Between 60 and 80 percent of all commercial airplane accidents are attributable to pilot error. NOVA looks at some shocking instances of pilot negligence and what airlines are doing to solve the problem.
Category
📺
TVTranscript
00:00Nobody is to land at the airport. We've had a crash.
00:12Most airline passengers take flying for granted, but sometimes something goes wrong. Our planes
00:19are getting safer, but the accident rate is no longer going down. The reasons may surprise
00:24you. Tonight on NOVA, why planes crash.
00:36Major funding for NOVA is provided by this station and other public television stations
00:40nationwide. Additional funding was provided by the Johnson & Johnson family of companies
00:47supplying health care products worldwide. And by Allied Signal, a technology leader
00:55in aerospace, electronics, automotive products, and engineered materials.
01:17I've been lucky in that I have not been involved in accidents or anything to cause me to be
01:25scared or feel unsafe. The natural concern for any airline passenger
01:31is mostly fearing mechanical failure, and I think that's where we worry the most.
01:37I worry more about, like, terrorist bombings and this type of thing than I do actual mechanical
01:42accidents. I rarely think of the pilot and the crew as
01:47being at fault. I feel like my odds are real good I won't
01:51be one of the people to crash. I think if it's my time, my time. If it's
01:57the pilot's time, it's my time also. Aviation is amazingly safe. Each year in
02:04the United States, several hundred million people fly. The chances of them being involved
02:09in an accident are minuscule. When we start talking about statistics in
02:14terms of the number of people killed in the airline industry in a given year, we're playing
02:17a very dastardly game, because what we're really saying is how many people can we kill
02:21a year and still say it's a safe enough system? Around 900 million people fly on airplanes
02:26worldwide every year. Almost all of them reach their destinations without incident. In an
02:31average year, only 800 people die in plane accidents. It's a safe system, but a growing
02:37number of industry observers are starting to ask tough questions about the margins of
02:41this safety. Author and airline pilot John Nance is one for whom the issue is not whether
02:46flying is safe, but whether it's as safe as it could be. In raw numbers, the fatality
02:52rate has been steadily dropping since 1950, and as a percentage of the total hours flown,
02:57there are fewer accidents now than there used to be. Since 1975, the overall rate of decline
03:03has leveled off. Now it is no longer improving, and U.S. analysts, including some airline
03:08pilots, are worried about forces which could actually set it back. Bill Raynard runs the
03:13ten-year-old National Aviation Safety Reporting System. The aviation system is probably in
03:23one of its more challenging periods right now, because you have essentially three things
03:28happening all at once. You've got the recovery from the air traffic controller strike. You've
03:33got the experience factor, because as the system accommodates new air traffic controllers
03:39and new pilots, you're going to have new people in the system, and you've got the deregulation
03:44environment. You've got more airlines. Safety is a very difficult thing to define. The lack
03:52of accidents does not in itself mean that something is safe. We have a potential that
03:58we're having to deal with here also. And I do believe that the margins are being reduced,
04:04that the operating philosophy of the airlines has changed based upon their increased competitive
04:14pressures. Early in 1982, a disaster occurred which suggested that all these influences
04:20had converged to strain the system to breaking point. Even more shocking to an industry normally
04:26preoccupied with nuts and bolts was the attention this accident focused on human beings. Air
04:32Florida was indeed a crucible of human factors. It was a quintessential example of what happens
04:37in an airline crash, a crash of any sort, when there's nothing wrong with the airplane. It's
04:42all a chain of human failures that led to that accident. January 13, 1982. At 4.01 on
04:50that snowy afternoon, Air Florida Flight 90 was at the bottom of the Washington, D.C.
04:54Potomac River. The rescue drama was about to be played out in real time on live television.
05:01The Boeing 737 had collided with the 14th Street Bridge shortly after takeoff from National
05:06Airport on its way to Fort Lauderdale. Four people lay dead on the bridge, and America
05:11watched in disbelief as helicopters hovered over the river searching for survivors.
05:25Only five people were pulled out of the water alive.
05:47Safety Board accident investigator Rudy Kapustin was on his way home when the news broke.
05:52Just as we got ready to go in the elevator, somebody came out and said, the airplane
05:55just hit the 14th Street Bridge. And it just didn't sink in for some reason to none of
06:02us that this was a big airplane. My response was, what the hell is a little airplane doing
06:06flying on a day like this? A major snowstorm had closed National Airport, and the Air Florida
06:12737 was held at the gate for two hours waiting for snow plows to clear the way to the ramp.
06:18At 3.58 p.m., the sound of laughter was heard in the cockpit as Flight 90 was being cleared
06:24for takeoff.
06:40Less than three minutes later, at 4.01, First Officer Roger Pettit tells Captain Larry Wheaton
06:45they are going down. One second later, the sound of impact ended the transmission.
07:01Seven days later, the cockpit voice recorder was dredged up from the river. Its grisly
07:05testimony helped explain the mystery of why the plane had been able to leave the runway
07:10but without enough clearance to miss the bridge.
07:13On the Air Florida cockpit voice recorder, we became aware that something just wasn't
07:18quite right with the engines.
07:20Paul Turner is the safety board laboratory technician responsible for analyzing the cockpit
07:25voice recorder.
07:26We all had this funny feeling about it not having enough power on it. So I brought the
07:31tape into the laboratory in the other room, and I ran an analysis on it, a spectrum analyzer
07:38on the tape, and was looking at the engine sounds. And it became apparent that when he
07:44ran his stand-up check with the throttles, and he pushed the throttles forward, and he
07:49checked his engines out, and then we should have seen an increase in the engine power
07:54and the engine should have gone up. And they didn't. They just barely rose, if at all.
07:58And then they were stabilized at that particular EPR setting, which is how you set this engine
08:03pressure ratio.
08:04The next test confirmed the idea that the crew had tried to take off with only three
08:08quarters of the necessary engine power. Boeing experts later confirmed Turner's hunch that
08:14this was due to an ice-blocked pressure duct. This, in turn, had created a misleadingly
08:18high indication of the engine pressure ratio. The first officer could see from the other
08:23gauges that something was wrong, but the captain paid no attention.
