Controlled Pod Into Terrain episode 1: Air Asiana 214

Re-upload of first episode!

this one is about the time a 777 booped the sea wall at SFO.

featuring J, the engineering one, Ariadne, the aviation business expert, and Kyra, who writes about plane crashes

Episode Transcript

(00:00):
Hello and welcome to Controlled Pod Into Terrain.

(00:03):
We are a multimedia podcast about air and space mishaps and putting them into a broader
context of how and why things went wrong.
Now, some of you are here because you're a fan of Admiral Cloudberg's articles and we
want to let you know that this is not going to be a companion podcast to those articles,
but we do want to make sure that we capture the spirit that she writes with.
We are a podcast with slides, but we're a different podcast with slides and any similarity

(00:27):
to other podcasts with slides is purely of a non-actionable kind.
To introduce myself and my co-hosts, my name is Ariadne and my pronouns are they and them.
Jay?
I'm Jay.
I do system stuff, communication, software and electronics.
I'm the one with the fancy accent and my pronouns are they and them.
And last but not least, I'm Kyra Dempsey, better known as Admiral Cloudberg, the aviation

(00:52):
writer specializing in accident breakdowns and my pronouns are she and her.
All right.
Today we are going to be talking about ASEAN Airlines 214, otherwise known as that one
time the plane hit the sea wall.
That has gone very poorly.
That has gone very poorly, yes.

(01:13):
First, there's not supposed to be a hole in the top of the plane.
No.
Normally.
No, no.
Generally planes only do that when they're very distressed.
All right.
Shall we kick this off?
Yeah.
Let's do some sort of news thing.
Okay.
That's going to be the official name, isn't it?

(01:34):
Some sort of news thing?
Yeah, I think so.
It sounds good to me.
And given the sort of level of professionalism on this podcast, we wanted to have an image
for our news that matched.
Yes.
This is Stable Diffusion's idea of a news show about planes.

(01:55):
And so we're going to do some news about planes.
All right.
All right.
The first thing we're going to talk about today is Alaska Airlines.
Now this was flight 1288.
This was leaving from Seattle, obviously Alaska's home base, and going to Santa Ana Orange County.
This is John Wayne Airport for those of you in the LA area.

(02:15):
This is the Rich People Airport or the Disneyland Airport, depending on how you look at it and
whether or not you have your own driver.
Yeah, it's totally obvious when an airline called Alaska has its home base in Seattle.
That's like completely intuitive and normal.
It is very much so.
Yeah.
Their presence in Alaska is actually more limited than you would think.

(02:37):
Also, so this particular plane was trying to land on the very rare occasion that SoCal
had some weather.
It had some weather that completely caught it by surprise because of course SoCal doesn't
have weather normally.
It just has hot and then it has damp.

(03:00):
But it never really has this, which was a tropical storm.
And it seems to have caused some problems for this particular plane.
Another victim of Hillary.
Another victim of Hillary.
Yeah.
So this obviously dropped a wing.
Her kill list is always expanding.

(03:24):
Obviously the wing seems to have dipped on landing, pushing the landing gear up through
the fuel tank.
Now, obviously this was very, very lucky that there was no sort of fuel leakage, but I think
it's pretty clear that this aircraft is almost certainly never going to fly again.
Yeah, it looks, I think the technical term for it is fucked.
Yeah, it's very fucked.

(03:45):
Yeah, yeah, this aircraft, I think, I assume this has been towed off the runway and a ramp
and it will probably be tent it or towed into a hanger and dismantled in place.
Yeah, I don't know how you actually go about moving a plane where one of the landing gears
has gone through the wingspar.

(04:08):
I mean, it's not like you can just jack it up.
No, I'm pretty sure that's exactly what they do.
That's literally what you are.
There are designated jacking points under every aircraft.
Right, but I would imagine that one of those would be the mounting point for the main gear,
right?
Which is presumably not structural anymore on this particular plane.

(04:33):
Well, I would think that if you're not necessarily looking to care about the structural integrity,
you could probably mount it some here on this wing box, which looks to be relatively undamaged.
Obviously, you're not going to do that on a functional aircraft because you would do
serious damage to it.
But if the aircraft is already totaled, and I think your top priority is just getting
it the heck off the runway, you do what you have to.

(04:56):
I mean, as for the cause of this, we really don't know because it just happened.
But what we do know is that people who are saying, oh, the pilot simply should not have
tried to land are barking up the wrong tree because the weather conditions reported for
Santa Ana Airport at that time were fully within the Alaska company limits in terms
of wind speed and direction and all of those sorts of things.

(05:19):
So there's no reason to believe they shouldn't have attempted this landing.
But we don't really know what went wrong.
I think they just got unlucky.
And as anyone who's ever played Kerbal Space Program can tell you, actually, there's two
ways to remove that plane from the runway.
One is jacking it up and putting a dolly underneath it.

(05:42):
The other is just to go to full power and try and fly it.
I think that would be a valid attempt.
I think it would leave more debris than the airport management would be acceptable.
Quite possibly.
Okay.
All right.
I think our next news story is another L for Moscow.

(06:04):
Oh boy.
So first we have the smashing success that is Shandriyan-3.
This was IASRO, so the Indian Space Research Organization, launched this and landed it.
It was their second moon attempt.
They did it for $75 million.
And a lot of people are very fondly pointing out that this is little less than half the

(06:27):
cost of Interstellar.
And I would like to point out that it is also even lower than the much lower budget and
much worse at ASRA from 2019.
$75 million.
How many Xboxes can you buy with that?
I would say, yeah, it would be a few thousand.
I would say $75 million would buy you all of the Xboxes in at least one Best Buy, but

(06:52):
probably not two of them.
Now on the other hand, we have Luna 25.
This one came from the storied and historic Roscosmos, proud pioneers in space.
Obviously the first man in space, the first man to walk outside the first space station.

(07:16):
First woman in space.
First woman in space.
Yeah, by decades, in fact.
Yep.
But like much else about Russia, it has seen better days.
It has.
Yes.
But you could call them the runners up in the moon race.
Twice.
Twice.
Exactly.
Yeah, these guys cannot stop coming second to the moon.

(07:37):
Obviously they did make it to the moon.
Now as you can see over here, this is an impact point.
So this, if you were playing Kerbal Space Program, is when you would press F5 to revert
to orbit.
Sadly, you can't do that in real life as it turns out.
I think maybe that was their plan.
Just didn't know that.

(07:58):
So there is a bit of a conspiracy theory going on about Luna 25.
Now the official story out of Roscosmos is that the vehicle made it into lunar orbit
and was in process of entering its braking phase, where it fires retrograde to circularize
its orbit and sort of get out of a free return trajectory.

(08:19):
And the story out of Roscosmos is that it fired its engine for too long and landed on
the surface with quite a bit more speed than it was expecting to.
Litho braking.
Litho braking.
Now the fact that this is an official story is kind of sad because what that means is
this is the best excuse they could come up with, was just pure incompetence.

(08:41):
But it turns out it's very likely it was much worse than that.
Roscosmos?
Wait, okay, how?
Explain.
Because what happens is that Roscosmos publishes the radio frequencies that they'll be using
for each mission.
Basically this is like their deep space network only much older and much, much worse.

(09:03):
Now a lot of radio amateur radio astronomers saw the vehicle have a good trans lunar injection
and then never heard from it again.
So what it looks like is that the forget upper stage was in communication, was tracking correctly
and fired itself towards the moon on an impact trajectory.

(09:25):
And then nothing ever happened.
So it looks like the probe may have been dead either in orbit or may have even been dead
on the launch pad.
That would be embarrassing.
It never woke up and it hit the moon at much higher than the five meters per second squared
target.
Seems like one of your launch pad checks would be is the probe working?
Yeah, I mean just to send it to you.

