Original subtitles

For investigators trying to solve a plane crash,

the most important tool can be the black box.

It records every detail in the cockpit...

Look at this.

Where's Charlie at?

...and tells investigators about vital conversations.

Damn, it's starting to rain.

Northwest 255, runway 3 Centre, clear for takeoff.

But in the crash of Northwest Airlines Flight 255...

I've never been to an accident of that scale.

...it wasn't what investigators heard on the tape.

- Control... - TCI was un-set.

It was what they didn't hear... - Checked.

...that would lead to an astonishing conclusion.

Mayday, mayday.

Detroit, Michigan.

August the 16th, 1987.

It's 8:00 p.m., and the city is sweltering.

Twenty-five kilometres from downtown...

...Detroit Metropolitan Airport

is one of the busiest airports in the United States.

More than 1,100 airplanes use its four runways each day.

Today, one of those is Northwest Airlines Flight 255,

bound for Phoenix, Arizona.

Captain John Maus is in command.

A Las Vegas native, 57-year-old Maus is a veteran pilot.

His first officer is 35-year-old David Dodds,

of Galena, Illinois.

What don't you tell 'em we're ready to go?

Both have years of experience on this type of aircraft.

The MD-80 is also known as the Super 80,

and is the second generation of the DC-9.

The MD-80 was quite a bit longer,

it had more powerful engines, it could carry more people.

For that reason, it was a better money-maker

for the airlines than the DC-9 was.

Look at this.

The sky between Detroit and Phoenix is filled with storms.

Several are moving quickly towards the airport.

There's a line here...

For the crew, it's been a long day.

...about 25 miles wide.

Well, if we get outta here pretty quickly, we won't have a delay.

They've already flown from Minneapolis, Minnesota,

to Saginaw, Michigan, and then Detroit.

Phoenix is their next stop on the way to Santa Ana, California.

If we wait till after the storms here,

there'll be delays going over to Waterville.

If they're delayed by weather,

they may not make their final destination.

Let's get outta here before it starts raining.

The plane's 149 passengers are also eager to leave.

Paula Chican and her family have been visiting relatives.

They're heading home to Arizona.

Her daughter Cecilia is only four years old.

- Looks like bags are all in. - Why don't you tell 'em we're ready to go?

Ramp 255 at Delta 15, uh...

Flight 255 is running half an hour late.

Northwest 255 clear to go.

- Okay, we're clear to push. - Let's do the checklist.

- Brakes. - Set.

Windshield heat...

is on.

Boost pumps. We've got six.

Cabin-pressure controller: Checked.

Auxiliary hydraulic pumps... pressure:

On and checked.

Damn, it's starting to rain.

To beat the storms, they need to leave immediately.

...beacons: They're all on.

Before Start checklist is complete.

Flight 255 begins moving from the gate to the runway.

Northwest 255...

But because of the weather...

Northwest 255, now exit at Charlie, runway 3 Centre.

...the ground controller gives them a last-minute runway change.

Okay, out to Charlie for 3 Centre, Northwest 255.

Charlie for 3 Centre, right.

Okay.

Ladies and gentlemen, uh,

we're currently number one for departure.

We should be rolling in a couple minutes.

Flight attendants, be seated. Thank you.

Blacker than hell out there.

Where's Charlie at?

By the time they get to the new runway,

they're 45 minutes behind schedule.

Northwest 255, runway 3 Centre,

clear for takeoff.

Within 17 seconds,

65,000 kilograms of passengers and aircraft hurtle down runway 3C.

But moments before liftoff,

Maus discovers he can't engage the autothrottle.

It won't stay on.

TCI is unset.

His computer isn't in takeoff mode.

They're on now.

Okay.

Clamp, 100 knots.

At 313 kph...

V1, rotate.

...the pilots angle the plane's nose up for liftoff...

...then something else goes wrong.

Just under 50 feet from the ground,

the aircraft begins rolling from side to side.

Tower, lifeguard copter, 120.

It rolls left and strikes a light pole.

Out of control, Flight 255 slams into the ground...

...skids along a highway...