08:27Determining why the duct was blocked raised the question of whether the engine de-ice
08:31was on.
08:32We went through the portion of the checklist in which the captain reads, or the captain
08:39replies, to the co-pilot as to whether the de-icing system is on or off. He said, anti-ice,
08:46the answer would be on or off. And those two words are sometimes difficult to differentiate
08:53in the cockpit with all the other noises on. And the group was actually divided in this
08:58particular case.
08:59Clearly, the response was off. And because everybody that listened to it, including the
09:06people that had no reason to be partial towards the investigation, our own people couldn't
09:14believe it. And of course, the Air Florida people didn't want to believe it. The pilot
09:18group didn't want to believe it, that the response was off.
09:23Failure to turn on the engine de-ice in the middle of a freezing snowstorm was only one
09:27of the cold weather protocols flouted by this crew. They also ignored other de-icing
09:32procedures and knowingly took off with snow and ice on their wings.
09:38Patricia Felch, one of only five survivors, later recalled her concern during testimony
09:43at the accident hearings.
09:45I remember the last de-icing was about 3.15. And I remember we took off about between five
09:54and four o'clock. I turned to my boss and said, they're going to have to de-ice us again.
10:00And when they came back to the right side, I turned and said, see? And he just laughed.
10:06But they never came back to the left side of the plane.
10:09Seven months later, the National Transportation Safety Board report implicated the controllers
10:14for the long delay between de-icing and departure clearance and the plane's design for a winter
10:20performance defect. But chief blame for the accident was placed on the Air Florida crew
10:25for three different counts of faulty judgment before and during takeoff.
10:30There were a lot of people at the time of the investigation and during the formulation
10:34of the report that did say, you know, this is probably our first deregulation accident.
10:39But Air Florida had a pretty good screening process. With this particular pilot, his captain
10:48was picked up from another airline, Air Sunshine. And he kind of fell through the crack. He
10:54didn't go through the screening process.
10:56People have said that the crew was inexperienced or that the crew didn't know what they were
10:59doing. There are several ideas along this line. The fact that they flew out of Florida
11:04and were not used to icing. However, they flew into the north. And the co-pilot particularly
11:09was an old fighter pilot who had plenty of experience with icing. And so you would have
11:18to say in this particular case that they were almost led into this thing by the circumstances.
11:23The first officer was described by his friends as jovial, witty, a good guy. We...
11:34He was considered a fairly sharp pilot, though. Skills-wise, he was fairly sharp in knowledge
11:41on the equipment.
11:42Nobody wants to believe that pilots make mistakes, including pilots themselves. But they do.
11:48For all its complexity, Air Florida is not an unusual accident. Between 60 and 80 percent
11:54of all plane crashes are blamed on cockpit error. Certain U.S. airlines, like Continental,
12:00now make their crews attend classes designed to deal with this problem. In this seminar,
12:05instructor Frank Tullo reviews the Air Florida case, emphasizing the breakdown in communication
12:11between the first officer and his captain.
12:13And what I want to make sure you understand is that any act of assertiveness here would
12:17have prevented the accident. And if the co-pilot had just stopped nurturing this captain and
12:22said, we got to do something. We have to either push the power up or abort. Because he obviously
12:26saw something. Five separate times he made it a point.
12:30Ordinary human failings, not paying attention, being afraid to speak up, disrupt the teamwork
12:36so crucial to safe flying. Aviation today is more than handling a machine. It's handling
12:41people. And the accident rate suggests that pilots are ill-equipped for this management
12:46role.
12:48A study of all plane accidents occurring between 1959 and 1982 shows crew error as the most
12:55frequent cause, followed by the aircraft, weather, airport management or air traffic
13:00control, miscellaneous, and in sixth place, maintenance. Increasingly, the planes themselves
13:07are not the chief cause for concern. Thirty years ago, the likelihood the equipment would
13:12fail was ten times higher than it is now. Since the first turbojets of the 50s, each
13:18new generation of aircraft has become more automated, more reliable, and safer to fly.
13:24But if the planes have improved, the same cannot be said of the pilots. Captain Mel
13:29Volz is safety consultant and former head of training at United Airlines.
13:34The manufacturer has done such an outstanding job of providing us with just a remarkable
13:39product. The airplanes today are so far ahead of what they were only a few years ago. They're
13:43strong, they're reliable, the engines don't quit, they just run magnificently. If we've
13:50done so well in attacking the mechanical side of the equation, that of course has left
13:54the human side over here, and that's the side we have to attack now.
13:58Dealing with the human side has meant taking a close look at how pilots acquired the attitudes
14:03and cockpit behavior that is getting them into trouble. Part of the answer lies in the
14:07way many of them underwent their original training.
14:10Now remember, I want you to look around, keep your speed up, and if you do get a bounce,
14:17cut him off and drive him in range. When you get in range, shoot, and when you shoot,
14:22shoot to kill. Okay, let's go again.
14:31Korea was just one of the proving grounds for many of today's pilots. Airlines have
14:36traditionally recruited military airmen, those brave mavericks who excelled in the single-seater,
14:42high-performance fighter aircraft. By the time author Tom Wolfe had immortalized the
14:47combat flying ace in his book, The Right Stuff, these men had found their way onto
14:52most of the flight decks of America's passenger airlines.
14:58That macho self-sufficiency that doesn't admit to difficulties
15:02is now dangerously out of place in the team-oriented peacetime cockpit.
15:13At NASA's Aerospace Human Factors Division, psychologist Dr. Clay Foucher is studying
15:18this legacy and the problems it has caused.
15:21In the old days, if you're familiar with Tom Wolfe, Tom Wolfe's treatment, The Right Stuff,
15:28pilots tended to be very individualistic, the types of individuals who weren't used
15:34to relying on other people to get something done. And I think a lot of that was bred from
15:39the single-seat, high-performance fighter aircraft school.
