Mayday (2013) s10e04 Episode Script
Dead Tired
VOICEOVER: February 12, 2009.
Just outside Buffalo, New York.
Gear down.
- Looks alive.
- Gear's down.
- (RATTLING) - Oh! - Oh.
- Jesus Christ.
The airplane entered an aerodynamic stall.
- It did not recover.
- (SCREAMING) (SIRENS WAIL) It was one of the grisliest, nastiest scenes that I think I've seen.
The crash of Continental Connection Flight 3407 would be one of the NTSB's most important in decades and would see a grieving father fight for changes to the laws governing small airlines.
I'm focused and determined to change what exists and not have another dad sitting here.
Continental Connection flight 3407 operated by Colgan Air is en route from Newark, New Jersey, to Buffalo, New York.
It's been a busy flight for captain Marvin Renslow.
He's providing guidance to a new first officer, Rebecca Shaw, a former flight instructor from Seattle.
(SNEEZES) Bless you.
Shaw's only been with the airline for just over a year.
She must decide if she now wants to become a captain.
I don't know what I want to do with the upgrade.
Depends where I'm based.
Well, think of it this way, if you stayed on the Q, obviously you're not making the captain, right? Right.
But you may have a better quality of life, with regards to buying a house, having a Shaw trained to be a first officer on the plane they're flying now - a Canadian-made Bombardier Q400.
It's a twin engine turbo-prop, popular with regional airlines.
The 45 passengers have had a long night.
Their plane was held up for two hours at Newark, a delay considerably longer than their journey.
Flight 3407 is heading north-west over Upstate New York.
The trip is only 53 minutes.
Visibility is poor, and there's a forecast of snow and moderate winds in Buffalo.
Just some water, please.
Ellyce Kausner is a student at Florida Coastal School of Law in Jacksonville.
She's on her way to visit her family in Buffalo.
MAN: She had five nieces and nephews at the time and they had a love luncheon at the kids' school, a couple of the nephews, and they both wanted Aunt Ellie to be their guest.
Any excuse to come home, she came home.
She talked to all of us from Newark.
- Hey.
- Then she called us from the lounge.
Yeah, it's been over two hours.
She was PO'd because of the delay.
Folks, from the flight deck, your First Officer speaking.
It looks like at this time, we're about 10, maybe 15 minutes outside of Buffalo.
Weather in Buffalo is pretty foggy - it's snowing a little bit there.
I'd like to make sure that everyone remains in their seats so that the flight attendants can prepare the cabin for arrival.
Thank you.
Colgan 3407 descend and maintain 2,300.
OK, down to 2,300.
Colgan 3407.
Let's do a descent checklist, please.
We can do the approach checklist along with it.
Sure.
Bug set.
Set.
GPWS landing flaps, selected 15 degrees.
The pilots go through a list of settings for the plane to continue its landing approach.
Transfer off? Yeah.
Passengers can already glimpse the lights of Buffalo's suburbs.
Clarence Center is on the approach path for aircraft landing at Buffalo's airport.
(YAWNS) Gear down.
- Looks alive.
- Gear's down.
Right.
Flaps 15.
Extending the flaps provides more lift, allowing the plane to slow to its final approach speed.
- (ALARM BEEPS) - Er Oh! Oh! Jesus Christ! Suddenly, the control column starts to shake.
The Q400 is slipping out of control.
(SCREAMING) Captain Renslow struggles to keep his plane flying.
(GRUNTS) Come on there! Oh! But it seems to have a mind of its own.
Should gear be up? Gearup! Oh, damn! - Oh, we're down.
- (SCREAMS) Oh! The plane crashes into a house at 6038 Long Street.
At air traffic control in Buffalo, Flight 3407 disappears from radar.
Check Colgan there - see if he's on you.
MAN: (ON RADIO) What's his call sign? Colgan 3407.
- Nope, I don't have him.
- Do you see anything out there? Nothing.
Alright.
Call the fire department.
I'll never forget the dispatcher's words - "Plane crash, house on fire.
" I immediately got my shoes on and my coat on, got in my truck and started heading down the driveway.
Firefighters don't have far to go.
6038 Long Street, the home of the Wielinski family, is less than a block from the Clarence Center station.
As I turned towards the village, the whole village was aglow.
You could see the smoke.
You could see the flame.
Karen Wielinski and her daughter make it out of their house alive.
But her husband, Douglas, is killed.
My initial reaction was there's no .
.
there was no way that somebody made it out.
But the division reported that they had two people and they were taking them to the hospital.
I said, "Well, which firefighters are they?" and they said, "They're not firefighters, chief.
"They're survivors from the house.
" I actually physically had to go to the ambulance and look inside for myself to believe.
Firefighters have never seen a blaze like this.
If you've ever gone to a bonfire and stood five feet from it, and you couldn't take it anymore, that's what it felt like from 100 feet away.
MAN: The height of it was just unbelievable.
Obviously because of the fuel that was probably added to it and the debris area was very large.
It was a very, very horrific sight.
We live in the town where the plane went down right down the road.
My son was driving home from a soccer game and went right by the plane crash and called me up and said, "There's a plane down in Clarence.
" Never really entered our brain that it was Ellie's plane.
Flight 3407 has crashed 8km short of Buffalo airport.
All 49 passengers and crew are dead, including Captain Marvin Renslow and First Officer Rebecca Shaw.
But many more people could easily have been killed.
You picture a house and to see a plane on top of this house, no more house left, no more plane - only the tail section.
You're asking yourself, "How in God's creation did this happen "and not wipe the whole block out?" The next call was again from my son and he said in fact, it was Ellie's plane.
He said, "Dad, there's no survivors" .
.
at which point my wife shrieked and just hit the ground.
Chris said he's never heard his mother make a sound like that and I hope he never does again.
So, that's when I knew she was gone.
That was probably within 45 minutes of the accident.
The scene in the morning is one of utter devastation.
Clint Crookshanks from the NTSB.
Can I start poking around? Clint Crookshanks is one of the first investigators on the case.
When we arrived on scene, there was a fire still burning.
It turns out it was from a gas line that had been broken in the house.
The firemen would put the fire out and it would reignite every couple of minutes.
It was one of the grisliest, nastiest scenes that I think I've seen.
Flight 3407 may have been a small plane but it's the worst crash in the US in more than seven years.
It will become one of the NTSB's most important investigations.
The fire at the crash site of Flight 3407 continues to burn, preventing investigators from examining the wreckage.
That's where I think the box is.
The airplane had crashed into a house and then it had burned all night long, so all the debris was basically in the house.
WOMAN: Our concern is that we're losing evidence.
It's perishable and if we can't get in there and get the fire out, then we're not able to maybe get a hold of evidence that might help us during the investigation.
Clint Crookshanks urgently needs to recover the black box flight recorders, which could contain valuable clues about the accident.
We knew that the recorders were in the tail part of the airplane.
Ordinarily, investigators don't go near a crash site that's still burning.
But if the black boxes can't be rescued, they may never find out what brought down Flight 3407.
We started looking around and poking around into the wreckage and actually found out where they were.
OK, whoa.
Whoa.
