I'll clarify. One, it is NOT "strictly against procedures to be doing this, and a termination offense" to teach a fellow crew member about PBS while on duty, especially if you use a pen and paper. The no laptops thing is not a federal reg, but an airline specific policy. An example of federal regs specifying prohibited behavior in the cockpit is the 'Sterile Cockpit' rule - essentially, during critical phases of flight, non-essential conversation is prohibited. This is defined to include all flight below 10,000' MSL, taxi, takeoff, and landing.
I'll give you my personal perspective. Nobody obeys all the rules. At every company I've ever seen, people are always violating SOME rule or reg, generally not intentionally. Flying is not an exception. When you spend 10,000+ hours in an aircraft, when you know every quirk, knob, switch & controller, you get a pretty good sense of what's important and when it's important, and when it's not. When we were still in the 'learning phase' of aviation (up until the 1970's, roughly), the expertise of the pilots up front usually exceeded the expertise of the regulators by a large amount (kinda like IT today...). So, the Feds did something very intelligent - they came up with FAR 91.3 & 91.13:
91.3: (a) The pilot in command of an aircraft is directly responsible for, and is the final authority as to, the operation of that aircraft.
(b) In an in-flight emergency requiring immediate action, the pilot in command may deviate from any rule of this part to the extent required to meet that emergency.
(c) Each pilot in command who deviates from a rule under paragraph (b) of this section shall, upon the request of the Administrator, send a written report of that deviation to the Administrator.
This is the authority to do as the PIC sees fit. It is the entire basis of a Captain's authority - and it assumes that Captains know better than any one else how to command their craft. This has generally proven to be a good assumption - since the pilot is usually the first to arrive at the scene of an accident, he/she is highly motivated not to be in one... if you pick your pilots carefully. But to make sure that this authority is not abused...
91.13: (a) Aircraft operations for the purpose of air navigation. No person may operate an aircraft in a careless or reckless manner so as to endanger the life or property of another.
(b) Aircraft operations other than for the purpose of air navigation. No person may operate an aircraft, other than for the purpose of air navigation, on any part of the surface of an airport used by aircraft for air commerce (including areas used by those aircraft for receiving or discharging persons or cargo), in a careless or reckless manner so as to endanger the life or property of another.
The key words are 'Careless and Reckless Operation'. What it means is, if you fly in a careless or reckless manner, the FAA will nail you. Your actions, even if they don't specifically violate another reg, if they are 'careless' or 'reckless' - violate this one. How do you know what's careless & reckless? Well, if you have an accident, the presumption is you were careless, or reckless. These guys? They will DEFINITELY be hit for the 'careless' part. This is a catch-all - it means, at any time, your actions as a pilot can be held up to scrutiny, and if they are found to be 'careless or reckless', you are toast.
These laws were very good - it allowed pilots to experiment with how to run a plane, but held them responsible if something screwed up. Over time, best of breed techniques were discovered, spread from crew to crew, standardized by airlines, and ultimately adopted as Federal law - the FARs. Of course, accidents, once their root cause was discovered, frequently led to regulations as well. Overall, I think this was an excellent way to turn a fledgling, risky industry, into the mature and extremely safe one it currently is.
Techincally, they should also be monitoring Guard (121.50 MHz) the whole flight, and I'm pretty sure multiple attempts to contact them on Guard were made. But since they didn't bother paying much attention to the primary, they probably didn't montior Guard that well either.
There is really no excuse for both pilots completely losing situational awareness like this. They're both toast, and deserve to be.
As for the scheduling system they were going over - actually, that is probably the 'news for nerds' part. The old airline schedules were built in two units - 'pairings' and 'lines of time'. A pairing is a group of flights, typically from 1 to 6 days long, that begun and ended in a pilot domicile. The word 'pairing' was to indicate that an entire crew was 'paired' together that whole time. A line of time (or simply a line) was a month-long group of those pairings. There is a long list of legal requrements (min rest, max flight time, union contractual obligations, aircraft mx requirements, etc.) that these schedules had to meet.
Ultimately, from the pilot's point of view, these lines were published each month for the next month. Bidding was very straightforward. If you were the number 1 senior pilot in that base (technically, domicile, aircraft and status (capt. or F/O), you picked your line, and that was that. If you were #2, you picked your schedule, and got it.... unless the number 1 guy already got it, in which case you got your second choice. If you were number #300.... well, picking 300 schedules in the order you want them was a time consuming task, but the outcome was perfectly transparent. The line awards were public, so you could verify that the schedules you didn't get really did go to senior people. You can debate whether such a system is 'fair', but at least it is clear how it works, both globally and month to month.
Then, with the advent of more powerful computers, a system called 'PBS' was born - Preferential Bidding System. These systems, instead of having hard, published lines you bid from, instead only published the pairings. You expressed your 'Preferences' through a computer language. A computer program then ran, taking everybodys preferences, seniority, system constraints, etc. into account and generated schedules.
In theory, PBS sounds great. A pilot's preferences generally don't change that much month to month, so you could file your bid away and let it run automatically each month with little or no tweaking.
In practice, it's usually been highly disruptive and caused great angst for a year or two after being implemented, for many reasons:
1) The language used to express your preferences is generally designed for the programmers, not the users.
2) The results can be, to put it mildly, unexpected. When you have pre-published schedules, you have a pretty good idea ahead of time what to expect.
3) There are no month-to-month conflicts that generate additional days off, resulting in more work per pilot, a reason the airlines like them and pilots don't, on average.
4) Non-computer savvy older pilots (Captians) have a harder time getting it than younger pilots (F/O's), on average. It takes a vastly important piece of your life (when are you working? Where are you going? 28 hours in HNL or 32 hours in XNA?), and makes it tied to your comfort with learning, essentially, a primitive computer language.
I cringe when I see this, because I've done this - taught Captians while flying about PBS. So have many other F/O's. You just prioritize it where it belongs - below aviating, navigating and communicating. These guys made everyone else look bad.
I understand. I disagree with the death penalty as a matter of principle, but there are some people in the world who I would struggle to say don't deserve it.
...that hits all the same notes. E-books will take over the world, why are the German publishing houses sticking their heads in the sand, etc. I've thought about it quite a bit, since I have a strong personal preference for printed books, and have debated the topic with passionate advocates of e-books. I've come to a few conclusions:
1) The advantages that printed books have over e-books in terms of convenience will go away over the next 15 years. Limited resolution (200 ppi e-ink vs. 600+ ppi for print), limited battery life, bulk, storage capacity, etc., not to mention cost (not just direct, but transportation, storage, disposal, etc.), will all favor e-books in 15 years. Resolution (my particular nit) will probably take the longest, but it will happen.
2) I doubt a personal e-book 'reader' will last long in the marketplace. It's too big and bulky to be 'just' an e-book reader. Why not make it a web-browser? 95% percent of what you need to do that is there. E-mail? Terminal access? A cell phone with a bluetooth earbud? A movie watcher? It will become a general purpose computing device just like cell phones are becoming.