15:43Flying an airplane is something I've wanted to do since I was six years old.
15:47The real reason why I like flying, I'm not sure I can tell you,
15:50but I have to tell you that flying an airplane is the best job in the world.
15:55I've been flying for 30 years, and I wouldn't consider doing anything else.
15:59It's a very dynamic business. It's filled with a great deal of challenge and a tremendous amount
16:03of reward.
16:05Where else could you get this kind of experience?
16:07Where else could you get this kind of satisfaction?
16:11I take this gigantic airplane and fly it through the sky, and it's my sky, and it's my airplane.
16:20The responsibility of passengers is not overwhelming.
16:24The idea is that you take excellent care of yourself, and the passengers will be just great.
16:29Most pilots that I know are confident in their own ability to solve problems.
16:35They feel that they are ultimate professionals, that they know the job.
16:40They know the resources.
16:42It's a feeling of exhilaration.
16:44It's a feeling of accomplishment, and it's a feeling of mastery.
16:50We're not completely sure what today's right stuff is.
16:52Although, if I were to venture a guess, I would say that you would see a lot more emphasis on
16:59both instrumentality or concern with the task.
17:02That's clearly important, since technical skills will always be important in this job.
17:06But you might also see an emphasis on expressivity or the ability to relate well to other people.
17:15If stick and rudder skills are no longer enough, how will pilots with their technical orientation
17:20acquire the necessary new abilities?
17:23In search of a solution, and after two bad accidents, United was the first of a handful
17:28of U.S. carriers to embrace an idea called cockpit resource management, or CRM for short.
17:36At their Denver facility, captains, first officers, and flight engineers all undergo
17:41three days of annual recurrency training.
17:44In the United program, concern for personal relationships is valued as highly as technical
17:50competence.
17:51Using ideas borrowed directly from business management, crews are taught to recognize
17:55how their individual styles can create the communication blocks which cause accidents.
18:01I think so, too.
18:03So we would move this guy.
18:05Rob, would you have any problem with that, moving him up this way?
18:08Cockpit resource management has two aims.
18:11One is to a degree modify behavior.
18:13It's not sensitivity training, and we would not get into that.
18:17And the other is to also get the crew to work together as a crew in all circumstances.
18:23There are a number of techniques that are utilized in CRM training, but one of the most,
18:27I think, useful techniques is where pilots fly a full mission simulation, and that full
18:31mission simulation is videotaped.
18:34And they sit down with an instructor afterwards and go through exactly what occurred in the
18:39simulation.
18:39They see themselves very graphically.
18:42These cockpit simulators so closely imitate the real thing that pilots can be exposed
18:47to hazards far too dangerous for an actual training flight.
18:518 knots off.
18:52Interconnect closed.
18:53System depressurized.
18:56Instructor Tom Branch observes how well the team handles pre-programmed emergencies.
19:01This crew is about to lose electrical power, the kind of classic problem which could divert
19:06attention from flying the plane.
19:08In the debriefing to follow, the crew will be evaluated not only on how well they handled
19:13the technical challenges, but whether they did so in a coordinated effort.
19:17Start number three.
19:18Start number three.
19:19Number three.
19:20Start number three.
19:21Good.
19:24Well, we've lost the cabin now.
19:27Close.
19:30Rog.
19:31It says the outflow valve is fully closed.
19:33Disestablish communication.
19:36Video replay requires the individuals to confront themselves as team players.
19:41This exposure is sensitive, and the airline protects the pilots by erasing the tapes right
19:46after the postmortem.
19:48We had expected privileged access, but a computer malfunction made the flight problems more
19:53serious than necessary, causing, amongst other things, a dangerous loss of air pressure.
19:58The crew had performed well, but by their standards, not well enough to share much of
20:02their analysis with us.
20:04Usually, privacy ensures a more revealing discussion.
20:10Let's look and see what happened to us.
20:12You're at 35,000 feet, and you lost all your generators.
20:16And first of all, Frank, what was your problem?
20:19What was different back there than you expected to see?
20:22Well, one of the things you see in response to those situations are pilots turning around
20:29to the other crew members and saying, my God, do I come across that way?
20:33Whereas the other crew members will turn and say, yeah, you can be extremely overbearing
20:38in some cases.
20:39That type of feedback, video feedback, allows a person to see him or herself in a way that
20:46they're not used to doing.
20:49This need to make the crews aware of the impact of their behavior is underscored by the cockpit
20:54voice recorders recovered from accidents.
20:57Conversations amongst the crew betray patterns which are being acted out over and over again.
21:03One of the most straightforward examples is the tragedy of Eastern Flight 401, so clearly
21:09uncomplicated by any environmental factors that it has become a classic case study.
21:16On a dark December night in 1972, three able-bodied and experienced pilots flew an L-1011 into
21:23the Florida Everglades after becoming distracted by a burnt out light bulb.
21:28It sounds implausible, but crew preoccupation with minor mechanical hitches and an associated
21:34failure to monitor the instruments is a common cause of accidents.
21:38This disaster also illustrates some other recurring problems, like the failure of captains
21:43to act as leaders, make decisions, set priorities, and delegate responsibilities, shortcomings
21:48which are compounded by unassertive and complacent crew members.
21:54A simulator reenactment of the final eight minutes of this flight shows how all these
21:59ordinary human failings led to a complete breakdown in teamwork and ultimately to the
22:03deaths of 99 people.
22:08Miami Tower, you're in Eastern 401, just turned on final.
22:11Eastern 401 heavy, continue approach to 9 left.
22:15Continue approach, roger.
22:18Actual airline pilots play the parts of the ill-fated crew.
22:21The dialogue is taken from the cockpit voice recorder and altered only to remove or change
22:26expletives and names.
22:28Certain omissions have been made to shorten the elapsed time.
22:31Now I'm going to try it down one more time.
22:33Eastern flight 401 is on final approach to Miami International, runway 9 left.
22:39The nose landing gear indicator has failed to illuminate, so the crew cannot tell whether
22:43the gear is extended and locked.