It's probably in here, OK? The access panel's on the other side so we're going to have to cut a hole right there.
OK? The fire department produced a chop saw.
We were able to cut a hole in the side of the fuselage .
.
and go in and grab the recorders and pull them out.
Alright, that should do it.
To the immense relief of all, the recorders are undamaged.
Once we took the recorders out of the airplane, we put them on the jet and they were flown back to Washington, DC to our headquarters.
Now investigators are faced with a new hurdle.
What little is left of the aircraft is hopelessly jumbled together with human remains and debris from the house.
CLINT CROOKSHANKS: It all burned and settled into the basement so we had probably 10 feet of debris that we had to dig through in order to recover all of the airplane.
Authorities wonder how they can salvage any useful evidence from this chaos.
They get invaluable assistance from an unusual source.
A group of students learning to process crime scenes is enlisted to separate human remains from the rest of the debris.
They were graduate students from a local college - forensic anthropology students - and this was good experience for them to come dig through wreckage like this and look for human remains.
So, landing gear over there, please.
Thank you.
It's dirty, painstaking work but it frees up investigators to concentrate on the aircraft ruins.
We were on our hands and knees with brooms, with little shovels, scooping out debris - identifying it as to house debris or airplane debris and then putting it in different piles.
The first question for me as a structures engineer is to figure out if the whole airplane made it to the scene of the crash.
The wreckage is carefully studied to determine if the plane's four corners - nose, tail and both wing tips - are present.
If we find all four corners of the airplane, then we know that there was no inflight break-up.
There was nothing that departed the airplane during the flight that may have caused the accident.
One wing has been consumed by fire.
The other is shattered into pieces.
Hey, have a look at this! But investigators are gradually finding what they've been searching for.
Oh, yeah.
That's the last piece.
It's the left one.
It wasn't until several days into the investigation as we were scraping away some of the debris that we actually found evidence of the left wing and the nose.
Investigators now have all the pieces they need to conclude that the entire plane is at the crash site.
Whatever caused the disaster was not the result of a break-up in flight.
The tragedy of Flight 3407 is under intense scrutiny at the NTSB offices in Washington.
Officials there try to determine if some kind of malfunction caused the crash.
On this particular accident, we knew it was a landing accident, so we wanted to check the aircraft's performance and then also we checked for flight control continuity.
Much of this responsibility falls to Scott Warren.
My role in the investigation was to be the systems group chairman for the safety board, so I was in charge of looking at all of the aircraft systems.
Warren analyses the aircraft's cockpit voice recorder, or CVR, to determine if there were any indications of a problem in the cockpit.
REBECCA SHAW: .
.
prepare the cabin for arrival.
Thank you.
Is that ice on the windshield? He discovers that six minutes before the crash, the crew of Flight 3407 had noticed a build-up of ice on the aircraft.
Got it on my side.
On yours? Oh, yeah.
Oh, it's lots of ice.
Ice can be a deadly threat to any airplane.
If an aircraft has ice on it, it will have more drag on it, so it will require more power to maintain a given airspeed.
That's the most I've seen on the leading edges for a long time.
Until a year ago, Rebecca Shaw had no experience with ice on an aircraft.
I've never seen icing conditions.
Never de-iced.
Never experienced any of that.
When ice accretes on a wing, it adds weight to the airplane but most importantly, it changes the shape of the wing, and, of course, it's the shape, the curved shape of the wing that actually creates the lift.
So by changing the lift characteristics of the airplane, it makes it less able to fly.
You know, I'd have freaked out.
Like I'd have seen this much ice and thought, "Oh, my gosh" - we were gonna crash.
Observing that there was ice on the airplane was an important thing to do, but then it would be important to turn around and verify that your anti-ice and de-icing equipment was on and there was no indication on the voice recorder that they actually re-checked.
Flaps 15.
The CVR reveals that only minutes after the crew detected ice, a device called the stick shaker went off.
It was a warning that the plane was about to stall - literally to fall out of the sky.
Shortly afterwards, the aircraft did precisely that.
(SCREAMING) We know there were icing conditions.
The only question was were they bad enough to induce this airplane to have a failure? Ice is now the chief suspect in the crash of Flight 3407.
If it's ice, let's prove it.
We were all convinced that we had an icing accident.
We said, "Yeah, it's going to be an icing accident.
"We've just got to make sure "we look for these few key aspects of icing.
"Verify that and we're good.
" But some of the victims' family members are less convinced.
Initial thought was that it was an ice-related incident - it was an icy night that brought the plane down.
And as we began to talk to people who knew airline travel, who were pilots themselves, they would look kind of askance and say, "I don't think so.
"That doesn't make sense.
" Some investigators share that suspicion.
Scott Warren knows a plane like the Q400 has a sophisticated de-icing system.
It's designed to keep ice from building up on the wings and other critical parts of the aircraft.
To prevent ice accumulating, the plane has rubber bladders along the front of the wings, called 'de-icing boots'.
A series of valves uses air from the engines to inflate the boots and crack the ice off the wing.
Those boots are designed to inflate periodically, and that inflation breaks off the ice that's accumulated on those leading edges.
Warren now wonders if the crew of Flight 3407 actually turned on their de-icing equipment.
He studies data from the plane's other black box, its flight data recorder, or FDR.
It tracks the workings of crucial aircraft systems, including the de-icing mechanism.
We know from the FDR data that the de-icing system had been selected on by the crew and it was on during the majority of the flight and certainly at the end of the flight, it was recorded in the on position.
But now Warren needs to figure out if the device was actually working.
Just because the data records that the system is on, that's a start, but you can't necessarily believe that one piece of information.
The only way to know is to find what remains of the plane's de-icing system and determine if it was active when the aircraft went down.
Clint Crookshank's team hunts for a crucial component of the system - the valves.
Excuse me.
OK, everyone, I just want For the de-ice valves, we said, "Here's what it looks like.
Look for something like this.
"If we can find those valves "then we can test them and see if they're operating correctly.
" The valves were very important to us because they are the key component that moves air from the engines where the bleed air originates for the de-icing system.
OK, good.
Five more to go.
Five of the plane's six de-icing valves are eventually recovered.
Some were badly burned, some were in fairly good shape.
And we took those valves and we conducted as much testing on those valves as we thought was appropriate for the level of damage.
We looked at the level of electrical conductivity and we looked at the pressure testing.
We looked at a wide variety of things, depending on the condition of the valve itself.
And as far as we could tell, all those valves were working properly.
- Is that ice on the windshield? - I've got it on my side.
Now investigators have to consider the possibility Oh.
yeah.
Oh, it's lots of ice.
.
.
that there was so much ice on the wings that the de-icing system couldn't get rid of it fast enough.
Oh, yeah.
That's the most I've seen on the leading edges in a long time.
It has happened before.
In Roselawn, Indiana, a French-built commuter plane crashed in 1994 after suffering a catastrophic build-up of ice on the wing.
By analysing how much power was needed for the plane to maintain its airspeed, investigators reach a surprising conclusion.
Basically, the aircraft was flying as if it had a relatively small amount of ice on it, but a very manageable amount of ice.
Looks normal.
It was not overloaded.
It was not excessive.