...and disintegrates when it hits an overpass.

I prayed that everybody made it,

but I thought it was just a small plane 'cause it happened so quickly.

I didn't know it was a bigger plane.

And... it was just awful.

I saw the plane come through the viaducts; a big fireball.

When we got dispatched, it was a rainy Sunday night.

And ten he said there's an airliner down, and our mood just changed.

And the one guy looked at me, Dan, and said, "Well, I hope it's a small one."

When we pulled up,

we saw the cockpit and the word "West" written on the fuselage.

We looked at each other and he said it looks like a big one.

There's a trail of scorched bodies and debris more than a kilometre long.

Just minutes after impact,

paramedic Tim Schroeder is on the scene, looking for survivors.

I'd never been to an accident of that scale.

We were struck by the... the magnitude of what we were seeing,

the large scale of it.

It was just... it was almost overwhelming.

From the little that's left of Flight 255,

it is unlikely they'll find anyone alive.

I buddied up with Dan and we both started entering the wreckage.

It was probably a minute went by and Dan actually heard, um, a noise.

He asked me a couple times, you know, do I hear anything, and I said no.

And then finally I heard it.

And there was more like a faint cry.

When I turned my head to the right,

I saw an arm... underneath the seat.

One, two, three, lift!

She was covered in some blood and some soot.

Somehow, four-year-old Cecilia Chican has survived the crash.

But she's badly injured.

Tim Schroeder races her to hospital.

We have a four-year-old girl found alive in the wreckage. She has a very weak pulse.

If Cecilia survived,

perhaps others have, as well.

Rescuers spend hours looking through the wreckage for more survivors.

But their efforts will be in vain.

We actually covered anything that...

was a body or a body part with a yellow blanket.

It was just nothing but, like, a sea of yellow blankets basically.

Northwest Airlines said 154 passengers and crew aboard the plane died in the crash.

Both Captain Maus and First Officer Dodds are killed in the crash.

Two other people died when their cars were hit by the plane.

This is the second-deadliest airplane disaster in U.S. history.

Recovering in hospital with serious head wounds

is Flight 255's lone survivor,

four-year-old Cecilia Chican.

Despite her injuries, doctors say she will live.

Maybe, uh, God was on her side that night.

Within hours of the crash,

Investigator Jack Drake and his team

from the National Transportation Safety Board

begin looking for clues.

Drake is a former Navy pilot who's been involved in hundreds of crash investigations.

You know when you're at a crash site,

because you get this combination of burned plastic and kerosene,

and sometimes combined with fire-retardant foam

that has its own distinctive odour.

You know you've arrived when you smell it.

Drake and his team treat the crash site like a crime scene.

Our team consisted of 12 or 13 specialists,

some of whom go to the site and some of whom do their work elsewhere.

They set out to examine every piece of wreckage

to discover what went wrong.

They have responsibilities for looking at different parts

of the wreckage debris...

...and do qualitative analysis of those parts.

We always look for the recorders first.

They're frequently referred to as black boxes,

although they're usually orange.

The information is usually well protected,

because they're encased in a steel box

that is both heat-resistant and crash-resistant.

Since the 1960s,

commercial jetliners have been required to carry flight-data and voice recorders.

The CVR was first introduced in Australia,

following the 1960 crash of a Fokker F27.

The devices must be able to withstand an impact of 3,400 Gs

and temperatures as high as 1,100 degrees Celsius.

The cockpit voice recorder is intact.

But the flight-data recorder suffered some damage in the crash.

They may hold the only clues that can help solve this accident.

Both recorders are sent to the NTSB lab in Washington, D.C.

John Clark is Drake's flight-performance engineer.

His first task is to make a map of the debris left behind by Flight 255.

When I first started seeing the wreckage,

it, uh... your mind immediately starts turning to, uh...

sorting out where it hit, how it hit.

Not where the wreckage ended up,

but those first few inches where the airplane was coming down.

Clark looks for ground scars and other impact marks,

and interviews witnesses to piece together where the plane fell and how.

And, uh, that kind of gives you a sense of those last moments,

what the airplane was doing when it hit the ground.