3) It won't succeed until an Apple-like company makes it so stunningly easy to use and manage that its advantages are clear. A cellphone and a smart cellphone are quite similar, so the idea of an iPhone/Treo (a general purpose computer that happens to be a cell phone) was not so hard to get accepted. A tablet-like device has no commonly existing parellel right now, and the existing examples are weak, to put it mildly. It will have to be wildly simple and pleasant to use...
4) Once most books are no longer printed, it remains an open question whether it will make censorship of ideas easier or harder. I haven't been able to come up with a convincing argument either way. DRM is also still an open question, although you can make a good argument that a DRMed device will fail in the marketplace. Maybe.
There will be a great e-book reader one day, but it won't be called that. It will be part of a package that can do far more.
...the U.S. had the greatest rise in its living standards. Scientists, engineers, and other professionals from all over the world migrated here in seach of a better life, the opportunity to live pretty much in peace and quiet, or simply to survive. It was seen as the most desireable place to live in the world, and that seemed to become a self-fulfilling prophecy as 'the best and the brightest' came here to do their best.
I wonder, are more folks returning to their home countries' simply because of money and career advancement? Or do they feel less welcome in the culture? Or perhaps their own home cultures are changing to where they feel they can shape them for the better?
This seems more like an anecdote than a study; but there is something wrong when science and engineering and other technical fields are seen as undesirable by most Americans, and the immigrants who come here to learn them decide that they'll have better opportunites back home to use them.
It'll be great news for touch-up artists who still have their old-school airbrushes, though.
Perhaps a bit off-topic, but back in the 50's, my father worked for the National Enquirer as a photo retouch artist, with the air compressor/blower and all. His English was marginal back then, so many of the editing 'instructions' he got were quite confusing to him. Finally, his boss figured out how to explain it so he would understand 'the idea' behind the requested edits - "Make the boobs boobier, the flames flamier, and the blood bloodyer!" Suddenly, it all made sense.
He always said it was the greatest job he ever had after that!
I can't really argue with the first part - this part, I have a small nit to pick:
At any rate, flash drives containing thousands of books each spread accross the entire earth are actually better for archival purposes than paper. We can never have another burning of the library at Alexandria; there are too many copies. And the costs of perfectly preserving old data are effectively zero, as storage costs continue to plummet. We don't need monks dutifully making copies of scrolls as the scrolls age. We just copy our book collections instantly and flawlessly to the new storage tech as it becomes available.
You are assuming that all the books are recopied on a regular basis. NASA has had quite a bit of trouble retreving information from the magnetic tapes which stored the data from the Apollo missions, not so much because the tapes demagnetized, but because no records documenting the format of those tapes were located, and the designers of those formats weren't around anymore. Format loss is a signficant problem in the long term. In 50 years, do you think USB ports, FAT32 formatted flash-drives and UTF-8 will all still be in common use? Whenever I see someone use the words, 'never', ' instantly', and 'flawlessly' and other absolutes, I get suspicious of their argument. When it comes to preserving things for durations beyond a human lifespan, electronics are still relatively untested in practice.
But deep down inside, I know that e-books will probably win the day in the next decade or two; much like horses, physical books will probably become a niche item someday. Unless we screw them up royally with DRM or something. Amazing Xerox didn't appoint an armed guard by each photocopier they made to protect against possible copyright violations!
I'm in the same boat - my girlfriend and I have a combined library of around 2,000 books. There is a certain pleasure in browsing around, finding something you haven't read for a long time (or was from her library rather than mine)...
I realize that just about every argument I come up with for the superiority of printed books is pretty weak; the only four that are significant are 1) Books (good ones) can last 100's of years, 2) No power source needed, 3) Readability better (600 dpi+ vs. 160 dpi for typical e-book readers), 4) Better chance to survive civilization collapses.
Even these items are a bit of a stretch, and are likely to change within a decade. I suppose people with actual, physical libraries of books may be considered eccentrics fairly soon....
Actually, I'd be curious as to a reference for this. While I'm sure you are right once you start talking about 1000's of books, I'm equally sure the production of 1 paperback book is not 'extremely wasteful and bad for the environment' compared to one e-book reader. The key, of course, is how many books you read on your e-book reader before it, too, becomes e-waste.
A little googling revealed this Master's Thesis on exactly this topic. I haven't read it in-depth yet, but it looks to strongly favor e-readers.
This is a wonderful thing. It may make it much easier to publish new, low circulation books as well, since you don't need to reach a critical threshold sales number to make it worth printing. Of course, a 'book' (as in the physical form) may become obsolete over the next few decades as old curmudgeons like me who like reading printed material far more than reading off a screen drop off...
Sounds great. Let me know how that goes, okay? Tell me when the war is over, and teh stupid is vanquished. I'll hang a banner for you.
People with this attitude will have a hard time working with anyone outside of a very small group of very competent people; i.e. in the real world. Most people really aren't experts, they aren't always right, they frequently make mistakes.... but they are not evil. I try to reserve the word 'evil' for people who seek to hurt others for fun.
If all you have a choice of is 'competent' or 'nice', I suppose I would temporarily choose 'competent'. I would then seek to find a little more of BOTH in one human.
They are afraid terrorists will get a hold of the schedule and to keep that from happening they are going to stealth the whole process. Buses and trains will now be randomized. Numbers and routes will change spontaneously. Sometimes trains will run on bus routes and buses on train routes. Every once in a while one (either a train or bus) will cross over to NJ, drive off in the pine barrens on its own and self destruct on the off chance it is carrying a terrorist. That will solve everything.
The airlines already implemented that plan. Sadly, nobody noticed.
Flying manually without autopilot in the turbulence is like driving at 100mph on icy road without electronic traction control. I still think computer is in a better role in handling that.
Sigh - this is wrong. Many airlines recommend punching OFF the autopilot and autothrottles in severe turbulence because it will generally try to maintain altitiudes and courses when the best thing to do is minimize thrust and attitude changes to minimize loads on the airframe at the cost of altitude and course deviations. Similarly, when stuck in icing conditions, particularly in holding, many airlines recommend hand-flying to detect changes in flying qualities that may be occuring as a result of icing accumulation. The accident at Roselawn, IN lead to this procedure.
An autopilot is an aid to a human pilot. It has no decision making ability. It flies in certain conditions 'better' than a human, and 'worse' than a human in others. It is, ultimately, controlled by the human with an abstraction layer; instead of 'pull back stick to make nose go higher, which will probably result in a climb', it's 'select level change, set the level-off altitude to 15,000, and let the autopilot pitch and power to get me there'.
From what I've heard, you can get a payscale reset if you do lose your job, as well, since the payscale is almost entirely based on seniority.
The operative word is 'do' - you do get a payscale reset if you switch airlines - all of your previous experience just qualifies you for the new job at entry level. One of my least favorite features about the airline industry. You are generally paid by the flight hour, and the rate per flight hour is on a table, where the row is your years of service or longevity AT YOUR COMPANY (typically topping out at 12 years), and the column is your equipment and position; 757 CA, MD80 FO, etc. You can see these tables at airlinepilotcentral.com. Keeping in mind that the typical airline pilot flies about 800 hours a year, maximum of 1000, and you get a good idea of pay. Corporate is quite different, from what I hear (your pay has much less to do with years of service and more to do with equipment type and position), but they have different issues relating to corporate flight departments being favorite targets for accountants, consumers, cost-cutters, etc.