22:45You want me to test the lights or not?
22:47Yeah, check it.
22:48Put your seat back.
22:52Dougie, it could be the light.
22:54Could you jiggle the light?
22:55It's got to come out a little and then snap in.
22:59I'll put them on, up to 2,000.
23:04You want me to fly, Doug?
23:05What frequency did he want us on?
23:0728.6.
23:08I'll talk to him.
23:10All right, approach control, Eastern 401.
23:12We're right over the airport here and climbing to 2,000 feet.
23:15In fact, we've just reached 2,000 feet and we've got to get a green light on our nose
23:19gear.
23:20Eastern 401, roger.
23:22Turn left heading 360.
23:24Maintain 2,000.
23:25Vectors to 9 left final.
23:27Left to 360.
23:29I think it's above the red one.
23:31Yeah, I can't get it from here.
23:33I can't make it pull out either.
23:34You got pressure?
23:35Yes, sir.
23:36All systems.
23:37Put the damn thing on autopilot.
23:39All right.
23:40See if you can put that light out.
23:43No, you got to push this.
23:44Swish this a little bit further forward.
23:47Now turn it to the right a little bit.
23:49No, I don't think it's going to fit.
23:53Hey, get under and see if that damn nose wheel's down.
23:56OK.
23:57You got a handkerchief or something so I can get a little better grip on this?
23:59Anything I can do with?
24:00This damn thing just won't come out, Doug.
24:02If I had a pair of pliers, I could cushion it with that.
24:04The captain has neglected to divide up flying responsibilities.
24:08Everyone is absorbed by the crisis, so they don't hear the audio alert announcing a change
24:12in altitude.
24:13Hey, I'll give you the pliers.
24:14To help with this, go down and see if that red line is lined up down there.
24:18Don't screw around with that $0.20 piece of light equipment.
24:23Eastern 401, I'll go out west just a little further if we can here and see if we can get
24:27this light to come on.
24:29All right, Doug.
24:30The autopilot has somehow become disengaged.
24:33The plane is slowly descending, and nobody is paying any attention to the altimeter.
24:38It's always something we could have made scheduled.
24:40We can tell if the damn gear is down by looking down at the indices.
24:44An emergency landing with a possible nose gear problem is neither very risky nor all
24:49that unusual.
24:50It's an option the captain could be preparing for now.
24:54It's got to be a faulty light.
24:56Doug, this damn thing just won't come out.
24:58All right, just leave it there.
24:59Eastern 401, how are things coming along out there?
25:03The controller's inquiry is too vague for the crew to realize he's asking about 401's
25:08surprisingly low altitude.
25:10OK, 180.
25:19Hey, we did some of the altitude here.
25:21What?
25:22We're still at 2,000, right?
25:23Hey, what's happening here?
25:29Look at the time between when they first noticed that there was something wrong with
25:37the altitude and when they hit the ground.
25:40In that eight seconds, those two pilots, both of them, were so removed from their primary
25:46job of flying the airplane that they couldn't do a thing.
25:50They didn't reach up and grab the yoke and change the pitch or the roll of that aircraft
25:54one degree.
25:55That's how removed they were from their primary job of flying the airplane.
26:00Bill?
26:01We're sort of ingrained in what we've been taught over the years.
26:05And it's like the old story about the job description.
26:11Hours of sheer boredom with moments of stark terror.
26:15Just what he was saying before.
26:17We're focused on one item at a time.
26:23What you're saying is they're very task-oriented.
26:25We agree wholeheartedly.
26:26Pilots are task-oriented.
26:28They're used to having—as a matter of fact, sometimes you'll see a pilot who's actually
26:32excited about the fact that he's got a problem.
26:34Good.
26:34Let's get in here and solve this problem.
26:36I would like to prove to the world that I can solve this problem.
26:39Frank, you know, it's very difficult for a pilot, when he starts working a problem,
26:43to leave that problem before it's resolved.
26:46Well, how do we prevent that?
26:47How do we prevent that from happening in a cockpit, in your cockpit?
26:50How do you prevent that?
26:52Most of our pilots, prior to coming in, have a very low opinion of what we're going to
26:56do.
26:56They call it charm school.
26:58But as they leave the class, I usually get a healthy handshake and a thank you for a
27:02class they felt was very, very worth their while.
27:06We push sticky issues.
27:07We want the pilots to talk about the problems they've had in cockpits.
27:11It appears that just talking about them helps the situation.
27:15Our thinking is still back too many years ago, where the captain was the guard.
27:20He was Peter Pan and Mary Poppins all at the same time.
27:24And the first officers didn't have the input.
27:28The first officers weren't allowed to have the input.
27:31I'm flying with captains right now that, unless they get to a very dangerous point,
27:38I'm not going to say a whole lot.
27:40Because they are really going to jump.
27:42I know of one right now that has been reported.
27:45There's been nothing done because he's a senior guy.
27:48He's a check airman, but he flies good.
27:52But boy, you can't tell him anything.
27:54I think if we've learned anything over the past two days is the fact that us as first
27:59officers, that we can speak up and question anything that we don't agree with.
28:05It's my responsibility as a first officer to say something and have the guts that if
28:15I get called in, we'll take it from there.
28:18While pilots themselves may find this training valuable, the impact of CRM as an accident
28:22deterrent is hard to prove.
28:25Most airlines have yet to climb on the bandwagon.
28:27And it's not clear that ingrained, wrong-stuffed personalities are even amenable to change.
28:34Well, changing human behavior is a very difficult endeavor.
28:39And clearly, it's not something that you can bring about within a couple of days within
28:43the classroom.
28:44The program has to contain some recurrent training.
28:47Now a member of the National Transportation Safety Board, psychologist Dr. John Lauber
28:52pioneered much of NASA's ongoing research into cockpit behavior.
28:57Does he believe that CRM training should be mandatory?
