Fricking crashing airplanes

[rumborak]

Man, yet another one, and equally as weird. I saw a graph today of the altitude during the flight (similar to this one ), where it seemed the descent started right when the plane went over ground.

[MrBoom_shack-a-lack]

I read the news as I was about to board my plane from Düsseldorf - Copenhagen. Not the greatest thing to read before a flight. Such a tragic thing!

[El Barto]

This one will get sorted out pretty quickly. There are already people dumping ADS-B data and making charts and they have at least one black box (unknown which once at this point). This one won't be a big mystery.

It does seem that the AC was flying a phugoid after they left FL380, so the speculation is on pilot incapacitation; sudden decompression being the obvious guess there. Some airbus drivers over at PPRuNe are actually trying to piece together that ADS-B data to see if the flight director was calling the shots and what mode it would have been in.

[Chino]

This one will get sorted out pretty quickly. There are already people dumping ADS-B data and making charts and they have at least one black box (unknown which once at this point). This one won't be a big mystery.

They found the box with the voice recordings on it. They are still looking for the box with the flight data. I still don't understand why this data wasn't being written to a HD on land in real time. They could have had everything mapped before nightfall yesterday.

[El Barto]

This one will get sorted out pretty quickly. There are already people dumping ADS-B data and making charts and they have at least one black box (unknown which once at this point). This one won't be a big mystery.

They found the box with the voice recordings on it. They are still looking for the box with the flight data. I still don't understand why this data wasn't being written to a HD on land in real time. They could have had everything mapped before nightfall yesterday.

The CVR should be pretty helpful, though not as much as the FDR. And honestly, the current black box system really works pretty well. We had one instance where a plane gets sucked up by aliens so there's no flight data, but that was certainly the exception. There's really no reason to go reinventing the wheel here. Moreover, they already have a pretty large amount of data from the ADS-B telemetry. That's only half the story from the FDR, but it's still a huge improvement over 3 years ago.

And the best answer to your question is that any design change on an AC takes years of effort. Things as mundane as the inflight entertainment system take years and years to role out. As a rule, major changes like that only happens when a whole new family of AC are introduced so all of the regulatory stuff can be done at the onset. Once you consider that those regulatory things have to be done for a huge number of governing bodies then you're looking at a ton of effort for a system that A: you try your hardest to insure never gets used, B: only gets used in the tiniest fraction of flights, and C: works surprisingly well in the rare instances it is employed.

[rumborak]

I think much more surprising is that the plane can descend for 8 minutes, and ground control is totally clueless and powerless.

[El Barto]

I think much more surprising is that the plane can descend for 8 minutes, and ground control is totally clueless and powerless.

Yeah, I suspect that's a part of why the speculation right now is around pilot incap. I do know that centre ATC around Germany/Switzerland/France tends to be privatized rather than national, and there have been some lapses before. Wouldn't really be a factor here, but it might preclude them from getting the transcripts immediately.

I suspect you're already familiar with a particularly bad evening with regards to that system, but for anybody else that's interested here's a good read on how a whole bunch of factors (including some really bad luck) can result in an ATC meltdown: https://en.wikipedia.org/wiki/%C3%9Cberlingen_mid-air_collision

[rumborak]

Jeez yeah, that accident was horrible.

[cramx3]

This one will get sorted out pretty quickly. There are already people dumping ADS-B data and making charts and they have at least one black box (unknown which once at this point). This one won't be a big mystery.

They found the box with the voice recordings on it. They are still looking for the box with the flight data. I still don't understand why this data wasn't being written to a HD on land in real time. They could have had everything mapped before nightfall yesterday.

The CVR should be pretty helpful, though not as much as the FDR. And honestly, the current black box system really works pretty well. We had one instance where a plane gets sucked up by aliens so there's no flight data, but that was certainly the exception. There's really no reason to go reinventing the wheel here. Moreover, they already have a pretty large amount of data from the ADS-B telemetry. That's only half the story from the FDR, but it's still a huge improvement over 3 years ago.

And the best answer to your question is that any design change on an AC takes years of effort. Things as mundane as the inflight entertainment system take years and years to role out. As a rule, major changes like that only happens when a whole new family of AC are introduced so all of the regulatory stuff can be done at the onset. Once you consider that those regulatory things have to be done for a huge number of governing bodies then you're looking at a ton of effort for a system that A: you try your hardest to insure never gets used, B: only gets used in the tiniest fraction of flights, and C: works surprisingly well in the rare instances it is employed.