(09:46):
Is it turned on?
Also I gotta say I love how just by putting pictures of the two probes with flags on them
you've somehow made the Chandrayaan probe look like Swoll Doge and the Luna 25 is clearly
Cheems just from the positioning.
We have accidentally made a Cheems meme.
And of course obviously up in here you have Zelensky who is relentless in taking down

(10:08):
his targets.
Yeah, they actually hit this with a cardboard drone, knocked it straight into the moon.
Yeah, yeah, a lot of people don't know that.
It's interesting, we're taking a look at that.
Alright, and our last story is that the ALPA, so this is the Airline Pilots Association,
this is effectively the union for all pilots in the United States, has come out with a

(10:28):
statement saying that they will support a two-pilot cockpit forever.
And they're completely correct.
Yeah, they're completely correct, I think.
We have absolutely no objection to that.
Yep, okay, alright.
No, oh and I want to explain why for some people who are freaking out about having one
pilot in the plane is, so to have, we have two pilots because if something happens to

(10:51):
one pilot it's still totally possible to fly the plane safely with just that one pilot.
But if you have only one pilot to begin with then you have to prove that you can fly, that
the plane can fly itself with no pilots without any decrease in level of safety.
And if you've ever seen a self-driving car, which I know there are plenty of San Franciscans
in this podcast crew, I know we have, the technology is a long way away.

(11:16):
Right, and I think that we're going to talk about fly-by-wire systems a lot in this podcast,
a lot, and when necessary I think we'll bring in experts from, whether they be software
developers on the programming side or actual pilots who fly these aircraft.
But a self-flying plane would be a totally different beast than fly-by-wire, which is

(11:38):
a control system.
Yeah, I think people have a sort of a false idea, I think, of how automated modern aircraft
are.
Yeah, they're not as, they're perhaps not as automated as you think.
Well, I mean, this particular one that you chose the picture of the flight deck here,
this is an Airbus A380, it's the absolute acme in automation, right up until the point

(12:04):
where one of its Trent 900 engines explodes and suddenly the pilots get extremely busy
trying not to crash the thing.
Yeah, you've got to be glad there were two pilots there and not one.
Yes, seriously.
Yeah, I think that's the other thing to mention, is that it's not just the aircraft's ability
to fly itself that is the issue at hand, it's the problem-solving and cockpit resource management,

(12:29):
crew resource management I should say, that comes from having two people in the cockpit,
and I don't think that's even possible to ever replace.
Yeah, no, so I don't see the FAA ever approving single pilot operations for airliners.
It's a pipe dream by managerial people who don't understand airplanes.

(12:52):
I think the very closest you will probably ever see is short haul cargo flights, I think,
of an hour or two, the types that go to places like Wells, Maine, or rural North Dakota to
service them.
I think those may one day go to…
Things like that are already single pilot.
If you're flying a Metroliner or something, those are approved for single pilot operations

(13:12):
with cargo.
Yeah, most of your caravan flights are all going to be single pilot, especially if you're
flying somewhere in the bush.
But yeah, you're not going to have one pilot in your A380 anytime soon, thankfully.
Thankfully.
Okay, let's talk about ASEAN Airlines.
Is it any good?
And on one particular day, it was not a great airline.

(13:33):
What about the rest of the time, you know, when they're not crashing?
Well, I'll say that there are travel and frequent flyer podcasts.
We are not one of them, so I won't comment.
I know that they're a fairly well-done airline.
I think they get decent reviews, but I think we're going to talk about one time when maybe
they weren't the best airline in the world.
Yeah, 6 out of 10.
The flight was totally good, but the landing was a little hard.

(13:56):
Yeah, still better than Ryanair.
Okay, so on this screen, we have a helicopter, headphones, hospital, a car, an apartment
building and a theme park.
Now Jay, would you mind taking a guess at what all of these items have in common?
Ooh, is it that they're all made by the same family-owned conglomerate?

(14:18):
They are indeed, yeah.
This is the one, this is a small portion of the products that are made by the Samsung
Crime Syndicate.
I'm sorry, the Samsung Family Corporation.
AKA ChaiBall.
AKA ChaiBall.
Now Jay, can you explain to us a bit about why ChaiBall is and why six companies control
one of the largest economies in the world?
It's kind of strange.

(14:39):
So South Korea has a relatively short past as a democratic market economy.
It's only really been that since actually the early 90s.
And before that, it has been various different degrees of...

(15:00):
It's been various different degrees of...
Dictatorship.
Dictatorship.
You've had a series of authoritarian military coups that have replaced each other, but none
of them have really...
Were really effectual at enacting any sort of liberal reform until the very early 90s.
Right, so the very early 90s, they started reforming their economy a little bit, but

(15:25):
the problem was that by that point, so much of the economic activity in South Korea had
been sort of...
Well, if you look at the situation that currently exists in Russia, for example, where you have
these oligarchs who own everything, and then the government has to sort of suck up to them

(15:45):
because they control so much of the economic and industrial output of the country that
if any government was to try and go against them, then it would be really, really in big
trouble.
So you end up with this situation where there's a very few of these very large companies or

(16:08):
groups of very closely related companies that are all owned by these same few families.
It's different from a kai-retsu that you have in Japan because those are generally sort of
conglomerates of businesses that have...
Many of them have existed for hundreds of years.

(16:33):
Chai balls tend to be newer than that.
Korea doesn't really have such a long history of empire and that kind of thing as Japan does,
but what it did have is this very quick sort of land grab for these highly placed families

(16:57):
to control a huge amount of the resources of the country.
And it is quite a rich sort of modern country in terms of the businesses and the industries
that it supports, but in many ways on a government level, it's still actually kind of a feudal

(17:19):
state.
Someone who is an employee of one of these chai balls might spend their entire life working
for the same company or the same overarching business management thing.
And you won't just work there.
I mean, you'll be born in a hospital owned by Samsung.

(17:39):
You will probably live, grow up in an apartment block built by Samsung.
You will go to university sponsored by Samsung, after which you will go work for Samsung.
On vacation, you'll go to the theme park.
On your days off, you have a Samsung partner.
This is sort of a very real, by and large from wallet type model, where you have this

(18:07):
is the future that people think Amazon is going to do one day, just sort of own every
aspect.
But it's happened in Korea.
They've been operating this way for almost 30 years now.
And you have some very weird operations where you will have Jay, is it LG that makes plasma

(18:27):
television or I'm sorry, OLED televisions and cancer drugs in the same division?
Yeah, that's LG cam.
LG cam makes lithium ion batteries, cancer drugs, imaging supplies for hospitals, and
also the chemicals that LG display uses for making OLED TVs.

(18:51):
Yeah, they're actually the same, not just part of the same chai ball, but actually the
same business unit.
Right.
So the reason that all of this is relevant to what we're talking about is that Arexiana
was part of one of these conglomerates.

(19:11):
And these conglomerates are operated by people that do not fundamentally understand the industries
that they are running.
So the decisions they make are almost entirely driven by an Excel spreadsheet.
Right.
So they'll hire people who have experience in day to day running of an airline, but the
executive decisions about how is this airline going to expand?

(19:32):
What are its priorities going to be?
Those are being made by some guy.
Yeah, and they might be not just some guy, but some fail son.
Right, exactly.
Yeah, I think if you watch Succession, just imagine that.
Just imagine the scene where Kieran Culkin doesn't understand how a rocket launch works

(19:54):
and ends up blowing one up on a launch pad in Japan, and you kind of get an idea of how
a chai ball works and how a fail son might end up putting an airplane in multiple pieces
on a runway.
Not that any of these pilots were, to my knowledge, failed chai ball sons.
Well, no, we should say in a very legally correct way that we actually don't know anything

(20:20):
about the people who run the Kumho Group.
The Kumho Group was the chai ball that wanted to compete with Korean Air, KAL, which was
owned by Hanjin Transportation.
So they decided one day that they were going to have an airline and buy a bunch of planes.

(20:42):
They bought pretty good ones.
The Boeing 777, everyone agrees it's a pretty good plane.
And they just decided to set up.
And because these chai balls control such vast amounts of money, they were able to actually
get to the point where they were regularly operating flights because they didn't really

(21:03):
have to borrow anything to do this because they controlled a major proportion of the
Korean economy.
Yeah, and I should point out that before Asiana Airlines came along, Korean Air was the only
South Korean airline.
It was South Korea's aeroflot, you might even say.