Clark begins to understand the final moments of Flight 255.

According to witnesses, as it lifted off,

it couldn't climb and flew in a nose-high position.

V1. Rotate.

That could indicate that the plane didn't have

enough power to get off the ground,

that it didn't have enough speed,

or that high winds prevented it from lifting off.

Witnesses provide investigators with a critical clue.

You saw fire coming from the engine?

Several, including an air-traffic controller,

saw flames coming from the plane's engine before the crash.

Damn! No!

The engines become the first focus of this investigation.

They looked for evidence of an internal failure.

The plane suffered an engine failure...

Investigators soon learn that less than a month earlier,

one of the plane's engines was damaged when it was hit by a foreign object.

It was repaired and being monitored by mechanics to see how it performed.

The team studies the remains of the engine for clues that it had either caught fire

or shut down on takeoff.

Despite what the witnesses saw,

they find no evidence of fire or of a massive breakdown.

The information suggested that the engine operation had been normal.

Clamp, 100 knots.

The flames were the result of the fuel tank rupturing

after the plane hit a light pole.

The fire didn't cause the crash.

If Drake and his team are to solve this mystery,

they need to be certain about what happened

in the last few seconds before Flight 255 crashed.

Well, I think probably the best physical evidence

is what was on the flight-data recorder.

The flight-data recorder doesn't tell you about weather,

it tells you about aircraft parameters,

aircraft performance, essentially, second by second,

even at quarter-second intervals in some parameters.

But NTSB technicians can't recover

all the information from the recorder.

They send it to the manufacturer to see if can recover the lost data.

While he waits for news about the flight-data recorder,

Jack Drake looks more closely at the weather on the night of the crash.

There was some convective or thunder-shower-type activity that had moved through the area,

and its impact on the accident required a lot of analysis.

Drake wonders how the storm affected Flight 255.

He listens to the cockpit voice recorder for clues.

He discovers the menacing weather was a concern to the crew.

Let's get outta here before it starts raining.

Jesus, look at this.

Drake sees that there were several storms along the flight path,

and they were getting closer to Detroit.

There's a line here...

...and a line between these two.

Uh-huh.

And another one here about 25 miles wide.

Thunderstorms can create a very dangerous threat to pilots.

Since there was severe weather in the area,

we always worry about microbursts.

Microbursts occur when columns of air shoot down to earth.

As a plane passes through,

winds batter it from all directions,

making it difficult to control.

In a microburst condition,

you can get very shifting winds, and also vertical winds,

that will try to push the airplane into the ground.

So you can get a loss of airspeed, a very rapid rise in airspeed,

and then also actually push the airplane toward the ground.

And it requires very aggressive flying on the part of the flight crew.

This unusual weather condition had killed before.

In 1985, a microburst brought down a Delta Airlines flight in Dallas,

killing 137 people.

At the time of the Detroit accident,

there was no device at airports to accurately detect microbursts.

Instead, pilots relied on reports from other crews.

Jack Drake discovers that 27 minutes before liftoff,

Captain Maus and First Officer Dodds received such a warning.

Ground, this is 722.

You just had a microburst out here. The dust exploded down there.

Investigators suspect that a microburst

may have slammed Flight 255 to the ground as it tried to lift off.

It was right after departure when the accident occurred.

If they had struck a very strong microburst,

that would be a candidate for one of the possible causes.

Satellite images taken at the time of the crash,

and weather data from the airport's sensors,

show that there were storms near the airport at the time of the accident,

but there's no evidence of a microburst.

Wind and rain, but nothing that could be a microburst.

Around the time of the crash,

the airport's sensors did record a dangerous gust of wind on the runway,

powerful enough to set off alarms in the tower.

Upon further investigation,

Drake discovers that Flight 255 was still at the gate at the time of that alarm.

So the winds couldn't have brought the plane down.

Sign beacons.

But they did have a huge effect on Captain Maus's flight plan.

The crew's pre-flight dispatch package stated

they would take off from runway 21 Left.

But with the sudden change on wind direction,

Ground Control sends Flight 255 to runway 3C,

the shortest of three available runways.

