind is the idea that terrorists could issue a global “descend to elevation 000” command and crash planes into the ground, or even issue a command to “go to coordinates x,y” where x,y is the location of a bridge, or building.

First in my mind is the idea that the thing could (and eventually will!) fail. What if a power spike causes the remote-control system to engage unexpectedly? What if its default values are all zeros?

Look…this all might make for a good thriller but is very easily handled even absent fancy programming. Assume we are not talking about a system that can override pilots from the ground. Merely a way to send input values from the ground to a plane under autopilot. You could make it such that the plane would only accept groundbased inputs if certain criteria are met (cabin depressurized, no affirmative actions from the flight crew, ignore several specific queries from ATC, whatever…). If the criteria are not met the plane ignores any control inputs from the ground. Once the autopilot receives the groundbased input (assuming prior criteria were met) it could set off an alarm in the cockpit and not do anything for a minute allowing flight crew who are awake to overrride the system. Alarm keeps going till someone takes control back from the autopilot (e.g. just turn it off so the ground cannot possibly control the plane).

Remember we are assuming an unconscious (or dead) flight crew. At this point why would an autopilot taking some commands from the ground be any scarier than a pilotless plane? Frankly I’d be happier if they could point it away from a city.

[iamthewalrus_3]

@Broomstick

Yes, but there are still some important differences between machines and humans. As a general rule, machines deal best with unvarying routine. Humans deal much better than machines with the unexpected. This is why we have mechanical autopilots - to deal with routine flying - and humans to deal with emergencies.

Human pilots are much better at spotting and avoiding other airborne obstacles with unpredictable behavior such as other aircraft and flocks of birds than machines are.

Right, but when the humans are incapacitated, it’s better to have something to fall back on.

In which case, how do you secure the remote controls so terrorists do not take them over?

With a dead-man switch in the cockpit triggered when any of a set of Bad Things occurs. External controls can’t do anything unless the cockpit fails to respond within a reasonable amount of time.

@Whack-a-Mole

I would think a plane flying on autopilot is more dangerous than one plummeting to the ground (at least for other planes flying…not counting whoever is on the ground under the plane). The chances of another plane being struck as a 737 hurtled towards the ground seem pretty slim. There is only a few moments where the two planes would share the same altitude. However, if the jet descended to (say) 12,000 feet and started flying merrily along and if, as you say, it has no radar to see ahead of itself to avoid obstacles, then it may well be at the altitude of another plane that thought it was flying in clear airspace. Big surprise for that guy when the 737 rolls over him from behind. (Still seems like it’d take major bad luck for that to occur though.)

I’m pretty sure that unless you’re near a major airport hub, the chances of that happening are awfully slim. Much more slim than the chance that a failsafe system like this could save a planeload of people (or keep it from crashing into a known population center). Even with the slightly increased risk, there are plenty of solutions. The simplest one that comes to mind is to establish an “emergency” altitude. If all the planes on auto-emergency go to a particular altitude, all you have to do is not cruise at that altitude.

@gotpasswords

First in my mind is the idea that the thing could (and eventually will!) fail. What if a power spike causes the remote-control system to engage unexpectedly? What if its default values are all zeros?

Airplanes already have critical software running. How would this particular case make such software any more likely to fail?

[slaphead]

But I think it may be an overreaction to a vanishingly rare event. As has been pointed out, this type of thing has happened once or twice in the last decade out of probably millions of flights. While future ocurrences could be potentially be prevented by making some simple but non-trivial amendments to autopilots and flight routines, why bother?
This accident presumably happened because multiple other safety systems and procedures failed simultaneously. Given that an individual failure of any of those systems/procedures is serious, why not put the effort into fixing them instead?

Working out a way of having the spare wheel automatically deploy and attach itself if a wheel comes off a moving car seems a lot more complicated than just making sure the wheelnuts are fastened in the first place. KISS and all that. :dubious:

[Padeye]

KISS is a nice sloagan but airplanes don’t work that way. They need backup systems for failures. Maybe a spare wheel doesn’t come down in place of one that falls off but even the simple nut holding it in place has safety wire as a backup.