If ice didn't bring down the plane, something else must have caused it to stall.
Warren studies the Q400's operating manual to learn the plane's stall speed - the velocity at which it can no longer generate enough lift to stay in the air.
He discovers that in the conditions Flight 3407 was flying through, it's roughly 111 knots.
He now compares the plane's stall speed with its actual airspeed just before the stall warning went off.
131 knots - well above the danger zone.
Flaps 15.
When the stick shaker went off - Er Oh! - Jesus Christ! .
.
they were not necessarily at the edge of a stall.
They were still 20 knots or so away from the stall.
This new revelation deepens the mystery of Flight 3407.
If it wasn't stalling, why did it fall out of the sky? Investigators turn their attention to the critical seconds before Flight 3407 went out of control.
They look for clues that could explain why the stall warning went off when the aircraft was flying well within its safety margins.
They discover that this plane has a unique feature known as a reference speed switch.
It governs the sensitivity of the plane's stall warning.
ROGER COX: Very few airplanes in my experience have such a switch.
This airplane is the only one I know that has an actual switch on the overhead panel.
It was designed by the manufacturer to be an extra safety feature.
Some kind of variable ref speed? Pilots are supposed to turn on the reference speed switch when they are going to be flying through icing conditions.
And we'll probably be picking up some ice.
When in the increase position, it reminds pilots to fly faster to counteract any drag effect ice will have on the aircraft.
When you are in icing conditions and ice does accrue on the wing, it can cause the stall speed to go up, and so this ref speed switch correspondingly causes the warning to come on sooner or at a higher speed.
What that switch does is it basically changes the trigger settings for the stick shaker.
So we had to ask the manufacturer, "How does this switch work?" What we found was it was part of the system's description that the crews got when they went through training but they didn't get a lot of training on how to handle that switch.
It seemed like it was too simple to worry about.
Investigators need to know if the crew of Flight 3407 had turned on the reference speed switch .
.
triggering the stick shaker at a faster than normal speed.
The flight data recorder doesn't show whether the switch was on.
Investigators must find another way of determining its position at the time of the accident.
(PHONE RINGS) Clint here.
Clint Crookshanks is given a new priority Alright, I'll see if we have it.
.
.
recover the ice protection panel from the Q400's cockpit, where the reference speed switch is housed.
Since the panel was in the cockpit, finding it is a challenge.
Most of the front end of the airplane was consumed by fire, so we didn't find anything except for little balls of molten aluminium little wire bundles and a lot of ash.
But after an extensive search, Crookshanks discovers that the ice protection panel is one of the few pieces of the cockpit that survived.
Bingo.
However, the knobs and switches are barely recognisable.
Crookshanks examines the charred panel to check the position of the reference speed switch.
It was set to activate stall warnings at higher than normal speeds.
We did find the ref speed switch in the wreckage and it was in the increase position.
This discovery only raises more questions.
We'll be probably be picking up some ice.
The cockpit voice recorder indicates that as Renslow was beginning his descent into Buffalo, he commanded his plane to fly at the normal approach speed.
But what's strange is that with his reference speed switch on, he actually should have been flying faster (SNEEZES) .
.
as this is what the switch would remind him to do.
So why wasn't he? The plane's computer warned the crew to fly faster according to the settings they had configured by displaying a set of red bars on the airspeed indicator.
You may have a better quality of life with regards to These bars are meant to warn the pilots that a stick shaker activation is imminent.
If you're looking at the airspeed indicator, you should be aware that you're getting slow and the stall warning may come on.
Jesus Christ! It seems Renslow and Shaw were caught off-guard.
Still, they could have easily corrected the situation.
Once the stick shaker activated, they could have turned the switch off or they could have put the nose down and increased their airspeed.
It's clear to investigators that Flight 3407 wasn't in danger of stalling when the stick shaker went off.
So now they need to know exactly what happened after the stall warning was activated.
An animated simulation of the crash is constructed .
.
based on information from the flight recorders.
Watch what happens just after the stick shaker goes off.
It illustrates that just after the stick shaker was triggered, the plane suddenly pulled up.
This action dramatically slowed the aircraft and at this point, it did stall.
Essentially, the airplane entered an aerodynamic stall from which it did not recover.
It pitched over and hit the ground.
Investigators are dumbfounded.
Flight 3407 wasn't stalling when the stick shaker went off.
But a few seconds later it was.
We would expect that no flight crew would stall an airliner, so the question is why.
The crew's every action during that brief time now demands careful scrutiny.
What did they do? It's a puzzle.
How could a trained flight crew take a plane that wasn't stalling and in the space of a few moments, make it fall from the sky? The focus of the investigation now switches from the plane to the crew, specifically on the moves they made during the critical seconds after the stall warning sounded.
We wanted to see if the way they flew the airplane was the way they were trained according to the standard operating procedures that are portrayed in their flight manuals.
The flight data recorder retains information from more than 1,000 different aspects of the Q400's flight operations.
From the air speed and altitude to the position of the rudder pedals and throttles .
.
it also records the movements of the most critical flight control, the control column.
Pilots use the control column to change the position of the elevators and ailerons, which manage the direction of the plane.
The flight data recorder stores information not just about the control column's position, but how much force is applied to it as well.
SCOTT WARREN: The FDR records what the control positions were.
It has sensors built into the control column.
It has sensors inbuilt into the control wheel.
What Scott Warren finds when analysing the control column's position is stunning.
In response to the stick shaker, Captain Renslow should have pushed the column forward to bring the nose down and gain speed.
But for some reason, he did the exact opposite.
We found that the crew, instead of pushing forward, which is the normal response to a stick shaker triggering, the crew was actually pulling back on the control.
This had the effect of pulling the nose up, causing the air speed to drop and tipping the aircraft into an actual stall.
Captain Renslow had apparently mishandled one of the most elemental piloting manoeuvres - how to recover from a stall.
Above everything, it requires gaining air speed to get out of the red.
The recovery procedure is fairly simple and straightforward.
It requires pushing forward on the controls and adding full power.
At any point in time, had the captain pushed forward on those flight controls, he had a reasonably good chance of recovering quickly.
From everything we've gained, that stall was recoverable on a repeated number of levels on a repeated basis.
There was no reason for that plane to go down.
Investigators also learned that First Officer Shaw in trying to help Renslow deal with the crisis inadvertently made things worse.
I put the flaps up.
She retracted the flaps, reducing the amount of lift as the plane struggled to stay in the air.
Had the first officer simply called out, "You're stalled.
Advance the power.
Push the nose over," the airplane would have been able to recover.
From a human point of view, it's sad to recognise that those sorts of things happened and the tragedy that came from that.
It's concluded that Captain Renslow's failure to properly respond to the stall warning was the primary cause of the crash of Flight 3407.
As the issue is now pilot error rather than mechanical failure, human performance investigator Evan Byrne is brought onboard.
His first question - why hadn't either Renslow or Shaw noticed that their air speed was too low for the icy conditions? In this case, we can look back towards the fact that there were clear and conspicuous cues of the deteriorating air speed that were not heeded by the captain.
Byrne listens to the cockpit voice recording to try to understand what might have led to that oversight.