29:00If it were an idea and a concept that was fully developed, that we knew what techniques
29:07worked and worked effectively, and we knew unequivocally that it would be necessary to
29:13and we knew unequivocally how effective they all were, then it might be a good idea to
29:19consider making the concept mandatory.
29:22From my point of view, and I think others agree, it's still a concept that's in
29:26development.
29:27The airlines themselves are divided in their enthusiasm for CRM.
29:32Pilots from 46 air carriers belong to the Airline Pilots Association here in a Washington,
29:37D.C. hotel for their annual safety forum.
29:40But only two members, Pan Am and United, and two other airlines, Continental and People
29:45Express, run full-fledged CRM programs.
29:48Five others have partial training.
29:50Even among the majors, there are several who don't teach it at all.
29:54Captain Joe Oliver on the left is a 28-year veteran of Delta Airlines.
29:59Our philosophy is directed toward a captain-oriented flight crew.
30:05In such a philosophy, he is the director, the leader of the band, and by the time that
30:12the individual reaches captain status, he is either able to manage the resources that
30:20are his, or the four days or two days of training that you give him aren't going to
30:25materially change that.
30:27Charlie Finch is a Delta First Officer.
30:29The captain's always the final authority.
30:32You may be flying the aircraft, but he'll always be the one that maybe say, let's put
30:37a little bit more power on or carry it an extra five knots of airspeed.
30:40He reinforces your decisions in most instances.
30:42We have tremendous communication with Delta because everybody loves their job.
30:48And that's really important.
30:50But Delta doesn't have any training per se, not that I know of.
30:55I mean, I've never been told how to get along with the captain.
30:57You just assume that you get along with this person and make his job easier.
31:01Assertiveness training on Eastern Airlines as a standalone unit, as a specific objective,
31:07we've not felt the need for it.
31:10Most carriers have a corporate culture of their own, and each carrier is a little different
31:15one from the other.
31:17At Eastern Airlines, we have never had, to the best of my knowledge, any need to teach
31:22our pilots how to be assertive as such.
31:25Captain Burton Beach is in charge of training close to 5,000 pilots at Eastern Airlines.
31:30There is more evidence of the failure of the crew or the captain to effectively manage
31:38the crew than there is positive evidence, in my opinion, that it's a good thing.
31:44You can always point to accidents and incidents that have happened in the industry and say,
31:48well, if they had managed what they had a little better, that would not have happened.
31:52Accidents are terrible research criteria for a couple of reasons.
31:56First of all, if you saw a reduction in accidents, it wouldn't mean very much because accidents
32:01are so infrequent.
32:02If you saw an increase in accidents, you could argue it the same way.
32:05So in short, accidents really make lousy research criteria.
32:09If we haven't had an accident for five years, can we presume because we didn't have one
32:14for five years after the program went in, it was as a result of the program, would we
32:18have had another five years in the first place?
32:20Did we spend $5 million for nothing?
32:23Airlines may be skeptical about cockpit resource management.
32:27But given the fact of human error, is there an alternative solution?
32:31At the Boeing Aircraft Company, engineers are trying to enhance safety through improved
32:35design.
32:36Perhaps automation can compensate for erratic behavior.
32:40Maybe the answer to pilot error lies in doing away with the pilot.
32:45It's interesting to consider removing the pilot from the airplane.
32:49However, it's certainly not practical in the foreseeable future because the pilot's skill
32:56in cognitive decision making just can't be duplicated.
33:01Del Fadden is manager of flight deck integration at Boeing in Seattle, Washington.
33:06The cognitive process that the pilot goes through is extremely complicated, much of
33:12which is not consistent from person to person.
33:15And so we don't understand totally how everyone would do the cognitive task.
33:20The physical tasks are much more straightforward.
33:22They're much easier to replicate with the machine.
33:26The autopilot's the most obvious piece of automation in the airplane.
33:31It allows the pilot to be freed from the continuous flying task for major or minor portions of
33:38the flight.
33:39And that has been improving over the last 20 or 25 years.
33:43The original autopilots were able to maintain the airplane in straight and level flight,
33:47but could do little else.
33:50Subsequent autoflight systems could make approaches but couldn't land.
33:54Autopilots on 757 and 767 and some of the other contemporary airplanes can do very complex
34:02flight maneuvers in flight and do complete automatic landings.
34:06So the automatic ability there has been improved steadily with each airplane.
34:12The precision of autopilot saves fuel and is useful in crowded air corridors.
34:17It frees up the crew to scan for avoiding other planes and cuts fatigue on long flights.
34:22But the autopilot does little to relieve the pilot's invisible workload.
34:31The 757 and its precursor, the 767, ushered in the first generation of Boeing planes to
34:37tackle this problem of mental workload.
34:40The heart of the innovation is a computer which augments decision making, especially
34:44during times when pilots are liable to become overloaded and preoccupied.
34:49The engine, hydraulic, and electrical systems are now under computer surveillance.
34:54All the data is consolidated onto two screens.
34:57For the first time, the crew has the option of calling up what they want to see and when
35:01they want to see it.
35:02The information presented is very accessible and more easily read than scanning an array
35:07of instruments spread throughout the cockpit.
35:09Here the computer is flagging an engine fire, identifying which engine, and recommending
35:14a shutdown.
35:15Additional help can be had on demand.
35:18On the screen now are instructions for restarting the engine in mid-flight.
35:23The information advance here allows us to put information together in ways that the
35:28pilot could do previously but had to do himself.
35:32He can concentrate on the decisions, what to do with the airplane, what to do with the
35:36relationships that he can see in the display, rather than how do these two pieces of information
35:42that are physically separated relate.
35:45The system saves mental effort by assembling the information needed for a decision.
35:50But is this so-called glass cockpit a radical safety breakthrough or just another resource
35:55which the crew must manage?
35:58Earl Wiener is professor of human engineering at the University of Miami.
36:02In collaboration with several airlines, including Eastern, he is observing the impact
36:07of the 757 on pilots who are making the transition from less automated planes.
36:14Okay, we're going to St. Thomas.
36:15Is that it?
36:17Yeah, have you been down there before?