Reinvent the wheel?  How many planes have gone down in the last year and have had issues understanding what happened? We advance all technologies, why is re-thinking the black box idea not valuable?  Of course it will take time, that shouldn't be an excuse not to.  Technology is much more advanced now.  Keep them in place, but add more for real time data being sent "home" It makes too much sense. 

[rumborak]

It's a complex thing. I mean, sure, every passenger's smartphone far outstrips the processing and storage power of the black boxes. The problem is the requirements on them. They need to survive hell and back, and store their information for months with almost 100% reliability. Proving that takes years of very expensive certifications.
And the other factor is, as nice as it is to know why the accident happened, the device is needed in extremely rare cases (I read yesterday that this crash was the first for the A320 model after 80 million flights). Spending millions of Euros/dollars on debugging freak accidents is not exactly great ROI in an industry that has 0.5% profit margin.

[El Barto]

To the best of my knowledge there has only been one accident that remains unresolved, and that's the titular Malaysian plane. Like I said, FDRs and CVRs actually work surprisingly well. I think the problem here is that we live in a day and age where people demand answers immediately. This accident, like others, will take several months to issue a formal finding and that's honestly alright. If they discover a fatal flaw early in the process they'll work that problem sooner (as happened with the 737 problem a few years back).

Also, FDRs capture a minimum of 88 parameters (and >300 is the norm, I believe) with a polling rate around 3/second. That's a lot of data to stream in realtime when you're having to capture every AC in the air at any given time (7000+ in the USA alone).

[rumborak]

And it would only help if you're one water anyway. Over ground, even in an inhospitable area such as the Alps, it took them half a day to retrieve both boxes.

[cramx3]

Im not talking about replacing the black boxes, those things are rugged and do work.  Im talking about adding something so we can find lost planes and understand what happened faster and I don't think "it takes time and money" is a good excuse.  If we can add satellite TV to planes then I think we can add more technology for better tracking and real time data.  I should also add that supposedly the black boxes were never found from the planes during 9/11.  I think we would all love to have a better understanding of what went on with those planes.

[El Barto]

Im not talking about replacing the black boxes, those things are rugged and do work.  Im talking about adding something so we can find lost planes and understand what happened faster and I don't think "it takes time and money" is a good excuse.  If we can add satellite TV to planes then I think we can add more technology for better tracking and real time data.  I should also add that supposedly the black boxes were never found from the planes during 9/11.  I think we would all love to have a better understanding of what went on with those planes.

Yeah, but ADS-B is already doing what you want. The chart Rumbo posted shows a fair amount of flight data and was pieced together within minutes of the incident. And that satellite TV thing is occasionally cited as an example of how hard it is to change something on an AC. It took years to get regulatory approval for that and wound up costing a fortune.

As for the black boxes on 911, that's an interesting point and one which I always found fairly suspicious. In the two cases where giant building fell on top of them I can understand it. In the other two cases it's one of the many peculiarities that give truthers ammo to go on about (and one of the better ones, IMO). I suspect those were salvaged.

[rumborak]

@cramx3: Well, those devices are held to a much different standard though than TV screens. Entertainment stuff is in that "nice to have" category, and they can fail all they want (and often do as everybody knows). Telemetry are likely heavily controlled by FCC regulations.
To give you an example, I personally work on an Android app that soldiers will end up using in the field (not on the front line, only in benign situations). Nonetheless, we had to prove that our app can withstand a whole week of continuous use! Whereas the realistic use case of our app is 20 minutes at a time, at best.
I would think the requirements on telemetry equipment on a plane will be even harder, by orders of magnitude.

EDIT: I think the WTC black boxes simply didn't survive. Regarding the Pentagon ones, they're probably classified simply by virtue of their resting place.

[El Barto]

Actually, the failure of those Inflight Entertainment Systems was deemed a huge problem. For one thing they involve miles of wiring per AC. That's the sort of thing that gives everybody the willies. Furthermore, satellite systems are going to have to be guaranteed to not interfere with avionics and communications. In the case of the wiring, they're built as a completely separate unit from the rest of the AC, completely isolated. Obviously RF interference is a paramount concern.