(21:25):
So was it state-owned?
No, no, it wasn't.
It was owned by Hanjin Transportation, which is a different tribal.
Yeah, which also owned the state, right?
So then, no.
But the idea is there was no competition before Asiana Airlines.
Right, right.
This didn't last though because Hanjin Transportation ended up buying Asiana in 2020.

(21:49):
Yeah, so now there's really only one airline again.
It just pretends that there's two.
Do they not, does South Korea not have like a budget carrier?
They probably do.
I just can't name any off the top of my head.
Yeah, no, you're absolutely right.
But Asiana and Korean Air are their main international presences.
Yeah, and they're now one brand.

(22:09):
They actually do have a couple of budget airlines.
We're going to talk about some of them later.
How are we?
Yeah, I believe it's the Seoul to Busan route.
I think it may be the busiest air route in the world.
Is it not a domestic route within Japan anymore?
Yeah, I think it used to be Tokyo-Saporo.

(22:30):
I think that's still very much in the top three or five.
We could look it up, but we're not that kind of podcast that does research live on the
air.
Are we?
It could be.
All right, so let's talk about this plane.
And we're going to talk about this specific airframe because 742 was the one that broke
into pieces.
Yes, so the plane that you're seeing on your screen does not exist anymore.

(22:56):
And furthermore, until this, there had never been a Boeing 777 accident.
They first entered service in the 1990s and they lasted over 15 years without a fatal
accident in fact.
Without even a hull loss accident.
Well, no, there was British Airways Flight 38 in 2008.
It was a hull loss, but there were no fatalities.

(23:19):
And that was very distinctly not the air.
That was not the 777's fault.
That was the fault of the engine manufacturer.
Yes.
Specifically, I believe it was a fuel filter, but we're not going to talk about British
Airways today.
So basically, the 777 was the last good Boeing plane.
Okay, well we can fight about that.
But I mean, I think it's, yeah.

(23:40):
It's subjective.
No, I'm sorry, it was.
We're going to do an episode one day just on the 787.
So we're going to talk about the 777.
We don't have much to say about it.
It's just a really, really good airplane.
This was the first full fly-by-wire aircraft Boeing ever built.
It was famously the first aircraft that was ever designed 100% on a computer.
They still brag about that.

(24:01):
I don't know why, but they brag about it.
If you have ever flown internationally, there is a pretty good chance you flew on one of
these.
But not this one, because this one's 777.
I flew on one domestically, actually.
But yeah, it's really a basic airplane.
Wide body, two aisles, two engines, two wings, a tail.
It's not really that special.

(24:21):
It's like when you think of an airplane, it probably looks something like the 777.
Yeah, exactly.
You know, Boeing...
The 777 is a very large plane for a twin jet, though.
Yeah, it looks normal in photos.
Then you'll see it next to other planes, and it's just grossly disproportionately large.
But it looks normal.
Yeah, at what point, I believe, maybe we'll edit in a photo where you see that the engine

(24:46):
on a 777 is larger around than the entire body on a 737 that you flew to Pensacola.
Well, not this 777.
No, not this 777.
But newer, the newest one.
Yeah.
Obviously, the engines on this one got much, much smaller.
Hmm.
I think they got more detached, I think.

(25:07):
We will find.
Yeah.
So this is...
It's a deceptively long aircraft.
It's a deceptively wide aircraft.
The new 777X1000, wherever, if it ends or ever ends up coming out, will actually probably
have a fairly similar seating capacity to the old 747, which is what people think of
as a huge airplane.
So Kyra, where was this plane coming from?

(25:27):
And where was it going to?
Asiana Airlines flight 214 was a regularly scheduled flight from Sol Gimpo International
Airport to San Francisco International Airport in San Francisco, California, across the Pacific.
There were... how many people were on board this plane?
It was a lot.
Yeah, they had 291 passengers on board.

(25:48):
So it was not actually quite full, but it was fairly close.
And they had a cockpit crew of four, including two primary pilots and two relief pilots.
So these pilots, they take shifts, right?
Usually three and a half or four hours.
Yes.
So the primary captain was an instructor, Li Zhengmin, 49 years old.

(26:12):
And he was... and during the takeoff and landing, he was seated in the right seat, the first
officer seat, because he was fulfilling the role of a check and instructor captain on
the flight for a trainee captain who was 45 year old Lee Kang Kuk, who sat... who was
the primary... who flew from the left seat on takeoff and landing while receiving his

(26:35):
initial operating experience, which is he had just transferred to the Boeing 777 from
the Airbus A320.
And he was being checked out for the captain... he was going straight from captain on the
A320 to captain on the 777.
So he had just finished training and he was doing his initial operating experience where
he flies regular line flights with real passengers under the supervision of an instructor who

(26:58):
replaces the first officer.
In addition to these two, there were also the two relief crew.
There was 52 year old relief captain, Li Zhengju, and 40 year old relief first officer, Bong
Donghwan.
So those two flew during the middle of the flight.
And then during the approach, first officer... relief first officer Bong Donghwan actually
would come back into the cockpit to act as a third set of eyes because the right seat

(27:25):
instructor captain, Li Zhengmin, is... he has dual duties.
He's the pilot monitoring, but he's also instructing.
And that can be very distracting.
So it's helpful to have that third set of eyes there to catch things that he may miss
because he is busy monitoring the trainee captain's performance extra closely.

(27:47):
In his case in particular, it was even more distracting on the grounds that this was actually
the first time he'd flown as an instructor.
Yes.
He had never done this before.
He had just finished being checked out as an instructor and he had never instructed
a trainee on an actual line flight before.
And that's going to become important later.

(28:07):
Yeah.
So, but these guys were not necessarily... some of them were new to the airframe, but
they were not necessarily new to being pilots.
These are guys with thousands of hours of flying time, presumably.
Right, Kyra?
Yeah.
Yeah.
So even the trainee captain, Li... Li Li Kangkuk is his name.
He, he had, he had thousands of hours.
I can't tell you how much exactly...

(28:30):
Oh, he had over 9,000 flying hours, but he had only 43 on the Boeing 777.
So he was about as new as you can get.
Almost all his hours were on the Airbus A320.
Yeah.
Which is a much smaller plane.
It's also an entirely different kind of aircraft, right?
So...
Right.

(28:50):
The difference between a Boeing is a massive differential because you are going to a completely
new fly-by-wire system, a whole new control system.
We will also at some point when we talk about flight... Airbus crashes and fly-by-wire,
we'll have to talk about the different sort of levels of protection that Airbus offers
in a way that Boeing does not.
But basically Airbus has a totally different philosophy of control of the airplane than

(29:14):
Boeing does.
Yeah.
And that's what, that's what's really important.
You don't need to know the details yet, but we may get into some of them.
Yeah.
All right.
So, Jay, what are we looking at?
This is the West Plan.
By West Plan, I mean that this is the complicated set of aircraft movements that makes the three

(29:43):
major airports in the San Francisco Bay Area.
You can see them there, SFO, Oakland, and San Jose.
They're all within 50 miles of each other.
And as a consequence, movements of planes in and around the bay can get really quite
complicated.
Yeah.

(30:04):
So it would be a lot simpler if you could just, you know, come in and land at SFO from
the West and, you know, totally not have to think about Oakland and San Jose to the East,
right?
You can't because there's a line of 800 foot hills immediately west of SFO.
So it's not practical to land from the West.

(30:24):
You have to come in either from the Southeast or the Northeast.
So either directly from San Jose or directly from Oakland are where those runways point
back to.
And so in fact, most traffic has to, especially traffic, traffic coming from the West to SFO
has to loop around the airport, dodging the Oakland and San Jose approach corridors, and

(30:44):
then come back around and land on one of the two parallel runways from the Southeast.
This gets even worse if you have to do a go around because if there is any amount of traffic
and there is always traffic around SFO, it's a very busy airport with kind of crappy runways.