SHAW: Exactly.
RENSLOW: Where you could be home with your husband to take care of and all that stuff.
He learns that the crew had been talking throughout the flight.
The conversation continued during the landing approach.
It's a violation of a rule known as the sterile cockpit, which bans non-essential conversation during critical phases of the flight.
You're going to be upgraded in six months.
- Blah, blah, blah.
- (CHUCKLES) Quite simply, it prohibits conversations that aren't related to the operation of the flight.
Let's do a descent checklist, please.
We can do the approach checklist along with it.
Yeah, sure.
Byrne also discovers that, because of the cockpit banter, the crew performed critical checklists and briefings late.
Fuel transfer off.
Hydraulic pressure and quantity Distracted, the crew probably didn't see the red bar indicating they were flying too slowly for the conditions the plane had been configured for.
Approach checklist complete.
Rock'n'roll.
When crews deviate from standard operating procedures and perform checklists late or don't make the required call-outs, they become more vulnerable to subtle mistakes that they may make inadvertently.
That could lead to startle and surprise or unanticipated events that they have to respond to.
The evidence is unequivocal.
The crew of Flight 3407 was badly distracted throughout the approach.
Came in when we interviewed and he said, "Yeah, you're gonna be upgraded in six months.
" They had forgotten a key setting they had made that required them to fly faster than normal.
They had missed indications that they were flying too slow for icing conditions.
Then Captain Renslow had reacted incorrectly to a stall warning, sealing the fate of the plane.
Evan Byrne wonders what could have caused a trained airline flight crew to have made such missteps.
He finds a clue on the cockpit voice recorder.
(YAWNS) Excuse me.
(GROANS) The crew was showing signs of fatigue.
Could Renslow and Shaw have been too tired to function effectively on the flight deck? It's a tough question.
Answering it will require tracking their movements during the 72 hours leading up to the crash.
It's basic gumshoe detective work in the investigation where we're trying to collect as many facts as we can.
Byrne interviews the families of the pilots.
Can I speak to Mrs Renslow, please? He studies the pilot's mobile phone bills and records of text messages.
He searches the airline's computer system to determine if and when the crew used it to check their schedules.
He needs to track their every move.
And what time did Rebecca leave the house? We're talking to colleagues or other pilots, check airmen, instructors, and we ask all those people about the pilots, about their recent activities.
He learns that neither pilot actually lived anywhere near Newark but could not afford to stay in hotels on their salaries.
Captain Renslow was earning 60,000 annually at Colgan Air.
First Officer Shaw was being paid less than 16,000 a year, substantially less than an average bus driver.
As a result, both pilots had made long cross-country commutes to Newark, Captain Renslow from his home in Florida and Rebecca Shaw from Seattle, Washington.
In fact, Shaw had commuted all night from Seattle on a cargo flight that connected through Memphis.
Captain Renslow had spent the night in the airline's crew lounge at Newark Airport after having already worked two days.
He was seen sleeping on a couch in the lounge.
It was against company rules but pilots who couldn't afford housing near the airport did it anyway.
Records show that at 3:10am on the morning of the crash, Renslow was awake.
He checked his work schedule on the airline's computer network.
Next stop Buffalo.
At 7:29am, Rebecca Shaw sent a text message to her husband, telling him she'd arrived safely in Newark.
Hi, honey, it's me.
Phone records indicate that later in the morning, Captain Renslow was on the phone several times.
Rebecca Shaw was noticed having a nap, catching up on the rest she'd lost flying in the night before.
It's not a lot of sleep.
We ultimately concluded that it was likely that both crew members were experiencing some effects of fatigue at the time of the accident.
Her job is to watch the air speed.
Her job was to watch the instrument panel and my view is she was fatigued to the point where it's right here and you're still dull - I think that's where she was, and he just was not capable.
He was just He shouldn't have been flying an airplane.
The revelation leads John Kausner to become a fierce advocate for changes to the industry.
He raises awareness among both lawmakers and the public about the need to improve regulations governing pilots at regional airlines.
This is just saying we support 3407 families, their fight for aviation safety and these representatives of congressmen and senators have done that for us and we're very appreciative.
He takes his fight all the way to Washington, DC.
We needed to do something and so we began to advocate in Washington weekly, every other week.
We made innumerable trips down there and immediately the families just gelled.
We all attended the hearings and began to say, "This is what we can do.
" The crash of Flight 3407 focused the world's attention on a growing safety problem.
The relatives of those who were killed will help eliminate that threat.
The crash of Flight 3407 exposed wide-ranging shortcomings in the regulations that govern regional airlines.
These smaller airlines now make up half of all daily passenger flights in the US.
Their pilots are generally younger, less experienced, earn less and work long hours.
Their levels of safety are way different from the majors.
They have a much lower threshold in training, in ability, in pay, obviously, so they can't attract a higher-qualified pilot.
There are pretty low wages, pretty difficult working conditions and we don't seem to attract the same level of applicant that we used to.
Some regional airlines get into a bind and they have to hire the first people that meet the minimum.
In the US, of the seven fatal accidents involving passenger jets over the past 10 years, five have involved regional airlines.
Those include the crash of Delta connection flight 5191 in August 2006, which killed 49 people when the crew took off from the wrong runway.
John Kausner's campaign to change laws governing pilots has paid off.
We relied on your support.
We needed your support.
You nurtured us.
We want to thank all of you from all of us.
A year and a half after the crash, under pressure from Kausner and other victims' family members, the US Senate passed a bill which toughens training requirements and forces the FAA to draw up new rules on pilot fatigue.
Studies into the problem of overtired pilots are already underway.
At the University of Iowa, researchers are developing a system that could help pilots resist fatigue, stay engaged with the critical task of flying and prevent future tragedies like that of Flight 3407.
Thomas Schnell is a human factors engineer.
We use a number of neurocognitive and physiological sensors that we apply on subjects, pilots that we invite for our studies.
He's studying how pilots stay alert on the flight deck as a test subject conducts a cross-country journey in a simulator.
Using sensors, he can determine how alert and engaged a pilot really is.
The research could lead to the development of fatigue detectors on airplanes.
We are trying to predict a pilot's state so that we can adjust something on the flight deck to prevent the problem from getting worse or starting in the first place.
We monitor brain activity, eye movement activity, heart, the EKG and respiration and other parameters in an effort to figure out what the pilot's or the crew's state is.
Are they fatigued? Overworked? Are they disengaged or distracted? This section is where they were really drowsy and you can see that gaze has become kind of bored.
Schnell instructs the flight simulator to trigger a major system failure in the cockpit.
MAN: It's no good.
Everything's dead.
Hydraulic systems have failed.
Flaps not deployed.
Crank up that heat map so we can see what his brain activity was.
Brace for impact.
(SIRENS AND ALARMS WAIL) When a crew is fatigued, what you might see is their reaction to events may slow down.
So you will see mistakes being made on the flight deck.
It's precisely these kinds of mistakes on the part of Renslow and Shaw that ultimately crashed Flight 3407 .
.
and killed 50 people.
It's a tragedy that should not have happened, that was foreseeable, was preventable and it's repeatable if we don't do something about it.