36:18Yeah, I've been down once on this trip.
36:20St. Thomas, St. Kitt.
36:23Yeah, I was kind of hoping we'd start flying that from New York instead of just Miami because
36:26it's the perfect airplane for that route.
36:29One of the problems with any automatic device, and certainly the highly developed ones found
36:34in the cockpit of a 757, is that they are what I call the two D's, dumb and dutiful.
36:40Dumb in the sense that if you set up a condition which is incorrect or possibly even hazardous,
36:48equipment won't necessarily recognize it.
36:50Dutiful in the sense that it will carry out whatever it's programmed to do.
36:54In September 1983, Korean Airlines Flight 007 was shot down when it strayed into Soviet
37:01airspace.
37:02How it got there is still a matter of conjecture, but Earl Wiener puts it down to a typing error.
37:07I think that either the wrong heading was programmed into the computer at anchorage
37:12before takeoff, or the wrong initial position at anchorage, or perhaps it was put into a
37:18Perhaps it was put into a heading mode in flight and just stayed on that heading rather
37:24than following the pre-programmed navigation points.
37:27I do not believe that it was a conspiracy on anybody's part at all.
37:30It was a simple navigational error.
37:35When you go from manual flight to automatic flight, a new set of problems and conditions
37:41arise.
37:42One of the problems is the need to monitor, which humans don't do very well.
37:48Automation needs constant supervision, but pilots are alienated by this machine-minding
37:53role.
37:54They find it difficult to stay vigilant.
37:56The danger of this was dramatized by a 1985 Air China incident.
38:02Dynasty Flight 06 was at 41,000 feet over the Pacific when one of its four engines failed.
38:08The crew was not paying attention and didn't react in time.
38:12When the autopilot could no longer hold the nose straight, the plane started to yaw toward
38:16its dead engine.
38:18By the time the crew regained control, it had already fallen six miles.
38:22Passengers were injured and the plane damaged in an incident resulting directly from automation
38:27complacency.
38:32The most difficult part is the monitoring job or the watchkeeping job.
38:36And when pilots speak of complacency, that's what they have in mind.
38:39The tendency sometimes to set up the equipment and program it and then not keep track of
38:46it, not watch where it's going.
38:47And that, I think, is a critical problem in the mind's eye of the pilot.
38:52This is not the only source of concern for pilots who are trying to adjust to automation.
38:57Some of the more negative things that one hears in interviews is a sense of powerlessness,
39:04a loss of control.
39:05They talk about being along for the ride.
39:07They refer to the new aircraft cockpits as Ataris and Pacmans.
39:12And some of the more specific complaints would be the problem of complacency, that they recognize
39:19that tendency to be complacent, a concern about loss of skills.
39:24And that's a very realistic concern, that is, if they are spending most of their time
39:29flying along under automatic control, now do they have the skills to take over manually
39:34and land the plane should the automation fail?
39:37The heart of the problem of flying into an airport like St. Thomas, where the physical
39:44characteristics of the airport create a situation that is trying, you can't do much with automation.
39:53The wizards with the drawing board would like us all to believe that flying is an applied
39:57technology.
39:58But flying is still an art form.
40:00And automation is not going to do some of the things that have to be done.
40:06Pilots are error-prone but irreplaceable, and even at their best, entirely at the mercy
40:11of a system over which they have no influence.
40:14This system is currently imposing some severe pressures from many different points of view.
40:19Yeah, we've got some problems going on in the industry right now based on lack of management
40:24understanding of how close you can come to the line and not decrease safety.
40:30When you start putting pressure on pilots to fly with fewer and fewer amounts or lesser
40:36amounts of fuel, you start cutting the margin into an area that's subjective.
40:41In other words, you may be safe, you may never get close to crashing an airplane, but what
40:45you are doing is reducing the margin of safety.
40:47I personally was involved in an incident which was frightening, and there have been several.
40:53It involved a takeoff situation where an airplane was cleared for a landing at the same time
40:58that the airplane I was in was on takeoff.
41:01Well, the reports seem to indicate that we're seeing more runway transgressions.
41:04That's more than one aircraft or more than one vehicle on a runway when there should
41:08only be one there.
41:09We're seeing the automation issue, certainly because we're getting more automated aircraft.
41:14There's a lot of other new airlines that are coming in that aren't as quick on the
41:17radios, that aren't as confident or experienced in the approaches that we've used over and
41:23over, and you look out a little bit more than you used to.
41:27Although there are very strong pressures about keeping integrity with the schedule
41:34and accepting airplanes that have mechanical deficiencies and flying in weather that you
41:40certainly don't want to fly in, but it's not illegal for you to fly in.
41:583,000, Mike.
41:583,000.
41:593,000.
42:00Maintain 3,000 until the localizer cleared at ILS runway 17L approach.
42:04What's going on out there?
42:05Y-85, left of 1-9-0-12, ILS 17.
42:09What?
42:10Nobody is to land at the airport.
42:11We've had a crash.
42:13Everybody go around.
42:17Cockpit resource management theory meets reality out on the line, where crews must
42:22operate under all kinds of pressure and make judgments based on insufficient and conflicting
42:27information.
42:28Nowhere was this made clearer than in a disaster at Dallas-Fort Worth Airport on August 2,
42:331985, when treacherous winds forced Delta Flight 191 onto the ground a mile short of
42:39the runway.
42:41Wind shear has caused 18 major U.S. accidents since 1970, and in most cases, the pilots
42:47were held responsible.
42:49This one killed 173 people and involved a very experienced crew flying for an established
42:54airline.
42:55It's the kind of accident that could well happen again.
42:59How can pilots avoid danger which is difficult to detect, and can they be held responsible
43:04for the outcome of an entire chain of human errors?
43:09Jerry Chandler is an aviation journalist who happened to be at the airport at the time
43:13of the crash.
43:14He has since published a book about the safety issues it raises.
43:17On this road, Highway 114, we're heading west toward Fort Worth.