(31:07):
If you have to do a go around, you're going to get delay vectors and you're going to have
to go all the way around the pattern to make another attempt at landing.
It could be half an hour, 45 minutes before you get your chance at landing again.
Also not to be sure of this map, there are two other, there are two major US Air Force
bases that would be within the boundaries of this map and Moffett Field and Ames Research

(31:32):
Center, which is a NASA facility in a civilian airport about three and a half feet outside
of San Jose.
This is an extremely busy air corridor with very, very narrow parameters on what operations
can be done.
Don't forget the general aviation airfield in Palo Alto.
There's one there as well.

(31:53):
Yeah, so this is one of the factors that makes approaching San Francisco Airport challenging.
One of the things that SFO is infamous for is when you're flying to SFO, you get vectored
in hot and high, which basically means you are left to descend a little more steeply than

(32:17):
normal because when you loop around to pick up the approach path into SFO from the southeast,
you have to stay above the planes that are approaching Oakland and then you have to drop
quickly and so you have to lose a lot of speed and altitude simultaneously and that can be
potentially tricky, which is something that is going to become important.

(32:42):
The approach is also over water, which can make it difficult, especially if you're, you
know, maybe in your period of circadian low as you might be if you've just flown across
the Pacific.
You know, you try and avoid this, that's why they have this five hour break in the middle
with the relief captain and relief first officer flying the plane, but still you don't necessarily

(33:09):
sleep that well and you don't necessarily feel all that refreshed while you're doing
it and that can make it very difficult to judge the distance because, you know, water
is kind of featureless.
The other thing is that the western side of the city, of the actually the whole San Francisco

(33:30):
peninsula is quite often covered in fog and that can make it very difficult to orient
yourself while you're approaching it from the other side because you can't really see
the city.
You can't see the big sort of landmarks like the Golden Gate when you're approaching it
from the other side.
Yeah, so quite often SFO airport is in the clear and they're doing visual approaches,

(33:54):
but like everything from immediately west of the airport out is completely socked in,
it's zero visibility.
To make matters worse, on the day we're talking about, which is the 6th of June 2013, the
instrument landing system on a runway 28 left and 28 right, which is where planes were landing
that day, was partially inoperative.

(34:16):
So they had been recently expanding those runways, which meant they had to remove and
relocate the instrument landing system equipment.
And they had, by the time that this takes place, they had only put back in the localizer,
which provides lateral guidance to align with the runway automatically.
They had not put back in the glide slope, which provides vertical guidance.

(34:40):
So when you're normally flying an instrument landing system approach, the autopilot can
lock on to the glide slope and the localizer and follow this perfect, roughly 3 degree
descent path straight to the runway with relatively minimal help.
But if you don't have the glide slope, then you can align with the runway, but the pilot
has to manually achieve that 3 degree flight path angle.

(35:06):
And that can be tricky.
It's something that is really basic to flying and every pilot is sort of expected to be
able to do, but it's not intuitively easy.
And you may not have done it quite recently because a lot of airlines actually mandate
that you use as much of this automation as possible.

(35:29):
Or at least they did at that time.
ASEAN airlines did.
They said all automation, you know, automation must be used to the fullest extent.
So trainee captain Lee Kang-Kook, he had never actually flown an approach in the real Boeing
777 without a glide slope before.
He had not never done it.
He had maybe done it in a simulator, but that was it.

(35:52):
What gets even better is that the two runways, 28 left and 28 right at SFO are actually too
close together.
They're less than 300 feet centerline to centerline.
And so as a result of that, simultaneous operations of these two runways are only allowed if the

(36:14):
full ILS is operating, which means that traffic was even more backed up than usual that day.
Because San Francisco is famous for doing those simultaneous approaches where it looks
like the two planes are in formation flying all the way down to the ground.
And that helps, you know, obviously that being able to do that near doubles the capacity

(36:34):
in terms of airplanes per hour that can be landed at the airport.
So this was, you know, there was some pressure to not screw this up because you would have
to go way back in line, even worse than usual.
And of course the go around is very spicy because of those 800 foot high hills.

(36:55):
And if you don't have perfect guidance, you might end up buzzing a Daly City or San Bruno,
as has in fact happened in the past.
Yes, I used to live in Daly City and I could see those planes coming fairly low overhead
all the time.

(37:15):
But usually just after takeoff, I never saw any super low on a go around or anything like
that.
Should we talk about the autopilot a little bit?
So let's get, let's set the stage.
So it's the middle of the day, it's sunny and except for the fog that is covering over
San Francisco.

(37:36):
So Lee Kang-Kook, the trainee captain is in the left seat.
Instructor Captain Lee Jung-Min is in the right seat and really First Officer Bong Dong-Won
is in the observer seat.
So they come around, they are discussing, you know, what San Francisco landmarks can
we see?
And this is Lee Kang-Kook's first time in San Francisco as well.

(37:57):
So he's like, is that the Golden Gate Bridge?
But he's pointing at the Bay Bridge and Lee Jung-Min is like, no.
They are vectored in around to the southeast of the airport for what they expect to be
a visual approach to runway 28 left using the localizer but no glide slope.
And

(38:18):
If you remember the West Plan slide from earlier, it's the Teal Line.
Yes.
So, yeah, that one.
So they get to, they get to about 6,000 feet and they're starting to line up with the localizer
and so far so good.

(38:40):
Now once they are aligned with the localizer, they need to maintain that three degree glide
slope.
So how do you do this?
And this is a thing called energy management, which is basically you keep keeping the plane
on the glide path you want and at the speed you want and doing both of those things simultaneously

(39:00):
is what makes it so tricky.
When you descend normally, you are turning potential energy into kinetic energy, which
will cause your speed to increase.
But you have just come from cruising speed and you want to get down to landing speed.
So you have to both descend and decelerate.
You have to time all of your configuration changes in order to progressively increase

(39:23):
drag on the airplane faster than the descent attempts to cause your speed to increase.
So there's this constant balancing act that goes on.
And so the way that this works on the 777 is the pilot uses the mode control panel or

(39:43):
MCP, which is shown here, to select autopilot and autothrottle modes that help them achieve
the best combination of pitch, angle and thrust to both maintain speed and maintain the glide
path at the in the desired configuration.

(40:06):
So basically you can modify both speed and trajectory of an airplane using either thrust
or pitch.
So you can pitch up to go up or you can pitch down to go down.
But more commonly you can pitch up to speed to slow down or pitch down to speed up.
And then you can also use thrust to accelerate to go up or decelerate to go down.

(40:30):
Or you can use or you can increase thrust to increase speed or decrease thrust to decrease
speed.
So these are at speed and pitch, each of them can be used to control either.
So thrust or pitch each can be used to control either speed or.
The other thing is that your pitch will change when you change your engine power level as

(40:55):
well because the engines are not.
They thrust off axis.
In line with the center of mass.
Yeah, they thrust off axis.
So when you when you throttle up, the plane is going to sit up a little bit more.
And when you throttle down, it's it's going to go the other way a little bit.
Right.
So what you want to do is use an autopilot mode that best fits.

(41:22):
Does that profile you're going for?
The desired trajectory.
Right.
Right.
So we're going to discuss a couple of the autopilot modes and the autothrottle modes
that couple with them.
So vertical speed mode is an important autopilot mode on approach.
So in this mode, the autopilot pitches the plane up or down in order to achieve a climb
or descent rate selected by the crew in the MCP, for example, minus 1000 feet per minute.

(41:48):
And so if the.
So basically that's that's pitch controls vertical rate.
Right.
And when the autopilot is in vertical speed mode, normally the autothrottle, which can
be automatic engine thrust control, will switch to speed mode.

(42:12):
And in speed mode, the autothrottle increases or decreases engine thrust in order to achieve
an airspeed selected by the crew in the MCP.
So again, vertical speed and speed modes pitch modifies the descent or climb rate engine
thrust modifies the airspeed.
And then.

(42:32):
But you can also do the exact opposite of that with flight level change speed mode.
And this is a mode in which the autopilot can will pitch the plane up or down in order
to achieve an airspeed selected by the crew and the MCP.
Well, again, the exact opposite, the autothrottle will go into thrust mode in which it will
increase thrust to gain altitude or decrease thrust to lose altitude.