I'm focused and determined to change what exists and not have another dad sitting here.
Just outside Buffalo, New York.
Gear down.
- Looks alive.
- Gear's down.
- (RATTLING) - Oh! - Oh.
- Jesus Christ.
The airplane entered an aerodynamic stall.
- It did not recover.
- (SCREAMING) (SIRENS WAIL) It was one of the grisliest, nastiest scenes that I think I've seen.
The crash of Continental Connection Flight 3407 would be one of the NTSB's most important in decades and would see a grieving father fight for changes to the laws governing small airlines.
I'm focused and determined to change what exists and not have another dad sitting here.
Continental Connection flight 3407 operated by Colgan Air is en route from Newark, New Jersey, to Buffalo, New York.
It's been a busy flight for captain Marvin Renslow.
He's providing guidance to a new first officer, Rebecca Shaw, a former flight instructor from Seattle.
(SNEEZES) Bless you.
Shaw's only been with the airline for just over a year.
She must decide if she now wants to become a captain.
I don't know what I want to do with the upgrade.
Depends where I'm based.
Well, think of it this way, if you stayed on the Q, obviously you're not making the captain, right? Right.
But you may have a better quality of life, with regards to buying a house, having a Shaw trained to be a first officer on the plane they're flying now - a Canadian-made Bombardier Q400.
It's a twin engine turbo-prop, popular with regional airlines.
The 45 passengers have had a long night.
Their plane was held up for two hours at Newark, a delay considerably longer than their journey.
Flight 3407 is heading north-west over Upstate New York.
The trip is only 53 minutes.
Visibility is poor, and there's a forecast of snow and moderate winds in Buffalo.
Just some water, please.
Ellyce Kausner is a student at Florida Coastal School of Law in Jacksonville.
She's on her way to visit her family in Buffalo.
MAN: She had five nieces and nephews at the time and they had a love luncheon at the kids' school, a couple of the nephews, and they both wanted Aunt Ellie to be their guest.
Any excuse to come home, she came home.
She talked to all of us from Newark.
- Hey.
- Then she called us from the lounge.
Yeah, it's been over two hours.
She was PO'd because of the delay.
Folks, from the flight deck, your First Officer speaking.
It looks like at this time, we're about 10, maybe 15 minutes outside of Buffalo.
Weather in Buffalo is pretty foggy - it's snowing a little bit there.
I'd like to make sure that everyone remains in their seats so that the flight attendants can prepare the cabin for arrival.
Thank you.
Colgan 3407 descend and maintain 2,300.
OK, down to 2,300.
Colgan 3407.
Let's do a descent checklist, please.
We can do the approach checklist along with it.
Sure.
Bug set.
Set.
GPWS landing flaps, selected 15 degrees.
The pilots go through a list of settings for the plane to continue its landing approach.
Transfer off? Yeah.
Passengers can already glimpse the lights of Buffalo's suburbs.
Clarence Center is on the approach path for aircraft landing at Buffalo's airport.
(YAWNS) Gear down.
- Looks alive.
- Gear's down.
Right.
Flaps 15.
Extending the flaps provides more lift, allowing the plane to slow to its final approach speed.
- (ALARM BEEPS) - Er Oh! Oh! Jesus Christ! Suddenly, the control column starts to shake.
The Q400 is slipping out of control.
(SCREAMING) Captain Renslow struggles to keep his plane flying.
(GRUNTS) Come on there! Oh! But it seems to have a mind of its own.
Should gear be up? Gearup! Oh, damn! - Oh, we're down.
- (SCREAMS) Oh! The plane crashes into a house at 6038 Long Street.
At air traffic control in Buffalo, Flight 3407 disappears from radar.
Check Colgan there - see if he's on you.
MAN: (ON RADIO) What's his call sign? Colgan 3407.
- Nope, I don't have him.
- Do you see anything out there? Nothing.
Alright.
Call the fire department.
I'll never forget the dispatcher's words - "Plane crash, house on fire.
" I immediately got my shoes on and my coat on, got in my truck and started heading down the driveway.
Firefighters don't have far to go.
6038 Long Street, the home of the Wielinski family, is less than a block from the Clarence Center station.
As I turned towards the village, the whole village was aglow.
You could see the smoke.
You could see the flame.
Karen Wielinski and her daughter make it out of their house alive.
But her husband, Douglas, is killed.
My initial reaction was there's no .
.
there was no way that somebody made it out.
But the division reported that they had two people and they were taking them to the hospital.
I said, "Well, which firefighters are they?" and they said, "They're not firefighters, chief.
"They're survivors from the house.
" I actually physically had to go to the ambulance and look inside for myself to believe.
Firefighters have never seen a blaze like this.
If you've ever gone to a bonfire and stood five feet from it, and you couldn't take it anymore, that's what it felt like from 100 feet away.
MAN: The height of it was just unbelievable.
Obviously because of the fuel that was probably added to it and the debris area was very large.
It was a very, very horrific sight.
We live in the town where the plane went down right down the road.
My son was driving home from a soccer game and went right by the plane crash and called me up and said, "There's a plane down in Clarence.
" Never really entered our brain that it was Ellie's plane.
Flight 3407 has crashed 8km short of Buffalo airport.
All 49 passengers and crew are dead, including Captain Marvin Renslow and First Officer Rebecca Shaw.
But many more people could easily have been killed.
You picture a house and to see a plane on top of this house, no more house left, no more plane - only the tail section.
You're asking yourself, "How in God's creation did this happen "and not wipe the whole block out?" The next call was again from my son and he said in fact, it was Ellie's plane.
He said, "Dad, there's no survivors" .
.
at which point my wife shrieked and just hit the ground.
Chris said he's never heard his mother make a sound like that and I hope he never does again.
So, that's when I knew she was gone.
That was probably within 45 minutes of the accident.
The scene in the morning is one of utter devastation.
Clint Crookshanks from the NTSB.
Can I start poking around? Clint Crookshanks is one of the first investigators on the case.
When we arrived on scene, there was a fire still burning.
It turns out it was from a gas line that had been broken in the house.
The firemen would put the fire out and it would reignite every couple of minutes.
It was one of the grisliest, nastiest scenes that I think I've seen.
Flight 3407 may have been a small plane but it's the worst crash in the US in more than seven years.
It will become one of the NTSB's most important investigations.
The fire at the crash site of Flight 3407 continues to burn, preventing investigators from examining the wreckage.
That's where I think the box is.
The airplane had crashed into a house and then it had burned all night long, so all the debris was basically in the house.
WOMAN: Our concern is that we're losing evidence.
It's perishable and if we can't get in there and get the fire out, then we're not able to maybe get a hold of evidence that might help us during the investigation.
Clint Crookshanks urgently needs to recover the black box flight recorders, which could contain valuable clues about the accident.
We knew that the recorders were in the tail part of the airplane.
Ordinarily, investigators don't go near a crash site that's still burning.
But if the black boxes can't be rescued, they may never find out what brought down Flight 3407.
We started looking around and poking around into the wreckage and actually found out where they were.
OK, whoa.
Whoa.
It's probably in here, OK? The access panel's on the other side so we're going to have to cut a hole right there.