The plane that crashed out in the field is also suspect with regard to black boxes. There were instances of The Man ignoring or deflecting inquiries on the basis of "what difference does it make, anyway!" I suspect the boxes from both of those AC fell into that "you don't need to know about that" category (which is of course what feeds the truthers).

[cramx3]

I totally get that with the usage, but my point in relating them was in terms of costs, time, and purpose.  We can spend significant money and time for entertainment on a flgiht, but not on things that may help locate a potentially lost plane?  And if ADS-B does what I am asking, how come the malaysia plane is still missing?

If the black box didnt survive in 9/11 then it doesnt appear to be doing its job (albeit in an extreme circumstance, but isnt that the point of them?).  I think in general they were found and just classified, but thats my opinion and not meant for this discussion.

[El Barto]

And if ADS-B does what I am asking, how come the malaysia plane is still missing?

Valid question. The biggest problem is that it relies on ground based reception. When you're over the water you don't have that. I suspect that'll change soon enough, but I believe it's still relatively new, and of course subject to a long delay in full adoption.

Insofar as satellite reception, you do have ACARS which can provide some useful telemetry, but it is utilized for an entirely different purpose. More importantly it's implemented by the airlines and as such only reports back what they want it to. It wouldn't surprise me if there were enough clamoring to have it become more of a requirement to receive basic data, but then you run into the problem of forcing an industry that operates near the margins to implement new requirements that, once again, will almost never be necessary. It was ACARS which provided most of the data for that Malaysian plane, but as we saw, it wasn't exactly a goldmine.

[rumborak]

I totally get that with the usage, but my point in relating them was in terms of costs, time, and purpose.  We can spend significant money and time for entertainment on a flgiht, but not on things that may help locate a potentially lost plane?  And if ADS-B does what I am asking, how come the malaysia plane is still missing?

Airline A: "We spent $1 million on upgrading our entertainment system!"
Airline B: "We spent $1 million on reporting more accurately in the case of a crash!"

Which airline gets the customers?

[cramx3]

I totally get that with the usage, but my point in relating them was in terms of costs, time, and purpose.  We can spend significant money and time for entertainment on a flgiht, but not on things that may help locate a potentially lost plane?  And if ADS-B does what I am asking, how come the malaysia plane is still missing?

Airline A: "We spent $1 million on upgrading our entertainment system!"
Airline B: "We spent $1 million on reporting more accurately in the case of a crash!"

Which airline gets the customers?

   well when you put it that way...

[bl5150]

Early reports on the black box indicate that one of the pilots was locked out of the cockpit and trying to smash his way back in.

[El Barto]

Suddenly I find myself really hoping the FO was named Dieter von Hamburg and not Mohammad al-Bagdahdi.

[rumborak]

I wonder if this will be a case what we will never know what happened because audio recordings are very limited after all. It implicity assumes that people are talking.
Could be plain old suicide for example.

[El Barto]

Nah, they'll sort it out. We're not hearing anything official at all about the CVR, only a leak from the NYT, so it's likely they know a whole lot more than just the lockout (assuming the lockout is even real). Aside from that there's a ton of details that just need to be sorted out. In fact it's bording on information overload at some of the regular haunts for speculating pilots.

The CW among airbus drivers is that you don't just get locked out of the flight deck. It'd take a deliberate act from the person inside. However, how hard it would be comes down largely to the airline and their protocols. Pilots are reticent as fuck to discuss those protocols online, but the airline will certainly share the information with investigators.

And for the curious minded, here's how the door locking panel works for an A320: https://www.efbdesktop.com/airplane-general/sys-1.3.1.html

There's also some pretty good conjecture from a 320 pilot about the airspeed and the decent that suggests that it was definitely controlled flight; specifically a pilot fighting the FD. He actually posted this pretty early yesterday and posited Controlled Flight into Terrain. Again, just conjecture, but it demonstrates what can be reasoned with even basic transponder telemetry. At this point I suspect the folks at Airbus are doing the same thing and ascertaining exactly how that telemetry they have jibes with their own FD programming and will know (if they don't already) why the plane descended in the manner it did. The logic that makes the Airbus FBW system is insanely complex, but also remarkably easy to understand when you're looking backward at a problem like this. We know that all A320s will behave exactly the same way under the same circumstances, and that behavior is based on very specific condtions. That makes it easy to reason out.