(42:57):
And so in flight level change, change speed mode is primarily used for exactly what it
says, changing flight level.
So if you're at like 20000 feet and you want to go up to 25000 feet, you can enter 25000
feet in the mode control panel, select flight level change speed mode and the airplane will

(43:18):
climb will the autothrottle will accelerate to climb to that altitude while the autopilot
modifies the pitch of the airplane to maintain the same airspeed.
And the vertical the vertical rate is not directly controlled.
That's what if you want to directly control the vertical rate, that's what you use vertical
speed mode for.

(43:40):
So so thinking about thinking about this, the vertical speed is one leg of a right triangle
and the horizontal speed is the other leg.
And you want the hypotenuse to be at three degrees.
Yeah.
And it sounds like what you're doing with this panel is kind of the way you described
is there there is a way to tell the aircraft what you want it to do.

(44:02):
Go up or go down, go up or you know, and another that tells you how to do it.
So either with presumably the elevator trim or the throttle.
So here's what I want you to do.
Here's how I want you to do it.
Right.
So is it you basically have four four four squares on a planet square.
Yep.
So the yeah, the pilot will typically do a few things during this approach.

(44:24):
They will select vertical modes and they will enter airspeeds and altitudes using the airspeed
and altitude windows in the mode control panel.
And the autothrottle will typically change mode automatically the pilot and during a
normal approach, the pilot won't be manually switching the autothrottle or they shouldn't

(44:45):
be because the autothrottle will change modes automatically to complement the autopilot
mode autopilot and vertical mode that you have selected.
So then sorry to interrupt, but will these work on any sort of auxiliary systems, meaning
will the aircraft deploy its own flaps or or the spoilers?
No, it will not.

(45:05):
You have to do that.
OK.
So then lastly, last but not least, there's another autothrottle mode that's not thrust
or speed, and that's called hold mode.
And this is basically this is this just means the autothrottle motor is disconnected from
the thrust levers.
It can't make any inputs.
So unlike on an Airbus on a Boeing, the way that the autothrottle works is it has this

(45:26):
motor that physically moves the thrust levers for you.
And then that modifies the engine thrust.
But in hold mode, that motor is disconnected from the thrust, so it just can't do anything.
And the way that hold mode you can get into hold mode is there are basically two main
ways.
One of them is if the pilot overrides the autothrottle and moves the thrust levers manually,

(45:51):
the autothrottle will enter hold mode because then it's clear the pilot wants to make inputs
themselves and not the autothrottle.
And the other way is if the vertical mode is flight level change speed, thrust levers
are determining whether you go up or down, right?

(46:11):
If you're in so if you're in flight level change speed and the selected flight level
is so far below your current flight level that the autothrottle pulls the thrust levers
all the way back to idle, flight idle, which is the minimum flight thrust position in flight,
then it will also enter hold mode and it will just leave the thrust levers there at flight
idle until some new mode is selected.

(46:34):
Or until the MCP selected altitude is reached.
So if you're in flight level change speed and you've selected an altitude 5,000 feet
below you, the autothrottle will pull the thrust lever back to hold mode and the plane
will just descend in an open descent until it approaches 5,000 feet, then it will wake
back up and the autothrottle will come out of hold mode and accelerate the thrust levers

(46:58):
to level you off.
Okay, I think we understand that.
But surely the pilots weren't just told, hey, land on this runway.
There has to be a procedure for doing one of these one of these landings without a glide
slope, right?
So yeah, they're flying a visual approach, which is is literally just you.

(47:23):
I see the runway.
It's over there.
You see the runway and then you try to descend at the appropriate, along the appropriate
with the appropriate flight path angle, again, about three degrees.
And it's like you just do that you have a great you have fairly good leeway in terms
of how, although there's usually a correct answer.

(47:45):
So like, you know, but we're going to get into that what the pilots actually did versus
what they probably should have done.
So again, let's go back to the flight itself right now.
We'll pick them up as they are descending through 5300 feet.
And they have just come down off that hot and high approach because they're going over

(48:06):
the top of the approach path into Oakland.
So they're in flight level change speed mode to descend to their cleared flight level.
And the the autothrottle has pulled the thrust levers all the way back to idle to make that
descent work.
So the autothrottle is in hold mode, and the pilots have selected a target speed in the
mode control panel of 212 knots and the autopilot is maintaining this by pitching up or down

(48:32):
to modify their speed.
So so it's at this moment that they are told by the air traffic control, hey, could you
reduce airspeed to 180 knots, just, you know, keep your place in the approach pattern without,
you know, overrunning the aircraft in front of you.

(48:55):
So this is a pretty routine request.
And it wasn't it wasn't an extreme request or anything like that.
So immediately, the trainee captain Lee Kang-kook, he reaches up to the mode control panel and
he dials in 180 knots as the target airspeed.
So so to slow the plane down, they're in flight level change speed mode to slow the plane.

(49:18):
The autopilot is going to do what it's going to pitch the nose up.
So it does.
And immediately, this causes them to begin ballooning above that three degree glide path.
And this is a problem because they had they had discussed, you know, okay, we want to
cross the waypoint do yet at a distance of 5.4 nautical miles from the runway at an altitude

(49:43):
of 1800 feet as shown in the slide.
And you know, now they have this handy feature on their navigation display, which shows them
exactly what trajectory they have versus the ideal three degree glide path.
And it's clearly showing, hey, you're going to go above it.
You're actually going to go you're going to go way above it.
So what do you do in this situation?
The correct answer is you need more drag on the airplane because the thrust levers are

(50:07):
already idle.
You can't reduce thrust any more than they are than it already is.
And if you want to so if you want to decelerate to 180 knots without going above the glide
slope, you need to increase drag on the airplane by deploying the speed brakes.
But they don't do this.
In fact, the pilots are not going to deploy the speed brakes at any point during this

(50:27):
approach, which is very interesting.
You know, I've had pilots ask me, you know, why does this keep happening?
Why does it seem like in some of these crashes, pilots just don't ever does never seem to
occur to them to deploy the speed brakes.
And we really don't know.
But they did not.
Now, Kyra, can we can we talk about what it means to get behind the aircraft?
Right.

(50:48):
So so we've kind of touched on energy management, right.
So you have you have potential energy if you have altitude, you have kinetic energy if
you have forward speed.
And we're sort of constantly balancing that.
Obviously, jet engines, they don't they don't respond immediately.
Now, some of you drive gasoline cars.
Some of you have electric cars, which are even more responsive.
But you're sort of used to you push the gas pedal and it moves immediately.

(51:11):
Jet engines do not do this from the time you give them a command to the time they are providing
the amount of thrust you've requested can be a few seconds.
And that's going to come relevant very late in this flight.
Very late.
Not relevant yet.
Exactly.
Yeah.
But but we're going to talk about something called getting behind the aircraft.
And obviously, as Kyra is going through this, this is an energy management equation, right.
You need to bleed off energy in and turn that into friction and heat.

(51:35):
And you need to descend the aircraft.
And you sort of always need to be planning ahead because you sort of every every action
has an equalizer direction.
If you speed up, you will gain altitude.
And as Kyra is walking you through the ways that they're not shedding speed, they're starting
to get behind the aircraft, meaning that they are things are happening faster than they
are able to keep up.

(51:58):
Right.
And this begins to happen once, you know, you're off course and now you have to play
catch up to try to get back on course.
But things are continuing to happen.
Listen to at the same rate that they normally do with this added task on top.
So so the confusion is going to begin to mount as we go through here.
So they have begun to deviate above the glide path.