OK? The fire department produced a chop saw.
We were able to cut a hole in the side of the fuselage .
.
and go in and grab the recorders and pull them out.
Alright, that should do it.
To the immense relief of all, the recorders are undamaged.
Once we took the recorders out of the airplane, we put them on the jet and they were flown back to Washington, DC to our headquarters.
Now investigators are faced with a new hurdle.
What little is left of the aircraft is hopelessly jumbled together with human remains and debris from the house.
CLINT CROOKSHANKS: It all burned and settled into the basement so we had probably 10 feet of debris that we had to dig through in order to recover all of the airplane.
Authorities wonder how they can salvage any useful evidence from this chaos.
They get invaluable assistance from an unusual source.
A group of students learning to process crime scenes is enlisted to separate human remains from the rest of the debris.
They were graduate students from a local college - forensic anthropology students - and this was good experience for them to come dig through wreckage like this and look for human remains.
So, landing gear over there, please.
Thank you.
It's dirty, painstaking work but it frees up investigators to concentrate on the aircraft ruins.
We were on our hands and knees with brooms, with little shovels, scooping out debris - identifying it as to house debris or airplane debris and then putting it in different piles.
The first question for me as a structures engineer is to figure out if the whole airplane made it to the scene of the crash.
The wreckage is carefully studied to determine if the plane's four corners - nose, tail and both wing tips - are present.
If we find all four corners of the airplane, then we know that there was no inflight break-up.
There was nothing that departed the airplane during the flight that may have caused the accident.
One wing has been consumed by fire.
The other is shattered into pieces.
Hey, have a look at this! But investigators are gradually finding what they've been searching for.
Oh, yeah.
That's the last piece.
It's the left one.
It wasn't until several days into the investigation as we were scraping away some of the debris that we actually found evidence of the left wing and the nose.
Investigators now have all the pieces they need to conclude that the entire plane is at the crash site.
Whatever caused the disaster was not the result of a break-up in flight.
The tragedy of Flight 3407 is under intense scrutiny at the NTSB offices in Washington.
Officials there try to determine if some kind of malfunction caused the crash.
On this particular accident, we knew it was a landing accident, so we wanted to check the aircraft's performance and then also we checked for flight control continuity.
Much of this responsibility falls to Scott Warren.
My role in the investigation was to be the systems group chairman for the safety board, so I was in charge of looking at all of the aircraft systems.
Warren analyses the aircraft's cockpit voice recorder, or CVR, to determine if there were any indications of a problem in the cockpit.
REBECCA SHAW: .
.
prepare the cabin for arrival.
Thank you.
Is that ice on the windshield? He discovers that six minutes before the crash, the crew of Flight 3407 had noticed a build-up of ice on the aircraft.
Got it on my side.
On yours? Oh, yeah.
Oh, it's lots of ice.
Ice can be a deadly threat to any airplane.
If an aircraft has ice on it, it will have more drag on it, so it will require more power to maintain a given airspeed.
That's the most I've seen on the leading edges for a long time.
Until a year ago, Rebecca Shaw had no experience with ice on an aircraft.
I've never seen icing conditions.
Never de-iced.
Never experienced any of that.
When ice accretes on a wing, it adds weight to the airplane but most importantly, it changes the shape of the wing, and, of course, it's the shape, the curved shape of the wing that actually creates the lift.
So by changing the lift characteristics of the airplane, it makes it less able to fly.
You know, I'd have freaked out.
Like I'd have seen this much ice and thought, "Oh, my gosh" - we were gonna crash.
Observing that there was ice on the airplane was an important thing to do, but then it would be important to turn around and verify that your anti-ice and de-icing equipment was on and there was no indication on the voice recorder that they actually re-checked.
Flaps 15.
The CVR reveals that only minutes after the crew detected ice, a device called the stick shaker went off.
It was a warning that the plane was about to stall - literally to fall out of the sky.
Shortly afterwards, the aircraft did precisely that.
(SCREAMING) We know there were icing conditions.
The only question was were they bad enough to induce this airplane to have a failure? Ice is now the chief suspect in the crash of Flight 3407.
If it's ice, let's prove it.
We were all convinced that we had an icing accident.
We said, "Yeah, it's going to be an icing accident.
"We've just got to make sure "we look for these few key aspects of icing.
"Verify that and we're good.
" But some of the victims' family members are less convinced.
Initial thought was that it was an ice-related incident - it was an icy night that brought the plane down.
And as we began to talk to people who knew airline travel, who were pilots themselves, they would look kind of askance and say, "I don't think so.
"That doesn't make sense.
" Some investigators share that suspicion.
Scott Warren knows a plane like the Q400 has a sophisticated de-icing system.
It's designed to keep ice from building up on the wings and other critical parts of the aircraft.
To prevent ice accumulating, the plane has rubber bladders along the front of the wings, called 'de-icing boots'.
A series of valves uses air from the engines to inflate the boots and crack the ice off the wing.
Those boots are designed to inflate periodically, and that inflation breaks off the ice that's accumulated on those leading edges.
Warren now wonders if the crew of Flight 3407 actually turned on their de-icing equipment.
He studies data from the plane's other black box, its flight data recorder, or FDR.
It tracks the workings of crucial aircraft systems, including the de-icing mechanism.
We know from the FDR data that the de-icing system had been selected on by the crew and it was on during the majority of the flight and certainly at the end of the flight, it was recorded in the on position.
But now Warren needs to figure out if the device was actually working.
Just because the data records that the system is on, that's a start, but you can't necessarily believe that one piece of information.
The only way to know is to find what remains of the plane's de-icing system and determine if it was active when the aircraft went down.
Clint Crookshank's team hunts for a crucial component of the system - the valves.
Excuse me.
OK, everyone, I just want For the de-ice valves, we said, "Here's what it looks like.
Look for something like this.
"If we can find those valves "then we can test them and see if they're operating correctly.
" The valves were very important to us because they are the key component that moves air from the engines where the bleed air originates for the de-icing system.
OK, good.
Five more to go.
Five of the plane's six de-icing valves are eventually recovered.
Some were badly burned, some were in fairly good shape.
And we took those valves and we conducted as much testing on those valves as we thought was appropriate for the level of damage.
We looked at the level of electrical conductivity and we looked at the pressure testing.
We looked at a wide variety of things, depending on the condition of the valve itself.
And as far as we could tell, all those valves were working properly.
- Is that ice on the windshield? - I've got it on my side.
Now investigators have to consider the possibility Oh.
yeah.
Oh, it's lots of ice.
.
.
that there was so much ice on the wings that the de-icing system couldn't get rid of it fast enough.
Oh, yeah.
That's the most I've seen on the leading edges in a long time.
It has happened before.
In Roselawn, Indiana, a French-built commuter plane crashed in 1994 after suffering a catastrophic build-up of ice on the wing.
By analysing how much power was needed for the plane to maintain its airspeed, investigators reach a surprising conclusion.
Basically, the aircraft was flying as if it had a relatively small amount of ice on it, but a very manageable amount of ice.
Looks normal.
It was not overloaded.
It was not excessive.
If ice didn't bring down the plane, something else must have caused it to stall.