While it's true that starting pay is low, it's not in the $15K range; more like $20-$25K
Yeah.
Commutair - first year F/O, $18/hr * 1000 hrs. (absolute maximum a year) = $18k.
CapeAir - first year F/O on the Cessna 402, $9/hr. That's $9k a year, if the bugger flies 1000 hours in that thing.
Great Lakes - first year F/O, $16/hr -> $16k a year.
Take a look at a few others...
When I started at Continental Express in 97, flying ATR-42s and -72s, my first year pay was $13k. That doesn't include the $10k I paid for my own training - they had a great scheme going. If you fail out of training, well, you don't get the job, and you don't get the $10k back! If you pass, well, they just got a new, qualified pilot for 'free'. Also, see how far the princely salary of $30k/year goes when your domicile is SFO, EWR, or BOS. Here's the reality - you will be at poverty to lower-middle class wages for about a decade in a typical aviation career. It could be much shorter, if you get lucky. It could easily be much longer. It might be interspersed with periods of no pay, when you have been furloughed for some reason (e.g. 9-11, $140/barrel, a merger, bankruptcy, etc.) If you continue to be lucky and persistent, you can start making over $100k a year. It takes a long, long time to get there. So, frankly, spun has a point. I'm glad that your family of pilots has done quite well, but I'm quite sure you probably know a few pilots, and their families, who have not.
The bigger issue is why pilots flew the plane into a thunderstorm.
I'd like to talk to you a bit about this, seeing as I just spent all day avoiding doing just that, and only partially succeeding.
The problem is the following. What is a thunderstorm? Where is the edge of it? Laterally? Vertically? How do you detect these edges? How rapidly are they changing? How do you transmit all this data (assuming you can get it) to a pilot in a form that is useful and timely? This is actually an extremely difficult problem that is not solved. Whenever you hear a weather forecaster predict 'scattered thundershowers', he/she is admitting ignorance in determining exactly where the thunderstorms will and won't be, and instead provides a probabilty which correlates roughly to what percentage of an area will be hit by thunderstorms based on models in a given time period. I think you can see that's of little use in actually avoiding the real, specific thunderstorms that pop up, although it might be somewhat useful in planning - as in, "50% of Pennsylvania is going to be covered with wx, it's probably a good idea to carry extra gas to route around the entire area."
Thunderstorms have a life cycle; generally they have a growth stage, a mature stage, and a dissapating stage. While an airmass thunderstorm may go through all the stages in a hour or so, thunderstorms associated with a squall line can last in the mature stage for many hours. Each stage has very different characteristics, with different dangers. Severe turbulence that can rip an airplane apart, or simply put it out of control (after which the ripping apart may occur) is common; lightning strikes can knock out instruments and avionics; hail can physically damage the airplane; severe updrafts, downdrafts and downbursts can physically put the airplane at an altitude above it's coffin corner or into the ground... so let's agree, for the purposes of this discussion, that thunderstorms, in whatever phase, are dangerous. How do we detect them, so as to avoid them?
The best way, in terms of confidence and precision, is by eye. The towering cloud of condensing water vapor is incredibly distinctive, and gives a clear marker where not to go. Unfortunately, this only works if the visibility between the thunderstorms is excellent, which is frequently not the case. As a hot summer day wears on, the thunderstorms that have bloomed and dissipated have blown so much moisture high into the atmosphere that the visibility frequently is nil. Nighttime also does a great job of obscuring things. So, visual detection, while great when possible, is quite often not possible.
Onboard weather radar is the most common tool used in airline operations to avoid thunderstorms. The radar reflects off of (generally) rain and hail, and weakly off of snow and ice. Unfortunately, not all thunderstorms have much rain to detect, particularly in their growth stage! In addition, the radar is tilted at a specific angle above or below level, so really only sees a horizontal slice at a time; since the earth curves, and the altitudes of the thunderstorms curve with it, the height above ground of the horizontal slice varies with distance, making it exceedingly tricky at times to decide whether a radar echo is a threat or not. In addition, heavy rain is not the threat - turbulence, strong drafts, and lightning are. How does one decide whether the echos are part of a convective cell or a broader rain producing system? What if they are mixed together? In addition, at typical crusing altitudes, most of the rain visible to the radar is below you - but if you tilt the radar dish down to find it, you hit the surface much sooner, and lose the ability to detect cells further away...
Obviously, a real-time data link to ground doppler radar would be incredibly helpful. Unfortunately, it's not required by law, and it's expensive, so most airlines don't have it. Lovely. It probably wouldn't have done the Air France guys m
Actually, when everyone was worshiping that guy who crash-landed a jet into a river (Tully? Sully?), I was wondering what the hell the big deal was. I find it odd that our expectations are so low that we merely brand competence and doing your job well as heroism. I would expect most pilots of large passenger jets to be up to similar feats. If they aren't, I'm very scared.
Dude, that wasn't competence, that was a tour de force of skill. There is no airline training for a twin engine flameout on departure, follwed by a deadstick landing into a crowded, narror river. It's not supposed to happen. It could have gone wrong in a thousand and one ways, from hitting the water at an angle and tumbling the aircraft, hitting a boat or barge, touching down with too much speed, chosing not to land in a freezing cold river but attempting to make a piece of pavement, recognizing both engines were toast and not wasting time in a fruitless attempt to restart them, etc. He had been handed a slender reed of luck - that the bird hits occured at several thousand feet MSL, it was good, daytime VFR, and he was close to the Hudson - if any of those conditions were not so (IFR, below 1000' when the engines bit it, etc.), he would have had no chance at all. He had no room to make any mistakes, and no do-overs. He did it all right, made all the right choices, the first time, and the only time, he had ever been in that situation, and not a single person died. If you aren't impressed by that, you haven't talked to your step-brother enough about what it's like.
In the airline world, if a student in a sim ride has passed his checkride/PC/AQP/whatever and there is extra time, the instructor will frequently do a Kobiashi Maru on the student just to see how they react - -60 knot windshear on departure, followed by a complete loss of all instruments but the peanut gages and then an engine failure, etc. They try to load them up with everything they can, partly for fun and partly to see how far the student can go. It almost always ends in a crash anyway, but that's part of the point, to never, ever give up making smart decisions while controlling the aircraft to the best of your ability. Sully got a real-life Kobiashi Maru and pulled it off.
But last time I checked, most commercial airline crashes were due to technical problems, and not pilot error.
You have it in reverse. Over the last 20 years, the percentage of airline crashes due to pilot error has been around 80% - planes have gotten far more mechanically reliable. Pilots have gotten better too, but to nowhere near the same degree as the machinery, sadly.
I would argue that "simple fact". IMHO any pilot who decides to fly directly into a large thunderstorm when going over it is a viable alternative has already committed pilot error, the computer probably let him fly further before crashing than he would have solo.