(52:20):
And because the autopilot has pitched up to slow them down and they can fix this by deploying
the speed brakes, which increase drag on the wings, but they don't do that.
Instead, the trainee captain Lee Kang Cook switches the vertical mode of the autopilot
to vertical speed mode.
And in this so if you recall in this mode, the autopilot will pitch down to achieve the

(52:45):
descent rate that he has selected in the in the mode control panel and the auto throttle
will either reduce or increase thrust to achieve the MCP selected airspeed.
So this isn't going to solve the problem because the okay the autopilot is going to begin pitching
down to, you know, achieve the descent rate he has selected, which is minus a thousand

(53:09):
feet per minute.
But the auto throttle has already held pinned the thrust levers at idle, which is the lowest
they can go.
So it can't reduce thrust any further.
So because they're now pitching down more, the airspeed is going to increase.
So the airspeed never reduces to the selected MCP speed of 180 knots, it gets down to about

(53:31):
185 knots and then it starts going up, which is not what they want.
And are they are they discussing in the cockpit, right?
So are we using CRM?
Are we are we trying to figure out a problem?
No, not not really.
There's not really.
There's basically a discussion that's like, you know, somebody says, hey, you're high.
And Lee Kang Cook goes, oh, I'm high.

(53:52):
Okay, I will descend more.
You know, that's basically that's the extent of the CRM that's going on here.
Okay.
And is it do we think what's the culture inside the cockpit?
Because I know we are we have a rookie captain on an entirely new airframe that he's very
inexperienced with, with a Czech airman, right?
So this is a person that he knows is is grading him.

(54:13):
It's not it's effectively holding a clipboard, right?
Yeah.
You know, and he so he's he's obviously nervous about, you know, can he do it right?
But then the instructor is also himself uncertain of what his his own role is, because it's
his first time instructing.
And so, you know, he's like, how much you know, he's probably thinking, how much guidance
do I give?
Do I give him?

(54:33):
Do I tell him, no, you're doing it wrong.
You should do it this way.
You know, he doesn't really give very much guidance.
He's sort of watching as the trainee captain Lee Kang, who, you know, makes these mistakes.
And it's not totally clear whether the extent to which he's aware of them, but he's probably
aware to some degree that, you know, he's not doing things right, but he's not saying
anything.
He's just quietly grading.

(54:55):
And so anyway, they begin trying to extend the the flaps and landing gear because these
both also substantially increase drag because their speed is now increasing with the thrust
levers at idle.
You know, that's a problem because, in fact, as you, you know, as the plane is descending

(55:19):
and now its speed is increasing away from the target speed, it's now it's going to
overshoot the altitude by the, you know, planned altitude at the waypoint do yet the final
approach fix, but even more than before.
Because if you're going faster, you need to have a higher descent rate in order to achieve
the three degree glide path, just because, you know, you're covering more ground and

(55:42):
less time.
So you need to descend at a faster rate, obviously.
Now, so, okay, we can.
So you have to increase drag, the increased drag by deploying the landing gear and beginning
to deploy the flaps.
And you know, to this isn't enough even by itself.
So Lee Cancuga also increases the vertical speed to minus 1500 feet per minute.

(56:04):
This is immediately after he says, okay, I will descend more.
So now they actually start descending at a pretty good clip.
They are descending toward the, the glide path.
You know, they're quite high above it, but they are coming, you know, the trend line
is good and their speed is decreasing.

(56:26):
So it's like, okay, we're back on track, right? But this is where Lee Jung Min, the instructor
captain suddenly comes in and he says, you know, change your vertical speed back to minus
1000 feet per minute.
Because he thinks, you know, we're getting pretty close to the three degree glide path.

(56:48):
We should not be descending this quickly.
And it's not entirely clear to me why he did this so early, but it was too early to do
this.
They had not reached the optimal glide path yet.
And in fact, when Lee Cancuga put the MCP vertical speed back to 1000 feet per minute,

(57:08):
minus 1000 feet per minute, they started buoying up above the glide, the optimal glide path
again, which was predictable.
Yeah, the official accident report doesn't actually go into any real detail behind why,
why did he make this call at that moment?
But I think he may have seen, you know, okay, we're finally in a relatively high drag condition,

(57:31):
you know, but we need to get on with extending the flaps more and the flaps are airspeed
limited.
So, you know, each notch on the flaps that you extend, you have to be below a certain
airspeed or else you will, you will overspeed the flaps, you'll potentially even damage
them.
So at one point here, they're decelerating and the trainee captain Lee Cancuga calls for

(57:55):
more flaps and Lee Jung Min says, you know, I can't give you that because we're flying
too fast.
So his decision to decrease the vertical speed back up to minus 1000 from minus 1500 may
have been because he wanted to make sure they bled off speed faster, airspeed faster in

(58:20):
order to extend the flaps on time.
Because obviously if you're descending at a higher vertical speed, your airspeed is
also going to increase because you're trading more potential energy from more kinetic energy.
So he did this and he was able to extend the flaps more, but now they're too hot, they're
on track to be too high again.
They're still going to overshoot their runway.

(58:42):
And so what happens, so what happens next?
They cross the do-yet waypoint at 2250 feet instead of 1800 feet like they wanted.
So they're 450 feet too high at the final approach fix.
At this point, if you're 450 feet too high at the beginning of the final approach phase-
Call a go around.
Yeah, you should, you should go around.

(59:03):
I mean, you can in some, it is physically possible to salvage an approach after that,
but like you shouldn't, you know, it's, it's not always possible.
And sometimes, you know, you just, you end up digging a bigger hole for yourself.
Yeah, this is, this is an aircraft at this point probably weighs 350,000 pounds, right?

(59:27):
So this is a lot of kinetic energy that this is, and it doesn't, it has a lot of inertia
and it does not respond well to, to sudden asks.
Yeah.
So also, so around this point, as they reach the, the final approach fix at do yet, the,
they call the, they decide, oh, it's, they realize, oh, it's time to set the MCP target

(59:51):
altitude in case of, to the go around altitude in case we have to go around later.
And this is standard procedure on every approach.
Ominous.
Yes, you set the, you set the target go around altitude, which for this approach into SFO
was 3000 feet.
You set that in the MCP window.
And at this point in the approach, as in fact, they are the vertical mode of the autopilot

(01:00:14):
should be vertical speed with a negative descent rate selected.
So, and if you're in vertical speed and descending and you select an MCP target altitude, that's
above your current altitude, it's not going to do anything because the, in vertical speed
mode, the vertical speed takes priority.
So if it's not going to, the autopilot isn't suddenly going to ignore the fact that you

(01:00:36):
want a negative vertical speed in order to climb to the selected MCP altitude.
It's just, it's just going to ignore the MCP altitude.
You're going to keep descending, which is in fact what they wanted it to do.
And that's why you can set the go around altitude in advance.
And the purpose behind doing that is so that if you suddenly have to go around, you can
select go around mode and the autopilot will automatically climb the plane to 3000 feet,

(01:01:01):
the go around altitude without you having to even think about it.
And so, you know, everything is fine with that, right?
But we're about to get back to that in like a few seconds here.
So they're still above the glide path.
The trainee captain calls for flaps 30 to, you know, you progressively increase the flap

(01:01:25):
angle as you slow down and right.
So you can maintain lower, lower speed flight and increase drag and bleed off your speed.
Right.
But the instructor captain says, no, I can't give you flaps 30.
We're too fast still.
So at this point, the training captain Lee Cancuk, he needs to do something about the

(01:01:46):
situation.
So what does he do?
He decides he gets the incredibly genius idea.
I'm going to select flight level change speed mode.
And why?
And we still have not.
For the love of God, why?
And nobody has found the spoiler handle yet.
No, nobody has touched the spoiler handle.
In fact, no, nobody touches the spoiler handle this entire, this entire flight.

(01:02:10):
So what he really needed to do right now, if he was going to try to salvage this approach,
which he shouldn't have, but he could have maybe if he was good, was deploy the speed
brakes, increase drag on the plane, bleed off that speed.
And then you can get flaps 30 and that'll increase drag even more.
And it'll be this nice positive feedback loop that gets you slowed down nicely to where
you want to be.

(01:02:30):
Well, also descending to where you want to be.
But he doesn't.
He selects flight level change speed mode.
And this is just a terrible idea.
And we don't really know why he did this.
Like in interviews after the accident, he said, well, I thought, you know, if I selected
flight level change speed mode, then the, the auto throttle would decrease thrust more.