43:21Delta 191 first touched the ground off to the right, dug a trench about 260, 270 feet
43:31long, hurtled this lane airborne.
43:37Its left engine, the port engine, hit a Toyota driven by a mechanic from Mississippi coming
43:43out here for a job.
43:44That slewed the airplane about 10 degrees to the left, and that put it directly in line
43:49with the water towers.
43:51The tanks right over here to the left are the tanks which the airplane finally impacted.
43:58The nose hit the southernmost tank, pivoted the craft around, and actually snapped off
44:04the tail like a bullwhip.
44:06As it was, that tail was like an escape capsule for the 26 people who did survive.
44:11We had a water tower.
44:14I think it was a wind shear.
44:15Did you see any lightning?
44:17No.
44:20Delta 191 had flown into a type of wind shear called a microburst.
44:25A thunderstorm produces a funnel of rapidly downward moving air, which slams into the
44:30ground, causing violent wind eddies and swirling horizontal tornadoes.
44:36The plane was approaching to land when the first officer saw lightning.
44:40A minute later, violent tailwinds and driving rain dropped airspeed and altitude, but 191
44:45recovered.
44:46Another tailwind cut airspeed.
44:48A downdraft forced the nose down.
44:50It continued to descend into a tailwind, which slammed the plane onto the road at 250 miles
44:56per hour.
44:58Dr. John McCarthy is investigating wind shear for the National Center for Atmospheric Research.
45:03Wind shear alert.
45:04Center field wind 300 at niner south quadrant wind 170 at 1-0.
45:08Well, from what we know of the wind shear events, particularly microbursts that we've
45:12studied, approximately 25 percent of the microbursts we know about are no big deal.
45:18Anybody can fly through them.
45:21There's probably 25 percent on the other end of the distribution, which is probably
45:25fatal.
45:26If you get into it, you're not likely to get out of it.
45:29Is this what happened to the crew of Delta 191?
45:34I'm acting chairman of the National Transportation Safety Board.
45:37We are here today to consider the aircraft accident report of the Delta.
45:42The wind shear that caused this accident was the most violent ever recorded.
45:46Should they have been able to escape?
45:47And why did they fly into it in the first place?
45:50These were the issues on which the accident would be judged at the safety board hearing
45:54a year later.
45:55Did the crew know the severity of the weather they were flying into?
45:58The controllers had informed them of the rain shower and the fact that the winds were
46:04variable at the field.
46:05That was the only additional information the controllers had provided the flight crew,
46:10which effectively was all that they were in possession of.
46:13The forecast was only for variable winds and rain showers, so the airport was not expecting
46:17wind shear conditions or thunderstorms that afternoon.
46:21This lack of accurate information was linked with the operations of two different weather
46:25stations 72 miles apart.
46:27The weather radar specialist at Stephenville makes and disseminates weather radar observations
46:33at least once an hour at about 30 minutes past the hour.
46:38On the evening of August 2nd, the specialist left his radar position at 1735 for dinner.
46:45The other weather service is at Fort Worth.
46:48On the evening of the accident, the meteorologist took a supper break at 525 p.m.
46:54The meteorologist did not return from supper until about 610, which is approximately four
46:59minutes after the accident.
47:01The meteorologist testified that he would have issued a center weather advisory for
47:05the level four thunderstorm north of the airport if he had been at his position at six o'clock.
47:12There were two weathermen monitoring the gestation of the storm.
47:16In the busiest time of day at the world's third busiest airport, both of them went to
47:21dinner at the same time.
47:24Leaving, in effect, the radar scopes unmanned.
47:29The air traffic controllers saw lightning at different times but did not advise incoming
47:33planes.
47:35The controllers had also stated during the interview that they had observed lightning
47:41on at least one occasion, and in the case of the local controller, he had observed one
47:46flash approximately one and a half miles east of runway 17 left at about the time he observed
47:52the Delta aircraft emerge from the cloud mass and rain shower.
47:56The controllers, the tower controller, he had an excellent view of that end of the runway.
48:02He saw that it was pretty severe weather moving in, and he didn't say anything.
48:10None of these things alone would have caused the accident.
48:12Then he put them all together, and then he got in an accident.
48:15And as I mentioned earlier, we felt that the traffic load in the tower was such, particularly
48:19on the east side of the airport, that full attention should have been focused on the
48:24traffic situation.
48:26Runway, the taxiway to runway 17 right, right over here in the evening of the crash, the
48:34time of the crash occurred at 16 aircraft lined up waiting to take off on runway 17
48:39right.
48:40It's the companion runway for 17 left.
48:43Both of these were being handled by one controller, a supervisory controller handling the aircraft.
48:48He had aircraft approaching, 16 aircraft waiting to take off.
48:52The FAA calls it moderate traffic.
48:55I don't see how you can call it moderate traffic.
48:58Weather falls into the category of additional services.
49:02Additional services are required duties of controllers, but are based on the time available
49:08to the controller to provide those services.
49:11Had the controller found time to report lightning, it would probably not have changed the outcome
49:15because Delta 191 was already in the wind shear.
49:19The next line of questioning concerned the airport wind shear alert system and other
49:23potential sources of information about the bad weather.
49:26In this case here, there were no alarms received in the tower until 10 to 12 minutes after
49:31the accident.
49:32Therefore, nothing was provided to the flight crew.
49:36The low-level wind shear alert system in its current format, most airports has only six
49:42sensors.
49:42And the spacing between the sensors is like three kilometers or so.
49:47And unfortunately, many microbursts are like only one or two kilometers across.
49:52So that the problem is the net that we're trying to catch these things in is so coarse
49:57that they slip right through undetected.
49:59When Delta 191 encountered the microburst, it occurred outside of the network.
50:04And consequently, the aircraft got into big trouble.
50:07Once formed, the storm grew very rapidly.
50:10But from the Delta 191 approach side, it did not look as ominous as it did from the control
50:14tower.
50:16Not once before the accident did any arriving pilots complain about weather problems or
50:20request to discontinue an approach.
50:23Several of these crews had noticed lightning, but said nothing at the time.