You don't have the slightest idea what you are talking about. Thunderstorms are such dynamic systems, and frankly our ability to gather data on them in a timely fashion is so limited, that there is not a single airplane flying in the world where computers decide what course, altitude, or speed to take to avoid one. So the option to 'trust your life' to one doesn't even exist, because their decision-making ability is so primitive. Computers are used to control airplanes, in the sense that a computer figures out the angle to pitch the elevator, how much fuel to meter to the engine, etc., in response to commands by the pilots, either through a 'direct' mechanism, such as the yoke and thrust levelrs, or through the autopilot, which is being directed by the pilots. Computers have almost nothing to do with the decision-making that occurs in a cockpit, and there is not even a system on the drawing board where they do, because no one has any realistic idea of how to make one.
I would rather trust my life to a computer whose bugs can be ironed out and which will always perform the same way in a situation than a pilot who may or may not have gotten enough sleep, be drunk, or somehow be distracted. I've seen enough car crashes to know that humans are not the godlike infallible beings that the anti-computer controlled planes group seems to be preaching.
Yes, performing the same way in a situation is great when all possible situations and the best way to get out of them are known ahead of time. That would have done zero good with the recent USAir incident in the Hudson. Unexpected situations crop up constantly in aviation - humans, while certainly not infallible, not even close, are still far better at dealing with unexpected situations than AI presently is. Failures are generally, by definition, unexpected, and have unexpected side effects. If you think AI is up to the decision making required in flying, why don't you join the AI team at MIT who is trying to develop fully computerized driving? Even though they only have two dimensions and a fixed set of paths called 'roads' to travel over, and the ability to pull over to the side of the road if the going gets rough, it's still rather difficult task.
Pilot error was blamed for almost 80% of crashes in '04...why would you want to trust your life to something that statistics alone dictate to be more likely to crash?
Now you are either trolling or just stupid. More likely to crash than what? A computer AI pilot that doesn't even exist that would avoid the 80% of the crashes above, but not fall victim to a single additional crash which a human would have dodged? What are you smoking? You might have just as well said: "Human error was blamed for almost 80% of autocrashes in '04...why would you want to trust your life to something that statistics alone dictate to be more likely to crash?" - does this sentence make any sense to you since I don't know of any other viable drivers than humans?
Oh, and FYI - going over thunderstorms is generally a bad idea, since they can grow at rates exceeding 6000 ft./min., far, far in excess of the climb rates of any commercial aircraft at the upper flight levels. As you fly higher and higher, the stall speed (effectively minimum flying speed) and the speed corresponding to maximum Mach (effectively maximum flying speed) converge, meaning you have far less ability to manuever or ride through turbulence. Perfect time to have a thunderstorm rise up from below and envelope you, eh?
Another factor to take into consideration is that not all airline pilots are experienced. I don't like to dichotomize (like the poor summary of the article, dammit KDawson) but a pilot's first storm could bring hardening experience or crushing defeat.
58-year-old flight captain Marc Dubois, who joined Air France in 1988, had approximately 11,000 flight hours, with 1,700 hours on an Airbus A330
58-year-old flight captain Marc Dubois, who joined Air France in 1988, had approximately 11,000 flight hours, with 1,700 hours on an Airbus A330
If true, that's rather interesting. Most airline pilots fly around 800 hours or so a year. That would mean he started relatively late in the commercial flying business (mid-40s') or he had a significant timespan in his career where he was not flying. Anyone have more details on that?
Yes, because in the middle of the night at FL350 when you see nothing but black outside intermittently illuminated by blinding lightening apparently from all directions, where you are groping around your way with radar, which is painting red in all directions anyway, and getting ATC clearance for significant deviations off of airway centerline can take minutes (at 8 miles a minute, that's rather a lot) because of the limitations of LF, is just so EASY...how could those dumb pilots have done that?
Oceanic night flight through the ITCZ is still a bitch. Radar is still primitive. People are tired on long oceanic flights. The ability to change your route by significant amounts over the ocean without delcaring an emergency is limited. The aircraft is already operating near its maximum altitude for fuel effeciency, so any deviations from that altitude and course will eat into its fuel plans. A computer making decisions instead of a human changes none of these facts, and computer decision-making in fuzzy situations is notoriously poor, as most AI designers will tell you.
One of the major justifications for envelope safety in normal mode is that the pilot can attempt manuvers and rely on the envelope safety to allow the maximum. For example: Suppose an aircraft is going to fly into a hill. The pilot pulls back hard on the yoke and engages maximum thrust.
In a conventional airplane, you have concerns about structural failures of the plane. Pull back too hard and you might damage the control surfaces. With normal law protections, the computer will allow you to pull back as hard as it thinks the is still within basic limits of aircraft safety, thus allowing the pilot to pull back without worrying about structural failure. Given the massive redundency in fbw systems, I am not entirely worried about a software glitch causing an issue.
This is probably one of the most UNSAFE features of 'normal law' flight in the modern Airbuses. It essentially detrains the pilot. You learn the idea not to worry about control inputs, since you can, in theory, never overstress/stall/spin the airframe. It means that if and when a pilot winds up in 'alternate law' flight, where those protections are not present, all of his/her instincts about how much input is enought but not too much are gone.
I feel the same way about autobrakes for landing (not antiskid, not wheellock protection, not RTO protection, just autobrakes on landing) - they detrain you when you fly with them. Braking properly in a large aircraft is a skill, and judging by the slow drizzle of runway overrun accidents, a declining one. The best way to keep that skill fresh is to exercise it on each and every landing, rather than wait for the magic autobrake to kick in, then manually overriding it anyway for the runway turnoff. If we are going to rely on skilled humans to be in charge of aircraft, the systems of those aircraft have to be designed to some extent to help KEEP them skilled.
BTW, your comment after 'There are pros and cons to both. On the whole, I think that' got cut off, so perhaps you addressed my concerns already - my apologies if this is the case.
The controller didn't lose his job or go to prison. He was just put on leave, IIRC. So the killer did the right thing, since the system failed.
It also doesn't matter if his mistake was intentional or not, since it resulted in so many deaths. If the controller had any honor at all, he would have committed suicide. Since he didn't, he needed to be killed.
Not everyone has a sense of responsibility. Just look at the people running our government and financial institutions. Now that this controller's been eliminated, it'll probably make other air-traffic controllers do a better job or find another job if they can't hack it.
That's fine, so long as I can execute drivers who speed, run yellow lights or stop signs, don't yield to pedestrians in crosswalks, and who even think of drinking before driving. With 40,000 fatal car accidents a year in the U.S.A. alone, perhaps if we execute some of the 25,000,000 drivers who are issued citations a year (perhaps 1 in 10 like the Romans - that's only 2.5 million a year), we might reduce the accident rate! It doesn't matter if thier mistake(s) were intentional or not, since they result in so many deaths./sarcasm
Air traffic controllers are generally highly skilled, responsible, underpaid professionals who do a thankless job cleaning up messes of overloaded and antiquated systems. If you believe extra-judicial executions are okay for them, guess what - you won't have an ATC system at all, because no one in their right mind would bother. In the real world, where serious adults are interested in systematic safety improvements instead of chest-thumping suggestions of violence, nearly all aircraft accidents have been decriminialized (the exceptions are made for 'careless or reckless operations' or willful violations of law), since nearly every professional in the aviation industry doesn't want to cause an accident, and is more likely to be truthful about all the circumstances that led to the accident if they are not in jeopardy. The results of these more candid examinations of accidents (look up ASAP programs if you want to learn more about them) have lead directly to multiple improvements in regulations, procedures, and the overall safetly level in aviation - in other words, they have saved lives.