Northwest 255, now exit at Charlie, runway 3 Centre.

Yeah...

Okay to Charlie for 3 Centre.

Northwest 255.

Charlie for 3 Centre, right.

Okay.

Controllers try to have planes take off into the wind.

The additional wind flowing over a plane's wings gives it more lift

and helps it get off the ground.

Taking off into the wind is safer,

but taking off on the shorter runway

now means First Officer Dodds must recalculate the plane's takeoff weight.

If there's a runway change,

you have to determine if the weight of the aircraft will permit it to accelerate

and climb out safely; and this varies,

depending on the length of the runway, temperature, altitude of the airport.

Perhaps First Officer Dodds made a mistake in his calculation.

4044? How can we be that light for a full airplane?

If he did,

it could explain why the MD-80 wasn't able to make it off the ground.

Runway 3C simply wasn't long enough.

Using calculations based on average weight of luggage and passengers on board,

Drake's team confirmed Dodds's estimate.

The plane weighed 144,047 pounds,

well below the allowable limit for runway 3C.

It should've been able to get off the ground.

Drake's investigation has hit another dead end.

He and his team are running out of possibilities.

Until technicians can decode the damaged flight-data recorder,

the team must rely on the physical evidence

that's been found at the crash site.

But when investigators study the cockpit's centre console,

they're forced to consider an almost unimaginable cause.

Is this the way it was found?

To get the plane off the ground,

the flaps on the wings should've been extended to the 11-degree position.

But the way the flap handle is damaged suggests

the plane's flaps were retracted when it crashed.

The pin... had a left mark.

This happens because the aircraft comes to a very sudden stop,

and the handle jangles around and...

and it's a metal-to-metal contact

that's exaggerated by the impact.

It won't stay on.

That indicator was that the flaps were zero, or fully retracted,

and the slats were retracted, as well.

Flaps and slats are extensions

that slide out of the back and front of the wing.

They make the wing bigger,

which increases the amount of lift they can provide.

They must be extended for takeoff.

If the slats are retracted, for the most part,

with today's modern jets, the airplane is not capable of flight.

If the crew tried to take off with the flaps retracted,

it would be an astonishing blunder.

V1. Rotate.

But a pilot who was lined up directly behind Flight 255 on the runway

is certain the plane's flaps were extended.

And you're sure the flaps were extended?

Pilots in other aircraft that were close to the point of where the takeoff had begun

were telling us that they thought the flaps and slats

were deployed to a normal position.

Investigators can't be sure whether the flaps were extended or not.

The clues they need lie somewhere in the sea of debris recovered from the crash site.

Eventually, investigators find the evidence they need

inside a section of the plane's left wing.

Each component of the slat system has its own drive system,

and one of those was interrupted

by the light pole that passed through the wing.

Eighteen feet of the left wing was severed.

The cable controlling the slats was sliced in two

when the wing hit the light pole.

Based on where the cable was cut,

investigators can tell whether the slats and flaps were extended or retracted.

It severed two cables,

and if you lined up those two severed ends,

it corresponded with the slats being in the full retracted position.

It looks increasingly likely

that the crew never extended their flaps.

Only the damaged flight-data recorder can verify this.

Fortunately, technicians have finally been able to rescue all its data...

...a digital history of Flight 255's performance until the moment of impact.

I knew that...

if we had a good recorder, we were gonna get data back.

The flight-data recorders, in combination,

give you that time history that goes together with the physical evidence,

or physical damage.

As expected,

the FDR confirms what the evidence has been showing investigators.

The flight-data recorders told us that the, uh,

the flaps and slats had not been extended.

It's a major breakthrough.

Drake now knows what brought down Flight 255.

But the flight-data recorder doesn't answer a more troubling question.

So... why weren't the flaps deployed?

For some reason,

a seasoned crew forgot one of the most basic steps involved

in getting an airplane off the ground.

Two months after the crash...

...Northwest Airlines Flight 255's sole survivor,

Cecilia Chican, is released from hospital.

We can't be sure why the little girl survived.

She's a very little girl buckled into a big seat.