Man, I guess it's a good problem to have, but we're really spoiled nowadays. Here in 2015 we want answers within a day of something like this happening. There have been investigations that took years to piece together, and this one is very likely going to prove quite simple to resolve. Yet we're already clamoring for answers and wondering if they'll even come.

[rumborak]

The more I read about it, the more it looks like suicide. The fact that you can't lock out a pilot without actively locking the door from the inside (and relocking, since there's a timeout), the fact that the plane on autopilot would have maintained altitude... it all seems to point to deliberate crashing.

[MetalMike06]

Bear in mind, the "pilot locked out of the cockpit" scenario has only been reported by news outlets, not any official investigation.

I have to agree though that it does seem suicidal. The descent looks so controlled. The only other scenario I can imagine off the top of my head is that maybe they had a rapid depressurization, dialed a low altitude into the MCP, iniated the descent, but maybe got  forgetful of their surroundings as they were running through checklists. That seems insanely unlikely though; I really can't imagine that someone can become the captain of an airliner and accidentally run a plane into mountains when you have this kind of equipment on board: The A320 (and most modern airliners, to my knowledge) have a terrain display:

...in addition to automated verbal callouts "TERRAIN, TERRAIN, PULL UP!"

So right now I can't see how this wasn't deliberate unless they were somehow incapacitated after they initiated a desent.

[Stadler]

I think that the thought process is that you want the black boxes to remain with the bulk of the AC. You find one you find both. Ejecting them doesn't really make much sense, and while floating them does in some ways, drift would cause them to travel pretty far from the wreckage. Plus, they're going to be diving to bring up important bits of the AC and bodies anyway. That said, it does seem like a fuselage mounted marker body that detaches from the AC when submerged seems like a swell idea.

Oh, and the next step in the evolution of telemetry is going to be to have the AC burst transmit the FDR/CVR data to satellites. That will provide timelier info but will likely introduce other problems, as well.

el Barto, you're my resident air travel expert, so I'd love your opinion (though I think you've already agreed to this in your own diplomatic way) but I read threads like this where everyone has the next "bright idea" ("we should have orange boxes with drone capability that transmit in real time the biorhythms of every crew member to a server farm we can install on the moon.  It's 2015 for God's sake") and it never ceases to amaze me that it doesn't occur to people that yes, while mistakes can sometimes be made and there is always room for improvement, this is one of the most well-thought out processes we have.   If a 6-hour ejecting data recorder would be better than what we have now, the odds are strongly in favor of the idea that we would already have it installed.

[millahh]

Well, the French government is saying this was a criminal act.  This would lend credence to the details NYT is reporting.

This is all kind of horrifying....

EDIT:  "Screams can be heard before the final impact"  "Death was sudden and immediate"

And apparently there is a code the pilot can use to get back in through the locked door, but can be manually overridden from inside the cockpit.  So this would have required at least two very deliberate actions on the part of the co-pilot.

[El Barto]

Alrighty, so they've announced that it was the FO, it was a murder suicide and that he was a German. Sorry, but I'm honestly glad to hear this  (that he was a Kraut, not the suicide part, obviously). Not the first time a pilot has done this and likely won't be the last. Just glad it won't lead to more kneejerk reactions (and the pilots are pretty quick to point out that locking the doors in the first place qualifies as such, which has now led to the deaths of 150 people).

Millah: the link I posted explains the double locking mechanism. In "locked" mode it disables keypad entry for 20 minutes or until it's deactivated. It's also possible that there's a sliding bolt (can't beat the classics). That said, the pilots over at PPRuNe are pretty clear that with company knowledge this shouldn't be a dealbreaker. They're just also not keen to explain that. It should also be pointed out that there's a fire axe in the FD to be used for: A. breaking out windows in the event that it becomes necessary, and B. splitting the head open of the first guy through the door. Not a factor here, obviously, and moreover the FO was having to manually fly the plane down since the Flight Director would be doing it's damnedest to prevent it. He wouldn't have had time to physically defend against entry.