Your casual suggestion that executing people is a good way to encourage others makes me wonder whether you are even serious.
Thanks for the interesting post - you sound like you have a keen appreciation of the politics there. So, let me ask you the question I've asked elsewhere:
How does a nuclear armed North Korea benefit China?
All the explanations I can come up with only make short-term sense. I don't think the Koreas' will remain seperate forever (or even for more than 50 years). A unified, nuclear armed Korea is probably not a good thing to have on your doorstep - why encourage it?
I DO think the North Korean government is pretty crazy, and once they get deliverable nuclear weapons, that gives them a big stick to use against anyone, including China! Allowing North Korea to develop nuclear weapons ultimately reduces China's control over them. Why would they do this?
North Korea and China are allies for strategic, historical reasons which are becoming less important. The basic relationship seems to be, "We give you arms, and food when you botch up your agriculture or have a drought -- you cause trouble/distractions for us when we ask". I wouldn't bet a lot on such a relationship enduring many decades.
The fact that North Korea would be a smoking hole within a half-hour doesn't seem to deter them much, but I think China might object to having Guangzhou or Hong Kong removed as well. That's the problem with nukes - their destinations are not hard-coded into them based on the political winds of the day they are done.
And, risking veering off into serious off-topic land: Why do you think Japan's culture is more compatible with the West than any other Asian country? I know it might be a long response...
Slashdot Mirror
Or, to paraphrase, Nature is subtle, but she is not malicious? ;-)
I'll clarify. One, it is NOT "strictly against procedures to be doing this, and a termination offense" to teach a fellow crew member about PBS while on duty, especially if you use a pen and paper. The no laptops thing is not a federal reg, but an airline specific policy. An example of federal regs specifying prohibited behavior in the cockpit is the 'Sterile Cockpit' rule - essentially, during critical phases of flight, non-essential conversation is prohibited. This is defined to include all flight below 10,000' MSL, taxi, takeoff, and landing.
I'll give you my personal perspective. Nobody obeys all the rules. At every company I've ever seen, people are always violating SOME rule or reg, generally not intentionally. Flying is not an exception. When you spend 10,000+ hours in an aircraft, when you know every quirk, knob, switch & controller, you get a pretty good sense of what's important and when it's important, and when it's not. When we were still in the 'learning phase' of aviation (up until the 1970's, roughly), the expertise of the pilots up front usually exceeded the expertise of the regulators by a large amount (kinda like IT today...). So, the Feds did something very intelligent - they came up with FAR 91.3 & 91.13:
91.3: (a) The pilot in command of an aircraft is directly responsible for, and is the final authority as to, the operation of that aircraft.
(b) In an in-flight emergency requiring immediate action, the pilot in command may deviate from any rule of this part to the extent required to meet that emergency.
(c) Each pilot in command who deviates from a rule under paragraph (b) of this section shall, upon the request of the Administrator, send a written report of that deviation to the Administrator.
This is the authority to do as the PIC sees fit. It is the entire basis of a Captain's authority - and it assumes that Captains know better than any one else how to command their craft. This has generally proven to be a good assumption - since the pilot is usually the first to arrive at the scene of an accident, he/she is highly motivated not to be in one... if you pick your pilots carefully. But to make sure that this authority is not abused...
91.13: (a) Aircraft operations for the purpose of air navigation. No person may operate an aircraft in a careless or reckless manner so as to endanger the life or property of another.
(b) Aircraft operations other than for the purpose of air navigation. No person may operate an aircraft, other than for the purpose of air navigation, on any part of the surface of an airport used by aircraft for air commerce (including areas used by those aircraft for receiving or discharging persons or cargo), in a careless or reckless manner so as to endanger the life or property of another.
The key words are 'Careless and Reckless Operation'. What it means is, if you fly in a careless or reckless manner, the FAA will nail you. Your actions, even if they don't specifically violate another reg, if they are 'careless' or 'reckless' - violate this one. How do you know what's careless & reckless? Well, if you have an accident, the presumption is you were careless, or reckless. These guys? They will DEFINITELY be hit for the 'careless' part. This is a catch-all - it means, at any time, your actions as a pilot can be held up to scrutiny, and if they are found to be 'careless or reckless', you are toast.
These laws were very good - it allowed pilots to experiment with how to run a plane, but held them responsible if something screwed up. Over time, best of breed techniques were discovered, spread from crew to crew, standardized by airlines, and ultimately adopted as Federal law - the FARs. Of course, accidents, once their root cause was discovered, frequently led to regulations as well. Overall, I think this was an excellent way to turn a fledgling, risky industry, into the mature and extremely safe one it currently is.
Times change. Aviation's period of discovery i
Techincally, they should also be monitoring Guard (121.50 MHz) the whole flight, and I'm pretty sure multiple attempts to contact them on Guard were made. But since they didn't bother paying much attention to the primary, they probably didn't montior Guard that well either.
There is really no excuse for both pilots completely losing situational awareness like this. They're both toast, and deserve to be.
As for the scheduling system they were going over - actually, that is probably the 'news for nerds' part. The old airline schedules were built in two units - 'pairings' and 'lines of time'. A pairing is a group of flights, typically from 1 to 6 days long, that begun and ended in a pilot domicile. The word 'pairing' was to indicate that an entire crew was 'paired' together that whole time. A line of time (or simply a line) was a month-long group of those pairings. There is a long list of legal requrements (min rest, max flight time, union contractual obligations, aircraft mx requirements, etc.) that these schedules had to meet.
Ultimately, from the pilot's point of view, these lines were published each month for the next month. Bidding was very straightforward. If you were the number 1 senior pilot in that base (technically, domicile, aircraft and status (capt. or F/O), you picked your line, and that was that. If you were #2, you picked your schedule, and got it.... unless the number 1 guy already got it, in which case you got your second choice. If you were number #300.... well, picking 300 schedules in the order you want them was a time consuming task, but the outcome was perfectly transparent. The line awards were public, so you could verify that the schedules you didn't get really did go to senior people. You can debate whether such a system is 'fair', but at least it is clear how it works, both globally and month to month.
Then, with the advent of more powerful computers, a system called 'PBS' was born - Preferential Bidding System. These systems, instead of having hard, published lines you bid from, instead only published the pairings. You expressed your 'Preferences' through a computer language. A computer program then ran, taking everybodys preferences, seniority, system constraints, etc. into account and generated schedules.
In theory, PBS sounds great. A pilot's preferences generally don't change that much month to month, so you could file your bid away and let it run automatically each month with little or no tweaking.
In practice, it's usually been highly disruptive and caused great angst for a year or two after being implemented, for many reasons:
1) The language used to express your preferences is generally designed for the programmers, not the users.
2) The results can be, to put it mildly, unexpected. When you have pre-published schedules, you have a pretty good idea ahead of time what to expect.