And she was more protected than adults that might've been sitting around her.

So she was very lucky.

Jack Drake needs to know what contributed to the death

of her family and all the other victims.

He finds an important clue on the cockpit voice recorder.

It shows that the last-minute runway change

caused confusion in the cockpit.

Once the aircraft began to taxi...

- Blacker than hell out there. - Northwest 255,

now exit at Charlie, runway 3 Centre.

Other activities were introduced

that had the potential to cause distractions.

Did he say 3 Centre?

3 Centre, yeah. That's why I was thinking we had to go that way.

I was thinking 21.

Well, they made a wrong turn, which might've been confusing,

because they had to go a different route.

Where's Charlie at?

Right at the end of this ramp.

- I think Charlie was-- - No, it is Charlie.

You sure?

I think so.

The crew got lost on the way to runway 3C.

Ground, Northwest 255.

I guess we went by Charlie. We're going to 3 Centre, right.

Northwest 255, affirmative, make a left turn at Foxtrot.

They finally got to the runway 45 minutes late.

Ladies and gentlemen, we're currently number one for departure.

Should be rolling in a couple of minutes.

Flight attendants, be seated. Thank you.

Brakes. But Jack Drake finds something missing on the CVR.

It seems the crew overlooked a very important step.

We're okay for the... centre runway, aren't we?

Before they got lost,

the crew of Flight 255 performed a number of checklists.

But possibly due to the confusion of the runway change,

they seemed to have completely neglected the taxi checklist.

They apparently didn't consider the checklist,

and key in the checklist is the configuration of the aircraft for departure.

And the flight-data recorder showed that was never done.

There are hundreds of small steps for a crew to take

to get a passenger jet off the ground.

Most of them are covered by checklists.

The checklist is a means by which you ensure

that important items are positioned or done properly.

- Transponder. - Instead of doing it by memory,

and having the possibility of a lapse of memory...

...flight crews use a very rigorous and regimented procedure

of following the checklist...

- Cabin-pressure control. - ...to verify that each switch,

each dial, each lever is in the proper position

before taking the runway for departure.

The first item on the taxi checklist is flaps.

One of the things that would've been included in their checklist

was to configure the slats and flaps for low-speed flight.

But because they didn't run the checklist,

the crew never set their flaps to the takeoff position.

- I think Charlie was-- - No, it is Charlie.

You sure?

I think so.

They hadn't done this checklist at the time they normally would.

And as the activities piled up,

um, that were potential distractions,

they were further and further away from the point

at which they would normally perform that function.

Their mindset was probably that they had completed it.

The pilots got an indication that their plane wasn't properly configured.

During takeoff, they couldn't activate the autothrottle,

because their computer wasn't in takeoff mode.

Another step covered by the taxi checklist.

It won't stay on.

TCI is unset.

This should've alerted them

that they didn't perform the checklist...

If it had occurred to them at that point

that we might've missed something else on the checklist,

that could've led to a rejected takeoff.

They're on now.

Hundred knots.

Okay.

Apparently, that didn't happen,

and so the takeoff was continued.

...with disastrous results.

An alarm should've sounded when the pilots tried to take off with their flaps retracted.

But for some reason, investigators can't hear it on the cockpit voice recorder.

When it activates, it alerts the crew that the aircraft

is not in a configuration that's safe for takeoff.

Maybe it went off but we just can't hear it.

The investigation team is determined to find out

why the takeoff warning didn't sound.

Technicians analyze the cockpit voice recorder for more clues.

And they find something unusual.

They picked up, uh, some annunciations on the CVR that were not correct.

Stall.

Stall.

This warning is alerting the crew that the plane is about to stall.

But it should be coming from two speakers in the cockpit.

Stall.

Technicians notice it's only coming from one.

Stall.

Stall.

As the airplane lifted off, there was a stall warning,

and it has a typical characteristic

of a sound like "stall-all,"

because there are two annunciations,

and the purpose of that is to provide redundancy.

But that redundancy wasn't there. We had a single "stall."

I went to an MD-80 sitting on the ramp at Detroit.