Warren studies the Q400's operating manual to learn the plane's stall speed - the velocity at which it can no longer generate enough lift to stay in the air.
He discovers that in the conditions Flight 3407 was flying through, it's roughly 111 knots.
He now compares the plane's stall speed with its actual airspeed just before the stall warning went off.
131 knots - well above the danger zone.
Flaps 15.
When the stick shaker went off - Er Oh! - Jesus Christ! .
.
they were not necessarily at the edge of a stall.
They were still 20 knots or so away from the stall.
This new revelation deepens the mystery of Flight 3407.
If it wasn't stalling, why did it fall out of the sky? Investigators turn their attention to the critical seconds before Flight 3407 went out of control.
They look for clues that could explain why the stall warning went off when the aircraft was flying well within its safety margins.
They discover that this plane has a unique feature known as a reference speed switch.
It governs the sensitivity of the plane's stall warning.
ROGER COX: Very few airplanes in my experience have such a switch.
This airplane is the only one I know that has an actual switch on the overhead panel.
It was designed by the manufacturer to be an extra safety feature.
Some kind of variable ref speed? Pilots are supposed to turn on the reference speed switch when they are going to be flying through icing conditions.
And we'll probably be picking up some ice.
When in the increase position, it reminds pilots to fly faster to counteract any drag effect ice will have on the aircraft.
When you are in icing conditions and ice does accrue on the wing, it can cause the stall speed to go up, and so this ref speed switch correspondingly causes the warning to come on sooner or at a higher speed.
What that switch does is it basically changes the trigger settings for the stick shaker.
So we had to ask the manufacturer, "How does this switch work?" What we found was it was part of the system's description that the crews got when they went through training but they didn't get a lot of training on how to handle that switch.
It seemed like it was too simple to worry about.
Investigators need to know if the crew of Flight 3407 had turned on the reference speed switch .
.
triggering the stick shaker at a faster than normal speed.
The flight data recorder doesn't show whether the switch was on.
Investigators must find another way of determining its position at the time of the accident.
(PHONE RINGS) Clint here.
Clint Crookshanks is given a new priority Alright, I'll see if we have it.
.
.
recover the ice protection panel from the Q400's cockpit, where the reference speed switch is housed.
Since the panel was in the cockpit, finding it is a challenge.
Most of the front end of the airplane was consumed by fire, so we didn't find anything except for little balls of molten aluminium little wire bundles and a lot of ash.
But after an extensive search, Crookshanks discovers that the ice protection panel is one of the few pieces of the cockpit that survived.
Bingo.
However, the knobs and switches are barely recognisable.
Crookshanks examines the charred panel to check the position of the reference speed switch.
It was set to activate stall warnings at higher than normal speeds.
We did find the ref speed switch in the wreckage and it was in the increase position.
This discovery only raises more questions.
We'll be probably be picking up some ice.
The cockpit voice recorder indicates that as Renslow was beginning his descent into Buffalo, he commanded his plane to fly at the normal approach speed.
But what's strange is that with his reference speed switch on, he actually should have been flying faster (SNEEZES) .
.
as this is what the switch would remind him to do.
So why wasn't he? The plane's computer warned the crew to fly faster according to the settings they had configured by displaying a set of red bars on the airspeed indicator.
You may have a better quality of life with regards to These bars are meant to warn the pilots that a stick shaker activation is imminent.
If you're looking at the airspeed indicator, you should be aware that you're getting slow and the stall warning may come on.
Jesus Christ! It seems Renslow and Shaw were caught off-guard.
Still, they could have easily corrected the situation.
Once the stick shaker activated, they could have turned the switch off or they could have put the nose down and increased their airspeed.
It's clear to investigators that Flight 3407 wasn't in danger of stalling when the stick shaker went off.
So now they need to know exactly what happened after the stall warning was activated.
An animated simulation of the crash is constructed .
.
based on information from the flight recorders.
Watch what happens just after the stick shaker goes off.
It illustrates that just after the stick shaker was triggered, the plane suddenly pulled up.
This action dramatically slowed the aircraft and at this point, it did stall.
Essentially, the airplane entered an aerodynamic stall from which it did not recover.
It pitched over and hit the ground.
Investigators are dumbfounded.
Flight 3407 wasn't stalling when the stick shaker went off.
But a few seconds later it was.
We would expect that no flight crew would stall an airliner, so the question is why.
The crew's every action during that brief time now demands careful scrutiny.
What did they do? It's a puzzle.
How could a trained flight crew take a plane that wasn't stalling and in the space of a few moments, make it fall from the sky? The focus of the investigation now switches from the plane to the crew, specifically on the moves they made during the critical seconds after the stall warning sounded.
We wanted to see if the way they flew the airplane was the way they were trained according to the standard operating procedures that are portrayed in their flight manuals.
The flight data recorder retains information from more than 1,000 different aspects of the Q400's flight operations.
From the air speed and altitude to the position of the rudder pedals and throttles .
.
it also records the movements of the most critical flight control, the control column.
Pilots use the control column to change the position of the elevators and ailerons, which manage the direction of the plane.
The flight data recorder stores information not just about the control column's position, but how much force is applied to it as well.
SCOTT WARREN: The FDR records what the control positions were.
It has sensors built into the control column.
It has sensors inbuilt into the control wheel.
What Scott Warren finds when analysing the control column's position is stunning.
In response to the stick shaker, Captain Renslow should have pushed the column forward to bring the nose down and gain speed.
But for some reason, he did the exact opposite.
We found that the crew, instead of pushing forward, which is the normal response to a stick shaker triggering, the crew was actually pulling back on the control.
This had the effect of pulling the nose up, causing the air speed to drop and tipping the aircraft into an actual stall.
Captain Renslow had apparently mishandled one of the most elemental piloting manoeuvres - how to recover from a stall.
Above everything, it requires gaining air speed to get out of the red.
The recovery procedure is fairly simple and straightforward.
It requires pushing forward on the controls and adding full power.
At any point in time, had the captain pushed forward on those flight controls, he had a reasonably good chance of recovering quickly.
From everything we've gained, that stall was recoverable on a repeated number of levels on a repeated basis.
There was no reason for that plane to go down.
Investigators also learned that First Officer Shaw in trying to help Renslow deal with the crisis inadvertently made things worse.
I put the flaps up.
She retracted the flaps, reducing the amount of lift as the plane struggled to stay in the air.
Had the first officer simply called out, "You're stalled.
Advance the power.
Push the nose over," the airplane would have been able to recover.
From a human point of view, it's sad to recognise that those sorts of things happened and the tragedy that came from that.
It's concluded that Captain Renslow's failure to properly respond to the stall warning was the primary cause of the crash of Flight 3407.
As the issue is now pilot error rather than mechanical failure, human performance investigator Evan Byrne is brought onboard.
His first question - why hadn't either Renslow or Shaw noticed that their air speed was too low for the icy conditions? In this case, we can look back towards the fact that there were clear and conspicuous cues of the deteriorating air speed that were not heeded by the captain.
Byrne listens to the cockpit voice recording to try to understand what might have led to that oversight.
SHAW: Exactly.