3) There are no month-to-month conflicts that generate additional days off, resulting in more work per pilot, a reason the airlines like them and pilots don't, on average.
4) Non-computer savvy older pilots (Captians) have a harder time getting it than younger pilots (F/O's), on average. It takes a vastly important piece of your life (when are you working? Where are you going? 28 hours in HNL or 32 hours in XNA?), and makes it tied to your comfort with learning, essentially, a primitive computer language.
I cringe when I see this, because I've done this - taught Captians while flying about PBS. So have many other F/O's. You just prioritize it where it belongs - below aviating, navigating and communicating. These guys made everyone else look bad.
A Freudian slip, perhaps? ;-)
I understand. I disagree with the death penalty as a matter of principle, but there are some people in the world who I would struggle to say don't deserve it.
...that hits all the same notes. E-books will take over the world, why are the German publishing houses sticking their heads in the sand, etc. I've thought about it quite a bit, since I have a strong personal preference for printed books, and have debated the topic with passionate advocates of e-books. I've come to a few conclusions:
1) The advantages that printed books have over e-books in terms of convenience will go away over the next 15 years. Limited resolution (200 ppi e-ink vs. 600+ ppi for print), limited battery life, bulk, storage capacity, etc., not to mention cost (not just direct, but transportation, storage, disposal, etc.), will all favor e-books in 15 years. Resolution (my particular nit) will probably take the longest, but it will happen.
2) I doubt a personal e-book 'reader' will last long in the marketplace. It's too big and bulky to be 'just' an e-book reader. Why not make it a web-browser? 95% percent of what you need to do that is there. E-mail? Terminal access? A cell phone with a bluetooth earbud? A movie watcher? It will become a general purpose computing device just like cell phones are becoming.
3) It won't succeed until an Apple-like company makes it so stunningly easy to use and manage that its advantages are clear. A cellphone and a smart cellphone are quite similar, so the idea of an iPhone/Treo (a general purpose computer that happens to be a cell phone) was not so hard to get accepted. A tablet-like device has no commonly existing parellel right now, and the existing examples are weak, to put it mildly. It will have to be wildly simple and pleasant to use...
4) Once most books are no longer printed, it remains an open question whether it will make censorship of ideas easier or harder. I haven't been able to come up with a convincing argument either way. DRM is also still an open question, although you can make a good argument that a DRMed device will fail in the marketplace. Maybe.
There will be a great e-book reader one day, but it won't be called that. It will be part of a package that can do far more.
...the U.S. had the greatest rise in its living standards. Scientists, engineers, and other professionals from all over the world migrated here in seach of a better life, the opportunity to live pretty much in peace and quiet, or simply to survive. It was seen as the most desireable place to live in the world, and that seemed to become a self-fulfilling prophecy as 'the best and the brightest' came here to do their best.
I wonder, are more folks returning to their home countries' simply because of money and career advancement? Or do they feel less welcome in the culture? Or perhaps their own home cultures are changing to where they feel they can shape them for the better?
This seems more like an anecdote than a study; but there is something wrong when science and engineering and other technical fields are seen as undesirable by most Americans, and the immigrants who come here to learn them decide that they'll have better opportunites back home to use them.
Perhaps a bit off-topic, but back in the 50's, my father worked for the National Enquirer as a photo retouch artist, with the air compressor/blower and all. His English was marginal back then, so many of the editing 'instructions' he got were quite confusing to him. Finally, his boss figured out how to explain it so he would understand 'the idea' behind the requested edits - "Make the boobs boobier, the flames flamier, and the blood bloodyer!" Suddenly, it all made sense.
He always said it was the greatest job he ever had after that!
You are assuming that all the books are recopied on a regular basis. NASA has had quite a bit of trouble retreving information from the magnetic tapes which stored the data from the Apollo missions, not so much because the tapes demagnetized, but because no records documenting the format of those tapes were located, and the designers of those formats weren't around anymore. Format loss is a signficant problem in the long term. In 50 years, do you think USB ports, FAT32 formatted flash-drives and UTF-8 will all still be in common use? Whenever I see someone use the words, 'never', ' instantly', and 'flawlessly' and other absolutes, I get suspicious of their argument. When it comes to preserving things for durations beyond a human lifespan, electronics are still relatively untested in practice.
But deep down inside, I know that e-books will probably win the day in the next decade or two; much like horses, physical books will probably become a niche item someday. Unless we screw them up royally with DRM or something. Amazing Xerox didn't appoint an armed guard by each photocopier they made to protect against possible copyright violations!
I'm in the same boat - my girlfriend and I have a combined library of around 2,000 books. There is a certain pleasure in browsing around, finding something you haven't read for a long time (or was from her library rather than mine)...
I realize that just about every argument I come up with for the superiority of printed books is pretty weak; the only four that are significant are 1) Books (good ones) can last 100's of years, 2) No power source needed, 3) Readability better (600 dpi+ vs. 160 dpi for typical e-book readers), 4) Better chance to survive civilization collapses.
Even these items are a bit of a stretch, and are likely to change within a decade. I suppose people with actual, physical libraries of books may be considered eccentrics fairly soon....
Actually, I'd be curious as to a reference for this. While I'm sure you are right once you start talking about 1000's of books, I'm equally sure the production of 1 paperback book is not 'extremely wasteful and bad for the environment' compared to one e-book reader. The key, of course, is how many books you read on your e-book reader before it, too, becomes e-waste.
A little googling revealed this Master's Thesis on exactly this topic. I haven't read it in-depth yet, but it looks to strongly favor e-readers.
Sigh - I LIKE my printed books.
This is a wonderful thing. It may make it much easier to publish new, low circulation books as well, since you don't need to reach a critical threshold sales number to make it worth printing. Of course, a 'book' (as in the physical form) may become obsolete over the next few decades as old curmudgeons like me who like reading printed material far more than reading off a screen drop off...
... (shakes head) ...
Sounds great. Let me know how that goes, okay? Tell me when the war is over, and teh stupid is vanquished. I'll hang a banner for you.
People with this attitude will have a hard time working with anyone outside of a very small group of very competent people; i.e. in the real world. Most people really aren't experts, they aren't always right, they frequently make mistakes.... but they are not evil. I try to reserve the word 'evil' for people who seek to hurt others for fun.
If all you have a choice of is 'competent' or 'nice', I suppose I would temporarily choose 'competent'. I would then seek to find a little more of BOTH in one human.
The airlines already implemented that plan. Sadly, nobody noticed.
Sigh - this is wrong. Many airlines recommend punching OFF the autopilot and autothrottles in severe turbulence because it will generally try to maintain altitiudes and courses when the best thing to do is minimize thrust and attitude changes to minimize loads on the airframe at the cost of altitude and course deviations. Similarly, when stuck in icing conditions, particularly in holding, many airlines recommend hand-flying to detect changes in flying qualities that may be occuring as a result of icing accumulation. The accident at Roselawn, IN lead to this procedure.
An autopilot is an aid to a human pilot. It has no decision making ability. It flies in certain conditions 'better' than a human, and 'worse' than a human in others. It is, ultimately, controlled by the human with an abstraction layer; instead of 'pull back stick to make nose go higher, which will probably result in a climb', it's 'select level change, set the level-off altitude to 15,000, and let the autopilot pitch and power to get me there'.