(01:02:51):
But that's obviously impossible because thrust lever, the thrust levers were at flight idle.
They couldn't go any lower.
They'd been at flight idle this entire time.
So that, that logic never made any sense to me.
So we don't, we don't really know why he did this, but he selected flight level change
speed mode and obviously what does that do?

(01:03:12):
You know, the autopilot and flight level change mode goes, okay, what is the MCP target altitude?
Oh, it's 3000 feet.
That's above current altitude.
Therefore we must climb.
That's the wrong direction.
You're going, you're going the wrong way.
Yes.
Yes.
So he pitches, pitches the nose up to, to climb to 3000 feet.

(01:03:37):
It's slightly ironic that if he just let it do it and did a go around, everything would
have been fine.
Yeah.
We're going to use this phrase a lot on this podcast, but this would have been the perfect
time to just call for a go around.
And listeners, a lot of airlines, in fact, I would say the majority of airlines at this
point have what's called a no questions asked, go around policy, which means that if the
captain of an, of, of, of an aircraft or the pilot in command calls for a go around, that's

(01:04:02):
the end of the discussion.
The chief pilot will not talk to them.
They will not, there's not going to be a write up.
Nothing.
Nobody asks questions.
You just do the go around.
You don't even have to report it afterwards.
Yeah.
So so the, the autopilot, yeah.
Right.
So as soon as he selects flight level change speed mode, the autopilot pitches up to, to
slow the plane and the auto throttle increases thrust to climb to the selected go around altitude

(01:04:26):
of 3000 feet.
And this is not what he wants at all.
So within seconds, training captain Lee Cancuk disconnects the autopilot, pitches down manually
and then grabs the thrust levers and moves them to, to back to flight idle where they
were before.
So as soon as he does this, the plane actually begins to descend fairly rapidly because he

(01:04:51):
is pitching down more than the autopilot was and he's got the thrust levers back to idle
and now, and they're in a fairly high drag configuration now with all these flaps.
So they begin to descend at quite a, quite a rapid clip, which is what they want.
And I just, again, so, so this is something that we just keep saying is they could have

(01:05:13):
still salvaged it at this moment.
Right.
So if they had said, again, they shouldn't have tried, right about here was possible.
If these guys had said, Hey, we're going to try and salvage this.
We're in high drag configuration.
Start putting this, start pulling everything back, pull the flaps back in, increase the
thrust.
They could have salvaged this.
It was not, this was not a sure thing yet.
Yeah, right.

(01:05:34):
Again, they shouldn't have tried, but they could have, they could have done it.
But that would have required that they be ahead of the aircraft.
But they had missed one key thing that just happened.
So do you remember what one of the conditions for the autothrottle entering hold mode was?
That's when the pilot overrides the autothrottle.
It will go into hold mode, disconnecting the autothrottle motor from the thrust levers.

(01:05:59):
So as soon as Lee Kang-Hook pulled the, manually pulled the thrust levers back to idle, the
autothrottle went into hold mode.
And there are two ways to get it out of hold mode.
One is you select a new, or there are three ways.
One is you select a new autopilot vertical mode.
So if he switched to, you know, if he switched the autopilot back on, he had disconnected

(01:06:20):
it and changed it to vertical speed or whatever, the autothrottle would have come out of hold
mode.
Alternatively, it would have come out of hold mode if they had an MCP target altitude set
below their current altitude and the autopilot engaged, but they did not have that.
And the other way would be to change the auto flight overall system status.

(01:06:46):
So by that I mean, is the autopilot on or is it off or are the flight directors on?
And the flight directors are the overlay, are overlays on the pilot's displays, which
tell them basically where to fly, you know, up, down, left, right.
And it's, they're meant to make it easier to manually fly, hold, to manually hold an

(01:07:06):
aircraft trajectory.
And so when you switch the autopilot off on the, on the 777 like this, it defaults to
flight director mode.
So this happened before the autothrottle went into hold mode though.
So now they needed to change one way to get the autothrottle out of hold mode was to change
the auto flight status from flight director to something else, perhaps off.

(01:07:29):
And actually it was standard procedure at Asiana at this point in the approach to turn
the flight directors off because they're no longer needed once you have visual contact
with the runway.
And so the instructor captain Lee Jung Min turns off Lee Kang Cook's flight director
because he asked him to, but he doesn't turn off his own flight director.

(01:07:51):
So the, the, the system does not register a system status change.
It's still in flight director mode.
The autothrottle stays in hold mode.
And actually he was supposed to have turned off both of their flight directors, which
if he had done this, the autothrottle would have woken up from hold mode, but he didn't
and it did not.
So they, so now they're in it, they're descending with the autothrottle in hold mode, which

(01:08:13):
means the, the thrust levers are locked at flight idle unless the pilot moves them manually
and they will not move from that mode, that position automatically unless one of the aforementioned
things happens.
So, so now they have the MCP target airspeed.
They have selected it at this point, they select the target airspeed as the landing

(01:08:38):
reference speed plus five knots.
So which is known as V ref plus five.
And that's basically the speed you want to have right up until you flare for a touchdown.
And in this case, they had calculated that to be 137 knots.
So Lee Kang Cook put 137 knots into the MCP and he fully expected that as they decelerated
toward that speed, the autothrottle would kick in and maintain that speed.

(01:09:02):
But now because it was in hold mode, it was not going to do that.
The autothrottle can't do anything in hold mode, but he's not aware of this.
And so now they're in a situation where their speed is decreasing because they are now in
a very high drag configuration with all their flaps out and landing gear and everything
else and their airspeed is decreasing and it's not going to stop at V ref plus five

(01:09:28):
unless the pilot actively intervenes.
But he thinks that it will stop there automatically because he has not noticed that the autothrottle
has gone into hold mode.
In fact, neither of the pilots has noticed this.
The instructor captain called out the change in autopilot status from autopilot to flight
director, but he seemed to be distracted by that and did not notice the autothrottle mode

(01:09:53):
change to hold mode.
What about the guy in the jump seat?
So he didn't notice this either.
No, he's watching and he's looking at the situation.
Is he just playing Angry Birds?
And actually at this point his focus is on the vertical speed because he's seeing, hey,
we're getting relatively close to that three degree glide path and we're descending at

(01:10:13):
a pretty high rate.
They were descending at over minus 1800 feet per minute at this point because they're high
drag and thrust levers are idle and the trainee captain is pitching down.
So he actually says, hey, you should not be descending so quickly.

(01:10:36):
He says actually, sync rate, sir.
But as this is happening, trainee captain Lee Kang-Kook is looking at the precision approach
path indicator lights on the runway, which tell him whether he's too high or too low.
So basically there's a set of four lights which can be either white or red.
If two of them are white and two of them are red, great, you're on the three degree glide

(01:10:59):
path.
If they're all red, you're too low.
If they're all white, you're too high.
Three white and one red, you're slightly too high.
Three red and one white, you're slightly too low.
So he's looking at this and he's seeing, we are actually, we are very high right now.
He was seeing four white lights.
And so he really wanted to keep descending more.

(01:11:23):
So the relief pilot observing is repeatedly calling out sync rate.
You're descending too fast.
So now he's really in a bind.
Again, this is why they should have just gone around.
It's not an easy situation to solve.
And right around this time, they get clearance to land and they reach 500 feet, which is

(01:11:45):
the point at which they have to be fully stabilized for landing.
And that means they must be in landing configuration within five to 10 knots of the target landing
reference speed, VRF plus five, 137 knots with an appropriate thrust setting on a trajectory

(01:12:06):
that will require no further significant control inputs.
Now coincidentally, a couple of these things happen to be met at that exact moment.
They are actually, they hit 500 feet right as they cross the three degree glide path.
So hey, they've gotten back onto the correct altitude that they should be at momentarily.

(01:12:26):
So on glide path, check within five to 10 knots of the target speed actually also check.
They've pretty much reached VRF plus five at this point.
So they're at about 137 knots right as they cross 500 feet.
So on speed, check in landing configuration, landing gear is out, flaps are fully extended.
Check they're in landing configuration.