RENSLOW: Where you could be home with your husband to take care of and all that stuff.
He learns that the crew had been talking throughout the flight.
The conversation continued during the landing approach.
It's a violation of a rule known as the sterile cockpit, which bans non-essential conversation during critical phases of the flight.
You're going to be upgraded in six months.
- Blah, blah, blah.
- (CHUCKLES) Quite simply, it prohibits conversations that aren't related to the operation of the flight.
Let's do a descent checklist, please.
We can do the approach checklist along with it.
Yeah, sure.
Byrne also discovers that, because of the cockpit banter, the crew performed critical checklists and briefings late.
Fuel transfer off.
Hydraulic pressure and quantity Distracted, the crew probably didn't see the red bar indicating they were flying too slowly for the conditions the plane had been configured for.
Approach checklist complete.
Rock'n'roll.
When crews deviate from standard operating procedures and perform checklists late or don't make the required call-outs, they become more vulnerable to subtle mistakes that they may make inadvertently.
That could lead to startle and surprise or unanticipated events that they have to respond to.
The evidence is unequivocal.
The crew of Flight 3407 was badly distracted throughout the approach.
Came in when we interviewed and he said, "Yeah, you're gonna be upgraded in six months.
" They had forgotten a key setting they had made that required them to fly faster than normal.
They had missed indications that they were flying too slow for icing conditions.
Then Captain Renslow had reacted incorrectly to a stall warning, sealing the fate of the plane.
Evan Byrne wonders what could have caused a trained airline flight crew to have made such missteps.
He finds a clue on the cockpit voice recorder.
(YAWNS) Excuse me.
(GROANS) The crew was showing signs of fatigue.
Could Renslow and Shaw have been too tired to function effectively on the flight deck? It's a tough question.
Answering it will require tracking their movements during the 72 hours leading up to the crash.
It's basic gumshoe detective work in the investigation where we're trying to collect as many facts as we can.
Byrne interviews the families of the pilots.
Can I speak to Mrs Renslow, please? He studies the pilot's mobile phone bills and records of text messages.
He searches the airline's computer system to determine if and when the crew used it to check their schedules.
He needs to track their every move.
And what time did Rebecca leave the house? We're talking to colleagues or other pilots, check airmen, instructors, and we ask all those people about the pilots, about their recent activities.
He learns that neither pilot actually lived anywhere near Newark but could not afford to stay in hotels on their salaries.
Captain Renslow was earning 60,000 annually at Colgan Air.
First Officer Shaw was being paid less than 16,000 a year, substantially less than an average bus driver.
As a result, both pilots had made long cross-country commutes to Newark, Captain Renslow from his home in Florida and Rebecca Shaw from Seattle, Washington.
In fact, Shaw had commuted all night from Seattle on a cargo flight that connected through Memphis.
Captain Renslow had spent the night in the airline's crew lounge at Newark Airport after having already worked two days.
He was seen sleeping on a couch in the lounge.
It was against company rules but pilots who couldn't afford housing near the airport did it anyway.
Records show that at 3:10am on the morning of the crash, Renslow was awake.
He checked his work schedule on the airline's computer network.
Next stop Buffalo.
At 7:29am, Rebecca Shaw sent a text message to her husband, telling him she'd arrived safely in Newark.
Hi, honey, it's me.
Phone records indicate that later in the morning, Captain Renslow was on the phone several times.
Rebecca Shaw was noticed having a nap, catching up on the rest she'd lost flying in the night before.
It's not a lot of sleep.
We ultimately concluded that it was likely that both crew members were experiencing some effects of fatigue at the time of the accident.
Her job is to watch the air speed.
Her job was to watch the instrument panel and my view is she was fatigued to the point where it's right here and you're still dull - I think that's where she was, and he just was not capable.
He was just He shouldn't have been flying an airplane.
The revelation leads John Kausner to become a fierce advocate for changes to the industry.
He raises awareness among both lawmakers and the public about the need to improve regulations governing pilots at regional airlines.
This is just saying we support 3407 families, their fight for aviation safety and these representatives of congressmen and senators have done that for us and we're very appreciative.
He takes his fight all the way to Washington, DC.
We needed to do something and so we began to advocate in Washington weekly, every other week.
We made innumerable trips down there and immediately the families just gelled.
We all attended the hearings and began to say, "This is what we can do.
" The crash of Flight 3407 focused the world's attention on a growing safety problem.
The relatives of those who were killed will help eliminate that threat.
The crash of Flight 3407 exposed wide-ranging shortcomings in the regulations that govern regional airlines.
These smaller airlines now make up half of all daily passenger flights in the US.
Their pilots are generally younger, less experienced, earn less and work long hours.
Their levels of safety are way different from the majors.
They have a much lower threshold in training, in ability, in pay, obviously, so they can't attract a higher-qualified pilot.
There are pretty low wages, pretty difficult working conditions and we don't seem to attract the same level of applicant that we used to.
Some regional airlines get into a bind and they have to hire the first people that meet the minimum.
In the US, of the seven fatal accidents involving passenger jets over the past 10 years, five have involved regional airlines.
Those include the crash of Delta connection flight 5191 in August 2006, which killed 49 people when the crew took off from the wrong runway.
John Kausner's campaign to change laws governing pilots has paid off.
We relied on your support.
We needed your support.
You nurtured us.
We want to thank all of you from all of us.
A year and a half after the crash, under pressure from Kausner and other victims' family members, the US Senate passed a bill which toughens training requirements and forces the FAA to draw up new rules on pilot fatigue.
Studies into the problem of overtired pilots are already underway.
At the University of Iowa, researchers are developing a system that could help pilots resist fatigue, stay engaged with the critical task of flying and prevent future tragedies like that of Flight 3407.
Thomas Schnell is a human factors engineer.
We use a number of neurocognitive and physiological sensors that we apply on subjects, pilots that we invite for our studies.
He's studying how pilots stay alert on the flight deck as a test subject conducts a cross-country journey in a simulator.
Using sensors, he can determine how alert and engaged a pilot really is.
The research could lead to the development of fatigue detectors on airplanes.
We are trying to predict a pilot's state so that we can adjust something on the flight deck to prevent the problem from getting worse or starting in the first place.
We monitor brain activity, eye movement activity, heart, the EKG and respiration and other parameters in an effort to figure out what the pilot's or the crew's state is.
Are they fatigued? Overworked? Are they disengaged or distracted? This section is where they were really drowsy and you can see that gaze has become kind of bored.
Schnell instructs the flight simulator to trigger a major system failure in the cockpit.
MAN: It's no good.
Everything's dead.
Hydraulic systems have failed.
Flaps not deployed.
Crank up that heat map so we can see what his brain activity was.
Brace for impact.
(SIRENS AND ALARMS WAIL) When a crew is fatigued, what you might see is their reaction to events may slow down.
So you will see mistakes being made on the flight deck.
It's precisely these kinds of mistakes on the part of Renslow and Shaw that ultimately crashed Flight 3407 .
.
and killed 50 people.
It's a tragedy that should not have happened, that was foreseeable, was preventable and it's repeatable if we don't do something about it.
I'm focused and determined to change what exists and not have another dad sitting here.