50:27Delta 191 could see planes ahead landing safely.
50:30So there was no reason not to do the same.
50:32Or was there?
50:34We played the tape through, obviously, with the group around the table here in the cockpit
50:40voice recorder lab.
50:41And it was obvious to us that obvious to them that they did see a thunderstorm.
50:46And they did comment on the fact that there was lightning coming out of it.
50:49The and they chose to proceed into the thunderstorm.
50:53The pilots were clearly flying into thunderstorm conditions.
50:56They were flying into a known rain shower with lightning dead ahead.
51:01Dead ahead.
51:03We've been training pilots for years to stay out of thunderstorms at low altitude.
51:09Having determined how the plane flew into the storm, the board moved on to Delta policy
51:14on wind shear and thunderstorm avoidance.
51:17Whether the pilots had been given sufficient guidance and training in how to detect or
51:20how to fly through wind shears became quite controversial.
51:24And it took the board five report drafts to agree on whether the crew had enough information
51:28to have avoided this one.
51:31And what are what is official company policy?
51:33Yes.
51:35Once the pilot has been taught to recognize situations where wind shear might occur,
51:40company guidance, company policy is to avoid such areas.
51:45In all conditions of flight?
51:48Yes.
51:49There were two basic positions being developed within the board.
51:53We could have split the vote and it would have stood as as the record on that.
51:58Had we done so, I'm afraid that that would have been the story.
52:02That would have been the message to the public and to the airlines and to the FAA,
52:06a confusing one, a split message.
52:08And by taking all three factors and making them equal elements of the probable cause,
52:14it was it was an acceptable approach under the circumstances.
52:18And that's why the vote turned out the way it did.
52:22In effect, the blame was spread three ways on the absence of wind shear hazard information
52:26at the airport, on the crew's bad judgment in flying into a storm and then continuing
52:31to land after the wind shear became apparent,
52:33and on the airline's ambiguous policy for managing situations like this.
52:38We really don't know what more we could have done in the area of training.
52:42But what really concerns us and where we have an enormous concern
52:46is the fact that no warning was given to any of our pilots by the National Weather Service
52:50and by the FAA controllers about the presence of this storm, even though both could see it.
52:56It was clear that we had to indicate to the public and to the airline community and to the FAA
53:04and others that we were very concerned about the action of flying into that cell.
53:11I guess we're very disappointed in those findings because they point back to the crew as having
53:19the last chance at saving the airplane.
53:21They kind of like to point to that as the critical element into the chain of events that happened.
53:27There is a catch-22 involved here in that we are expected by our airlines and by our passengers
53:37to fly in all weather operations.
53:41And we do so, and just statistics prove that we do so very skillfully.
53:47But occasionally, due to information you don't have, or bad luck, or mechanical difficulty,
53:53or something else, it can result in a tragedy.
53:58And then we are blamed for doing what we are in effect and actually required to do on a daily basis.
54:07Whether any pilot alive could have flown through a wind shear of that intensity will never be known.
54:12But in an enterprise as complicated as our air transportation system,
54:17some accidents will inevitably occur.
54:20Pilot error is once again the weakest link in a long chain of human errors.
54:25Perhaps that flight could have been saved with better weather information,
54:28or if someone somewhere on the ground had made a better decision than they did.
54:33But ultimately, it is the crew that flies the plane and must bear the responsibility for decisions
54:39which no one else is in a position to make.
54:42Improved communications, better management of tasks, teamwork,
54:46cockpit resource management by whatever name.
54:49All these may not rescue any given flight, but couldn't they improve the odds on every flight?
55:12Before starting engines, circuit breakers and switches.
55:14Checked and on.
55:15Voice recorder.
55:16On.
55:17Tested.
55:18Oxygen.
55:18On 100. Speakers check and set.
55:20On 100. Speakers check and set.
55:21On 100 percent. Speakers check and set.
55:23Indicating lights.
55:24Check left.
55:24Check right.
55:25Check here.
55:26Flight control power switches.
55:27All on.
55:28Anti-skid.
55:28Test on.
55:29Stall warning.
55:30Tested.
55:30Speed command.
55:31Tested.
55:31Altitude alert.
55:32Tested.
55:33Instrument transfer switches.
55:34Renormal and reset.
55:35Auto brake.
55:36Off.
55:36Emergency exit light.
55:37Armed.
55:37Alternate flap switches.
55:38Pre-off.
55:39Flight recorder.
55:40Pre-off.
55:40Flight recorder.
55:41Tested and set.
55:42Seat belt no smoke.
55:43Both on.
55:43Transponder.
55:44Standby.
55:45Window heat is on. Anti-ice is off.
55:46Pedostatic heat is checked and on.
55:48Navigation lights.
55:49Position on.
55:51Fire warning.
55:51Switches in.
55:52Tested.
55:53Emergency brake and pressure.
55:54Handles wired off. Pressure is 1300.
55:56Flight instruments.
55:57Set and cross check left.
55:58Set and cross check on the right.
55:59Mach airspeed warning.
56:00Tested.
56:01Static sources.
56:02Normal left.
56:02Normal right.
56:04Gear lever.
56:04Down three green.
56:05Radar.
56:06Tested and standby.
56:07Radios and altimeters.
56:08We got a 92 readback.
56:09Weather Bureau 92.
56:11For a transcript of this program,
56:13send $4 to NOVA, Box 322, Boston, Massachusetts 02134.
56:19Please be sure to include the show title.
56:22To purchase film or video copies of this program for educational use,
56:26call toll-free 1-800-621-2131.
56:30In Illinois or Alaska, call Collect 312-940-1260.
56:36This program was produced by WGBH Boston,
56:38which is solely responsible for its content.
56:45Major funding for NOVA is provided by this station
56:48and other public television stations nationwide.
56:52And by Allied Signal, a technology leader in aerospace,
56:56electronics, automotive products, and engineered materials.
57:00And the Johnson & Johnson family of companies,
57:03supplying health care products worldwide.