And a captain took us through the process of checking out

to demonstrate those different sounds.

Let's start with the config warning.

The takeoff configuration warning is what would've alerted them

about the flaps and slats.

Flaps.

Slats. Flaps.

Can we get the stall warning to sound?

Stall all. Stall all. Stall all.

He activated the stall-warning system by a test switch,

and it said: "Stall-all."

The voice on the left channel is slightly different

from the voice on the right, as it should be.

But that's not what Clark heard on Flight 255's voice recorder.

Can you make it sound like this?

Stall.

Stall.

To get a singular stall,

he had to pull power to one side or the other.

One way he demonstrated that is he pulled the P40 circuit breaker.

A circuit breaker is the electrical switch

that protects the circuit from damage caused by overload.

The P40 circuit breaker is an important one in this investigation.

It handled both the failed takeoff-warning and the stall-warning systems.

What struck me was he said, "I hear people doing it.

I, of course, don't do it myself, but let me show you how."

And he reached around behind him,

around behind the seat and down low,

and pulled the P40 circuit breaker without looking.

And then, when he ran the stall-warning system, we got the singular "stall."

Stall.

Stall.

It's a major clue.

Clark could only reproduce the strange-sounding stall warning

by pulling the same circuit breaker that's connected

to the takeoff warning.

This tells investigators that the breaker was tripped

when Flight 255 tried to take off.

Then John Clark notices something else about the P40 circuit breaker.

You could see smudge marks around the decals

on each side of the circuit breaker.

It looked like fingermarks where oil had built up,

and dirt and grime, over the years.

So it told me that circuit breaker

was being used routinely by a lot of pilots.

Can you tell me why that is so worn?

It turned out that the takeoff-configuration warning

could be a nuisance to pilots.

If you're doing a single-engine taxi,

you have to push the throttle up further to get a power to taxi,

and you would set off the takeoff-warning system.

So they would pull the circuit breaker to silence it.

It's irritating. It's a warning.

It... it's meant to alert you.

And if it's going off routinely all the time,

it gets on their nerves and, uh...

so... apparently pilots were routinely silencing those takeoff warnings.

Investigators suspect that the crew of Flight 255

tripped the breaker to avoid the irritating takeoff warning.

And then, with the added delay from the runway change and the impending storm...

...they proceeded to take off without doing the taxi checklist.

That might explain why the alarm didn't sound

when they tried to take off with their flaps retracted.

We don't know if the pilot did pull that circuit breaker on that particular flight.

There was certainly one error and the potential for two.

I think the extensive use of the circuit breaker,

because of the smudge marks around the circuit breaker and the pilots' statements,

I think it's highly likely that he did.

It appears the downing of Flight 255 was caused by pilot error.

Now investigators can accurately piece together

what happened that night in Detroit.

But it would take another shocking accident

for the airline industry to learn its lesson.

Jack Drake's team has discovered what caused the crash of Flight 255...

...but cannot prevent it from happening again.

One year later, in Dallas,

Delta Flight 1141 tried to take off without their flaps extended.

The investigators who had been working on the Northwest crash are stunned.

I was very frustrated to learn that...

that another airline had done the same thing

in a different aircraft type about a year later.

The Delta crash would uncover potentially deadly flaws...

Okay, Before Start checklist, battery. Uh...

...in the checklists commercial pilots are trained to follow.

The Delta and Northwest crashes killed 170 people...

...and had eerily similar causes.

In both disasters, the workload in the cockpit increased...

Jesus... look at this.

Northwest 255,

now exit at Charlie, runway 3...

If we get outta here pretty quickly, we won't have a delay.

And in both,

the pilots failed to perform vital elements of their checklists.

It is very unusual for a crew to not perform a checklist.

They have done it hundreds upon hundreds of times.

Brakes.

Windshield heat... is on.

Cabin-pressure controller... is checked.

The normal procedures, uh, were a little bit out of the norm.

And as a result, it got overlooked.

To prevent this from happening again,

aviation officials turned to a government agency

that knows the importance of clear procedures:

- Four, three, two... - Discovery, you're go.

...one.

...NASA.