I absolutely agree with you though that it needs to be analyzed for risk/benefit before it’s implements. I said it was a good question and the answer may very well be that it’s a bad idea.

[Raguleader]

also, racer72 mentioned, I think, that the plane had a rather slip-shot maintinance record after the plane came to it’s current owner. Assuming in the future that a plane rigged with this fancy new, altitude-correcting, terrain avoiding, remote-controlled, tire-changing autopilot doohickey is in the hands of an airline with poor maintinance standards, THAT system might as well fail with the oxygen masks, cabin pressure regulamatrons, and the little AC vents above the seats that never seem to work on the airliners I fly on.

[Xema]

@Raguleader

THAT system might as well fail with the oxygen masks, cabin pressure regulamatrons…

It might, but since it’s independent of the other systems, there’s no reason to expect it will fail when they do. And if it basically consists of some additional software routines for the autopilot, it shouldn’t need much maintenance - most modern computers specify a long interval between oil changes.

[pilot141]

Well, I get the advantage of diving into this this discussion after a few days (damn scheduling!) and maybe some more information.

First off, I’ll say that there has never (repeat NEVER) been a passenger airliner lost because the pilots lost consciousness during a depressurization. The Payne Stewart Learjet accident was most likely caused by that, but it has never happened on a scheduled airliner.

That fact alone puts me in the mindset that installing any sort of remote-control system in the case of unconscious pilots is not worth the effort.

As to this accident specifically - something strange is going on here. The aircraft reported “air-conditioning” (read: pressurization) problems on climbout and requested to stay at 16,000 ft. Later they climbed up to 32,000 ft and radio contact was lost. I assume that they fixed whatever problem they had (or thought they fixed it) and proceeded to climb.

The other information passed on by the F-16s also has me intrigued. The captain was out of his seat, the FO was slumped over the controls, but someone was moving in the cockpit (speculation is that it was a Flight Attendant). There is only one scenario that I can come up with that would end up this way, and it involves a problem with the crew oxygen system.

In a passenger airplane (and even one of the spiffy cargo ones I fly now) if one crewmember leaves his/her seat above FL 250 (25,000 ft) the other crewmember is required to don the oxygen mask. The rationale is that if you’re above FL 250 with both people in the seat and lose pressurization at least one of you will be able to get the mask on in time. But if someone gets up (to use the lav, most likely) the other person should already be on oxygen so there is no chance of something going wrong.

In the Greek accident the Captain was out of the cockpit while the airplane was at FL 320. This would mean the the FO was wearing the oxygen mask. If the airplane then depressurized, the FO should have been OK as he was already on oxygen. The fact that he was slumped over the controls leads me to believe that there was some problem with the crew oxygen system. The F/ A might have grabbed one of the walkaround bottles and come up front only to find the FO incapacitated.

There is speculation that the F/A tried to land the jet and subsequently crashed. I don’t know if that’s true or not; what I do know is that if the airplane had gotten down to 15,000 ft or so under control the FO would have regained consciousness (unless he had some toxin in him) and would have been able to recover the airplane.

Note: I say “Toxin” but I don’t mean to imply some nefarious plot. Someone could have seriously screwed up and serviced the crew oxygen system with nitrogen (I don’t know how - the fittings are different, but that never seems to stop anyone). Nitrogen is prevalent around an airplane hangar - the tires of airliners are filled with nitrogen rather than ambient air (it’s that whole inert gas thing - in case a tire blows right next to a hot brake, leaking fuel or an electrical spark you won’t get a flash burn). If the FO got a 100% dose of nitrogen when the plane lost pressurization he’d be down for the count.

As with any accident we’ll have to wait until the final report comes out to find out what really happened (and even then it will be a series of events rather than one “gotcha” moment).

[Padeye]

If there was no depressurization what accounts for the text message from a passenger that said something to the effect of everyone being “frozen?” There was also a news report that a mechanic said there was a previous depressurization on that plane from one of the doors.

[Whack-a-Mole]

@Padeye

If there was no depressurization what accounts for the text message from a passenger that said something to the effect of everyone being “frozen?”

I heard on NPR that the text message was a hoax. Only heard it once and have not followed the story since though.