The operative word is 'do' - you do get a payscale reset if you switch airlines - all of your previous experience just qualifies you for the new job at entry level. One of my least favorite features about the airline industry. You are generally paid by the flight hour, and the rate per flight hour is on a table, where the row is your years of service or longevity AT YOUR COMPANY (typically topping out at 12 years), and the column is your equipment and position; 757 CA, MD80 FO, etc. You can see these tables at airlinepilotcentral.com. Keeping in mind that the typical airline pilot flies about 800 hours a year, maximum of 1000, and you get a good idea of pay. Corporate is quite different, from what I hear (your pay has much less to do with years of service and more to do with equipment type and position), but they have different issues relating to corporate flight departments being favorite targets for accountants, consumers, cost-cutters, etc.
Yeah.
Commutair - first year F/O, $18/hr * 1000 hrs. (absolute maximum a year) = $18k. CapeAir - first year F/O on the Cessna 402, $9/hr. That's $9k a year, if the bugger flies 1000 hours in that thing. Great Lakes - first year F/O, $16/hr -> $16k a year.
Take a look at a few others...
When I started at Continental Express in 97, flying ATR-42s and -72s, my first year pay was $13k. That doesn't include the $10k I paid for my own training - they had a great scheme going. If you fail out of training, well, you don't get the job, and you don't get the $10k back! If you pass, well, they just got a new, qualified pilot for 'free'. Also, see how far the princely salary of $30k/year goes when your domicile is SFO, EWR, or BOS. Here's the reality - you will be at poverty to lower-middle class wages for about a decade in a typical aviation career. It could be much shorter, if you get lucky. It could easily be much longer. It might be interspersed with periods of no pay, when you have been furloughed for some reason (e.g. 9-11, $140/barrel, a merger, bankruptcy, etc.) If you continue to be lucky and persistent, you can start making over $100k a year. It takes a long, long time to get there. So, frankly, spun has a point. I'm glad that your family of pilots has done quite well, but I'm quite sure you probably know a few pilots, and their families, who have not.
I'd like to talk to you a bit about this, seeing as I just spent all day avoiding doing just that, and only partially succeeding.
The problem is the following. What is a thunderstorm? Where is the edge of it? Laterally? Vertically? How do you detect these edges? How rapidly are they changing? How do you transmit all this data (assuming you can get it) to a pilot in a form that is useful and timely? This is actually an extremely difficult problem that is not solved. Whenever you hear a weather forecaster predict 'scattered thundershowers', he/she is admitting ignorance in determining exactly where the thunderstorms will and won't be, and instead provides a probabilty which correlates roughly to what percentage of an area will be hit by thunderstorms based on models in a given time period. I think you can see that's of little use in actually avoiding the real, specific thunderstorms that pop up, although it might be somewhat useful in planning - as in, "50% of Pennsylvania is going to be covered with wx, it's probably a good idea to carry extra gas to route around the entire area."
Thunderstorms have a life cycle; generally they have a growth stage, a mature stage, and a dissapating stage. While an airmass thunderstorm may go through all the stages in a hour or so, thunderstorms associated with a squall line can last in the mature stage for many hours. Each stage has very different characteristics, with different dangers. Severe turbulence that can rip an airplane apart, or simply put it out of control (after which the ripping apart may occur) is common; lightning strikes can knock out instruments and avionics; hail can physically damage the airplane; severe updrafts, downdrafts and downbursts can physically put the airplane at an altitude above it's coffin corner or into the ground... so let's agree, for the purposes of this discussion, that thunderstorms, in whatever phase, are dangerous. How do we detect them, so as to avoid them?
The best way, in terms of confidence and precision, is by eye. The towering cloud of condensing water vapor is incredibly distinctive, and gives a clear marker where not to go. Unfortunately, this only works if the visibility between the thunderstorms is excellent, which is frequently not the case. As a hot summer day wears on, the thunderstorms that have bloomed and dissipated have blown so much moisture high into the atmosphere that the visibility frequently is nil. Nighttime also does a great job of obscuring things. So, visual detection, while great when possible, is quite often not possible.
Onboard weather radar is the most common tool used in airline operations to avoid thunderstorms. The radar reflects off of (generally) rain and hail, and weakly off of snow and ice. Unfortunately, not all thunderstorms have much rain to detect, particularly in their growth stage! In addition, the radar is tilted at a specific angle above or below level, so really only sees a horizontal slice at a time; since the earth curves, and the altitudes of the thunderstorms curve with it, the height above ground of the horizontal slice varies with distance, making it exceedingly tricky at times to decide whether a radar echo is a threat or not. In addition, heavy rain is not the threat - turbulence, strong drafts, and lightning are. How does one decide whether the echos are part of a convective cell or a broader rain producing system? What if they are mixed together? In addition, at typical crusing altitudes, most of the rain visible to the radar is below you - but if you tilt the radar dish down to find it, you hit the surface much sooner, and lose the ability to detect cells further away...
Obviously, a real-time data link to ground doppler radar would be incredibly helpful. Unfortunately, it's not required by law, and it's expensive, so most airlines don't have it. Lovely. It probably wouldn't have done the Air France guys m
Dude, that wasn't competence, that was a tour de force of skill. There is no airline training for a twin engine flameout on departure, follwed by a deadstick landing into a crowded, narror river. It's not supposed to happen. It could have gone wrong in a thousand and one ways, from hitting the water at an angle and tumbling the aircraft, hitting a boat or barge, touching down with too much speed, chosing not to land in a freezing cold river but attempting to make a piece of pavement, recognizing both engines were toast and not wasting time in a fruitless attempt to restart them, etc. He had been handed a slender reed of luck - that the bird hits occured at several thousand feet MSL, it was good, daytime VFR, and he was close to the Hudson - if any of those conditions were not so (IFR, below 1000' when the engines bit it, etc.), he would have had no chance at all. He had no room to make any mistakes, and no do-overs. He did it all right, made all the right choices, the first time, and the only time, he had ever been in that situation, and not a single person died. If you aren't impressed by that, you haven't talked to your step-brother enough about what it's like.
In the airline world, if a student in a sim ride has passed his checkride/PC/AQP/whatever and there is extra time, the instructor will frequently do a Kobiashi Maru on the student just to see how they react - -60 knot windshear on departure, followed by a complete loss of all instruments but the peanut gages and then an engine failure, etc. They try to load them up with everything they can, partly for fun and partly to see how far the student can go. It almost always ends in a crash anyway, but that's part of the point, to never, ever give up making smart decisions while controlling the aircraft to the best of your ability. Sully got a real-life Kobiashi Maru and pulled it off.
You have it in reverse. Over the last 20 years, the percentage of airline crashes due to pilot error has been around 80% - planes have gotten far more mechanically reliable. Pilots have gotten better too, but to nowhere near the same degree as the machinery, sadly.