Stadler: you run into a situation with improvements like this since the airlines operate close tot he margins. For something that needs to be changed it has to be damned cost effective and honestly, improving something that happens after a crash doesn't qualify (unless it's to help survivors, of course). That's not to say that if it's a benefit that it will be done, though. There are plenty of things that the airlines install only after ages pass. Wind shear radar is an example of one of those, IIRC. Adding a new radar system to an AC is ridiculously hard/expensive to do and despite the obvious safety bonus it just didn't work out once the numbers get crunched. One of the problems here is that the FAA has the dual purpose of protecting safety but also promoting efficient air travel. They tend to make recommendations and very rarely issue directives. It's mostly the latter that the companies act on.

As for black boxes, I suspect that eventually they will be reporting back in realtime. There's just no priority for it. The current system is actually quite effective. Considering that they almost always survive and provide useful information on events that happen with increasing infrequency, it's hard to really make a huge case against them. The impetus is going to be something like ACARS, which benefits the companies in their own competitive way, but also provides useful investigative telemetry in the process.

The other thing to keep in mind is privacy. Pilots are understandably reticent to have their final words broadcast all over the world for the sake of morbid curiosity. Very rigid standards were put in place to protect them, which is why you only get transcripts (and occasionally not even that) and never the actual CVRs, which used to be made available. Given the rarity of something like this happening that doesn't trouble me much. None of us would want a camera/mic on us the entire time we're working, and the safety concerns here aren't nearly enough to warrant making pilots an exception to that.

[El Barto]

Oh, and there is one change that will be implemented after this. It varies by company but there are strict procedures for when a pilot has to use the head. Some companies mandate that a FA has to enter the FD and sit in the jumpseat while a pilot is out, that way he can make sure the exited pilot can get back in. I suspect we'll see this implemented across the board. In this case the FA would have been a part of that aforementioned axe scenario, but some fighting back can go a very long ways. Just ask Auburn Calloway. (And for people unfamiliar with it, this is serious movie of the week stuff; amazing story.)

[Scorpion]

I might be missing your point here, but why are you glad that he's German?

[Sir GuitarCozmo]

I assume so that the media and everybody else are less likely to go all "MUSLIM TERRORIST HIJACKING, ZOMG!!!"

[MetalMike06]

Completely unrelated to the crash, but it's pretty crazy that a new pilot in Europe can go straight from receiving a fresh commercial license/ATPL to the right seat of an A320. This guy barely had 600 (yes, six hundred) hours TOTAL flight time. In the states, at that point, you're still instructing in a small single or twin-engine prop with still at least another year or two ahead of you before you get into even a small regional jet. I'm jealous...in a way; I'd still like to have ample experience before I have that many lives in my hands.

[El Barto]

Completely unrelated to the crash, but it's pretty crazy that a new pilot in Europe can go straight from receiving a fresh commercial license/ATPL to the right seat of an A320. This guy barely had 600 (yes, six hundred) hours TOTAL flight time. In the states, at that point, you're still instructing in a small single or twin-engine prop with still at least another year or two ahead of you before you get into even a small regional jet. I'm jealous.

That's actually a pretty hot topic in the industry. The flipside of this is that there are a ton of US airline pilots living within the poverty level. Look up "food stamp pilots" and you'll see. Since the regionals (and even the big players) know that time is so important they can hire pilots for minimum wage and do avail themselves of that opportunity. The conventional wisdom is that being a pilot is an awesome job but one of the worst careers imaginable.

The truth is that under normal circumstances doing your training in the right seat isn't a big deal. The whole process is automation and rote procedure. The problem is when you get something like the Air France accident a few years ago and you discover that the short-timers never developed rudimentary stick and rudder skills.

[MetalMike06]

Oh I know. I believe the starting pay for a first year regional airline pilot is roughly $20k a year. It increases pretty significantly though the next year and so on.
At the last airline I worked at, I talked to a pilot who used to fly a Beech 1900 for Great Lakes Airlines...$13,000 a year, not kidding.  You'd make more money than that fueling their planes or being a ramper.
The regionals are feeling it right now though; even in light of the 1500 hr rule, they're getting less and less picky because they need people so bad and fewer people, for good reason, do not want to become pilots. I think they'll have to fork over a little more at some point.
I really think they only got away with this because there's a million people out there that wanna fly a shiny jet for a living, and they'll essentially work for free to do it. That might be a stretch, but if someone turns down an offer, chances are the next guy/gal will be willing to do it for less.