- So they're ready to go. - Alright.

Jack Drake and his team wanted the U.S. space agency

to help create checklists that decrease the odds of items being skipped.

Asaf Degani was a research scientist working with NASA.

After the accident,

he took on the project of improving a flight crew's pre-takeoff procedure.

We looked for any research that was done on checklists or procedures in general.

And, in fact, we couldn't find anything.

So Degani had to start from scratch.

But there are dozens of different checklists to examine.

Most of the ones on Flight 255 were printed on a single card.

They listed the tasks the crew had to carry out,

but didn't give them a way to keep track of what and what wasn't done.

At the time of the Northwest crash,

there were several types of mechanical checklists in use.

The U.S. Air Force used a scrolling checklist.

Once a checklist item is completed,

the pilot scrolls to the next one.

American Airlines used a system that allowed pilots

to cover up completed items with a plastic slide,

so only the non-completed items would be displayed.

Asaf Degani set out to see firsthand how pilots were using checklists.

He wanted to make it less likely for them to make mistakes.

He sat in cockpits and observed 42 different crews in action.

Degani concluded that many checklists were badly designed.

There's a certain flow by which you go about checking things.

And the idea is to prevent the case where you're doing one thing here, one thing there.

A checklist should have a certain flow,

which is a logical flow and not one which is kind of random.

And they're ready and they're going through...

Degani also found a much more serious problem with checklists.

If pilots are interrupted, they sometimes forget where they left off.

Transponder...

checked and on.

And there's many cases where people would do, A, B, C, D, E,

an Air Traffic call would come, they'd have to respond to it, and that's an interrupt.

Northwest 255, now exit at Charlie, runway 3 Centre.

Yeah...

They would go back to the checklist,

and skip a certain item and continue on the list,

assuming that the whole list was done.

People were very concerned about that.

To ensure no steps are missed,

airlines train their pilots to return to the top of a checklist

following the interruption and start over.

Again, Asaf Degani sees a problem.

And we found that if checklists are very, very long and meticulous, um,

that's overburdening the crew and they sometime would not want

to start again from the beginning.

To address the problem,

Degani suggests changes across the airline industry.

One of our recommendations from the study

is to try to take long checklists and chunk them to small pieces,

so that if an interruption happens,

then doing another four or five items,

it's not a big effort, as opposed to doing 20.

Degani even made recommendations

about the typeface that airlines use

so they can be more easily read by pilots.

Perhaps the biggest advance to checklists is the move from paper to computers.

Since, at the time, um, computer technology was coming into the cockpit,

it made a lot of sense to think about electronic checklists.

Today, Degani and his team are studying smart checklists

that keep track of checked items.

They provide pilots with a visual indication of where they are on the list,

and in some cases, verify that the task has been correctly carried out.

An electronic checklist shows you which item...

was completed and which item was not.

Computerized checklists are now slowly making their way

into the cockpits of commercial airplanes.

They make it far less likely that an accident like Northwest 255

could happen again.

The FAA also ordered a modification to the alarm system

of all commercial jetliners to prevent nuisance alarms.

The takeoff warning was redesigned

so it could not sound unless the plane was actually taking off.

Jack Drake went on to investigate hundreds of accidents

over a 26-year career with the NTSB.

The crash of Flight 255 taught him a valuable lesson.

This one is a worldwide example of...

the importance of following checklists

and configuration being completed correctly on every takeoff.

And so it became something that was a part of the training curriculum

in virtually every airline around the world.

The case of Northwest 255 is no different.

It was a series of events:

runway change, task saturation, an overlooked checklist,

a failed takeoff warning system.

Put all of those together, those links in a chain,

and you end up with the accident.

If you were to break any one of those links,

the accident wouldn't have happened.

Flight 255 will also be remembered for its lone survivor,

Cecilia Chican.

She's never spoken publicly about the death of her family.

But she's stayed in touch with the people who rescued her that day.

She's full of life, and the conversations we have,

it's more about, you know, her sports and her husband and her vacations.

She just said maybe one day,

she'll come out and tell the world what Cecilia is doing.

difuze