You don't have the slightest idea what you are talking about. Thunderstorms are such dynamic systems, and frankly our ability to gather data on them in a timely fashion is so limited, that there is not a single airplane flying in the world where computers decide what course, altitude, or speed to take to avoid one. So the option to 'trust your life' to one doesn't even exist, because their decision-making ability is so primitive. Computers are used to control airplanes, in the sense that a computer figures out the angle to pitch the elevator, how much fuel to meter to the engine, etc., in response to commands by the pilots, either through a 'direct' mechanism, such as the yoke and thrust levelrs, or through the autopilot, which is being directed by the pilots. Computers have almost nothing to do with the decision-making that occurs in a cockpit, and there is not even a system on the drawing board where they do, because no one has any realistic idea of how to make one.
Yes, performing the same way in a situation is great when all possible situations and the best way to get out of them are known ahead of time. That would have done zero good with the recent USAir incident in the Hudson. Unexpected situations crop up constantly in aviation - humans, while certainly not infallible, not even close, are still far better at dealing with unexpected situations than AI presently is. Failures are generally, by definition, unexpected, and have unexpected side effects. If you think AI is up to the decision making required in flying, why don't you join the AI team at MIT who is trying to develop fully computerized driving? Even though they only have two dimensions and a fixed set of paths called 'roads' to travel over, and the ability to pull over to the side of the road if the going gets rough, it's still rather difficult task.
Now you are either trolling or just stupid. More likely to crash than what? A computer AI pilot that doesn't even exist that would avoid the 80% of the crashes above, but not fall victim to a single additional crash which a human would have dodged? What are you smoking? You might have just as well said: "Human error was blamed for almost 80% of autocrashes in '04...why would you want to trust your life to something that statistics alone dictate to be more likely to crash?" - does this sentence make any sense to you since I don't know of any other viable drivers than humans? Oh, and FYI - going over thunderstorms is generally a bad idea, since they can grow at rates exceeding 6000 ft./min., far, far in excess of the climb rates of any commercial aircraft at the upper flight levels. As you fly higher and higher, the stall speed (effectively minimum flying speed) and the speed corresponding to maximum Mach (effectively maximum flying speed) converge, meaning you have far less ability to manuever or ride through turbulence. Perfect time to have a thunderstorm rise up from below and envelope you, eh?
Another factor to take into consideration is that not all airline pilots are experienced. I don't like to dichotomize (like the poor summary of the article, dammit KDawson) but a pilot's first storm could bring hardening experience or crushing defeat.
58-year-old flight captain Marc Dubois, who joined Air France in 1988, had approximately 11,000 flight hours, with 1,700 hours on an Airbus A330
If true, that's rather interesting. Most airline pilots fly around 800 hours or so a year. That would mean he started relatively late in the commercial flying business (mid-40s') or he had a significant timespan in his career where he was not flying. Anyone have more details on that?
Yes, because in the middle of the night at FL350 when you see nothing but black outside intermittently illuminated by blinding lightening apparently from all directions, where you are groping around your way with radar, which is painting red in all directions anyway, and getting ATC clearance for significant deviations off of airway centerline can take minutes (at 8 miles a minute, that's rather a lot) because of the limitations of LF, is just so EASY...how could those dumb pilots have done that?
Oceanic night flight through the ITCZ is still a bitch. Radar is still primitive. People are tired on long oceanic flights. The ability to change your route by significant amounts over the ocean without delcaring an emergency is limited. The aircraft is already operating near its maximum altitude for fuel effeciency, so any deviations from that altitude and course will eat into its fuel plans. A computer making decisions instead of a human changes none of these facts, and computer decision-making in fuzzy situations is notoriously poor, as most AI designers will tell you.
This is probably one of the most UNSAFE features of 'normal law' flight in the modern Airbuses. It essentially detrains the pilot. You learn the idea not to worry about control inputs, since you can, in theory, never overstress/stall/spin the airframe. It means that if and when a pilot winds up in 'alternate law' flight, where those protections are not present, all of his/her instincts about how much input is enought but not too much are gone.
I feel the same way about autobrakes for landing (not antiskid, not wheellock protection, not RTO protection, just autobrakes on landing) - they detrain you when you fly with them. Braking properly in a large aircraft is a skill, and judging by the slow drizzle of runway overrun accidents, a declining one. The best way to keep that skill fresh is to exercise it on each and every landing, rather than wait for the magic autobrake to kick in, then manually overriding it anyway for the runway turnoff. If we are going to rely on skilled humans to be in charge of aircraft, the systems of those aircraft have to be designed to some extent to help KEEP them skilled.
BTW, your comment after 'There are pros and cons to both. On the whole, I think that' got cut off, so perhaps you addressed my concerns already - my apologies if this is the case.
That's fine, so long as I can execute drivers who speed, run yellow lights or stop signs, don't yield to pedestrians in crosswalks, and who even think of drinking before driving. With 40,000 fatal car accidents a year in the U.S.A. alone, perhaps if we execute some of the 25,000,000 drivers who are issued citations a year (perhaps 1 in 10 like the Romans - that's only 2.5 million a year), we might reduce the accident rate! It doesn't matter if thier mistake(s) were intentional or not, since they result in so many deaths. /sarcasm
Air traffic controllers are generally highly skilled, responsible, underpaid professionals who do a thankless job cleaning up messes of overloaded and antiquated systems. If you believe extra-judicial executions are okay for them, guess what - you won't have an ATC system at all, because no one in their right mind would bother. In the real world, where serious adults are interested in systematic safety improvements instead of chest-thumping suggestions of violence, nearly all aircraft accidents have been decriminialized (the exceptions are made for 'careless or reckless operations' or willful violations of law), since nearly every professional in the aviation industry doesn't want to cause an accident, and is more likely to be truthful about all the circumstances that led to the accident if they are not in jeopardy. The results of these more candid examinations of accidents (look up ASAP programs if you want to learn more about them) have lead directly to multiple improvements in regulations, procedures, and the overall safetly level in aviation - in other words, they have saved lives.
Your casual suggestion that executing people is a good way to encourage others makes me wonder whether you are even serious.
Thanks for the interesting post - you sound like you have a keen appreciation of the politics there. So, let me ask you the question I've asked elsewhere:
How does a nuclear armed North Korea benefit China?
All the explanations I can come up with only make short-term sense. I don't think the Koreas' will remain seperate forever (or even for more than 50 years). A unified, nuclear armed Korea is probably not a good thing to have on your doorstep - why encourage it?
I DO think the North Korean government is pretty crazy, and once they get deliverable nuclear weapons, that gives them a big stick to use against anyone, including China! Allowing North Korea to develop nuclear weapons ultimately reduces China's control over them. Why would they do this?
North Korea and China are allies for strategic, historical reasons which are becoming less important. The basic relationship seems to be, "We give you arms, and food when you botch up your agriculture or have a drought -- you cause trouble/distractions for us when we ask". I wouldn't bet a lot on such a relationship enduring many decades.
The fact that North Korea would be a smoking hole within a half-hour doesn't seem to deter them much, but I think China might object to having Guangzhou or Hong Kong removed as well. That's the problem with nukes - their destinations are not hard-coded into them based on the political winds of the day they are done.
And, risking veering off into serious off-topic land: Why do you think Japan's culture is more compatible with the West than any other Asian country? I know it might be a long response...