(01:12:49):
Appropriate thrust setting?
No, they do not have the appropriate thrust setting.
At this point, when you're in a high drag configuration like this with all landing gear
and flaps fully out, you actually need a fairly high thrust setting to be able to maintain
the three degree glide path.
Whereas earlier they need, they weren't even able to maintain the three degree glide path
with the thrust levers at idle.
That was because they didn't have any drag on the airplane.

(01:13:10):
They didn't have any flaps landing gear, anything.
Now that they have all of that, the drag is so much more.
They actually need a fair amount of thrust to maintain that glide path, but they don't.
So that is a strike.
That's not a stabilized approach.
Furthermore, they have to be on a trajectory that will require no further significant control
inputs, but they are descending at well over a thousand feet per minute still at this point.

(01:13:36):
And that is much too fast.
So they will need to make significant control inputs to avoid landing short of the runway.
So again, not stabilized.
Yeah, a safe landing speed for the 777 is probably around 100 to 150 feet per minute.
And it'll be more than that until right before the flare.

(01:13:57):
But yeah, one thousand minus a thousand feet per minute, way too fast.
And so this approach is not stabilized.
If this approach is not stabilized at 500 feet, they have to go around.
But the instructor captain, Li Zhengmin, for whatever reason, he doesn't do the stabilized
approach check.
He looks and he sees that they're on speed and on glide path, but he doesn't look at

(01:14:20):
the rest and he doesn't call it out stabilized.
And we don't really know why he missed this, but it's probably just because a lot was going
on right at this moment.
And he saw those two things and he thought, oh, it's probably not a big deal.
Or he didn't even consciously think about it.
You know, just events.
Again, the pilots are behind the airplane, as Ari discussed earlier.

(01:14:45):
Well, so the other thing, the other thing is if they had been at the sort of 60 percent
thrust that would be a normal sort of setting for a landing with all of these flaps out
and the gear down and, you know, all of this, these very high drag devices and the slats
open and, you know, the whole thing, it's much quicker to actually get to go around

(01:15:11):
power from 60 than it is from flight idle, which might only be 10 or 20 percent.
Yeah, that's about to be.
Yeah, presumably if it was running the CM56, that is a very large engine and that is a
lot of momentum to keep going.
It takes a lot of spin now.
Yes.
So, so within seconds of this, they begin to descend below the 3D glide path because

(01:15:38):
their descent rate is too high.
So as the observer was in fact correctly calling out.
So trainee captain Lee Cancuk begins to pull the nose up to reduce their descent rate.
Now as he pulls the nose up, pulling up causes speed to decrease, right?
So the speed decreases as it is wont to do and he believes when he pulls up, the autothrottle

(01:16:02):
will increase thrust if necessary to maintain VREF plus five, 137 knots.
But remember it's in hold mode, so it doesn't.
So as he pulls the nose up, their speed begins decreasing well below VREF plus five.
And yet he's still, so he's, when the speed is now this low, pulling up isn't as effective

(01:16:23):
and they continue to descend below the glide path because the plane simply does, the plane
does not have enough kinetic energy to climb even if he's pointing the nose up.
And in fact, do we know why Lee Cancuk did not at this point just take manual control
of the throttle and push the levers forward?
He did not because he believed, he fully believed that every last bit of his heart, the autothrottle

(01:16:45):
was going to do that for him, literally.
He was so conditioned to believe that the autothrottle will maintain under all circumstances
the MCP selected airspeed that he, it simply didn't even occur to him that it might not,
right?
And he, so he wasn't even looking at the thrust levers.

(01:17:06):
You know, he, if they're at, you know, he must, he must, he probably doesn't even know
where they're positioned at this point.
Is that different from the Airbus that he's used to flying?
Yeah, in the Airbus, yes, it will always maintain a minimum airspeed.

(01:17:27):
And thrust also works quite a bit different on the Airbus.
Say the pilot selects a thrust category, you might say.
And then the, yeah, yeah, yeah, yes, exactly.
You select a thrust category and then the aircraft will use a combination of throttle
and an elevator trim to maintain the speed that you request.
But yeah, so, so he's, so his attempts to pull up to maintain the glide path are ineffective

(01:17:54):
because the plane has no kinetic energy at this point.
So he's, they're now descending too low.
He looks at the precision approach path indicator lights, the PAPIs, and now they're displaying
one white and three red lights indicating he's below glide path.
So he pitches up even more, but this makes the situation even worse because when he's

(01:18:15):
pitching up and they don't have enough thrust to actually, you know, go up, he's just increasing
the angle of attack, the, you know, the angle of the plane into the airstream, which causes
even more drag, which means that we lose even more speed.
So he's just creating this feedback loop.
He's on the backside of the power curve.
I know you had a slide showing that we removed it.

(01:18:36):
I guess I shouldn't go into too much detail about what that, what that means then.
But I think we do need the next slide though.
Yeah.
So he's pulling up and the automated voice calls out 200 and their airspeed is now 122

(01:19:01):
knots, which is 15 knots below the MCP target airspeed.
And they're still descending at 900 feet per minute with their nose is, they have their,
their pitch angle is more than seven degrees.
Their angle of attack is even higher than that.
They still have most of a mile to go until the runway thresholds.

(01:19:21):
This is not a situation you want to be in.
This is a situation where you need to go around immediately or there is threat to life and
limb.
This continues to develop.
The PAPI now shows four red lights indicating they are dangerously low.
And so what does Lee Kang-Kook do?
He pulls the nose up even more, but he still doesn't increase thrust because he still fully
believes he doesn't have to do that.

(01:19:44):
And so the instructor captain now also sees, Hey, we have four red PAPI lights.
And he glances down at his airspeed indicator and sees only 120 knots.
And he's confused because he also thinks the auto throttle should have automatically held
their speed at 137 knots.
So he's like, has it the auto throttle malfunctioned?
Even he doesn't realize, Hey, we're in hold mode and hold mode, the MCP airspeed doesn't

(01:20:06):
do anything.
So he calls out, he just calls out it's low.
And what does it, does he mean the airspeed or does he mean the altitude?
You know, nobody knows.
Well, I mean, they're both low.
Both.
Yes.
You are the concept of low right now.
He calls out it's low and Lee Kang-Kook says, yeah.

(01:20:28):
And he doesn't do anything.
He just says, yeah.
Yeah, exactly.
So then, so nobody has advanced the thrust levers.
And then at that point, a quadruple chime sounds indicating abnormally low airspeed because
they're now down to 114 knots at 124 feet above the water, descending at 600 feet per

(01:20:50):
minute with half a mile to the runway.
So basically they are on track to impact San Francisco Bay imminently.
And the automated voice calls out 100, that's the ground proximity warning system.
And then suddenly instructor captain Lee Jung-Min has like a heart attack.
He's like, he's like, speed, you know, we're too slow, we're too low, we're going to crash.

(01:21:12):
And he's, and the trainee captain isn't doing anything about it.
So he immediately grabs the thrust levers and he pushes them to full power immediately
and says, go around.
You know, oh shit, go around.
But at this point, it's already too late.
In fact, later the calculations would show the last point at which they could have successfully

(01:21:33):
executed a go around without striking the ground was at the moment of the quadruple
chime, the low airspeed alert, which is about 11 seconds before impact.
And he, he advances the thrust levers to full power seven seconds from impact.
And it takes eight seconds to spool up from flight idle to full power on these engines,

(01:21:53):
as you guys were saying earlier, and they don't have eight seconds.
You know, the plane is going to continue to descend until that full power is applied.
They're out of time, they're out of altitude.
And they're out of energy.
Yes.
And so you have this, these last moments of the cockpit voice recorder.

(01:22:14):
Oh shit, go around.
20, go around.
10.
Oh.
I also like the sound similar to stick shaker lasting for approximately 2.24 seconds, which
you would generally take that as a warning that you're about to crash the plane, right?
You would.
Yes.

(01:22:35):
You should.
Did this fundamentally, could this been prevented if we had taken Jay's suggestion to fit all
aircraft with Jato bottles?
I suggested afterburners.
Afterburners.
Well, I don't know how fast would it, you know, how fast would afterburners have changed
the trajectory of this airplane?

(01:22:55):
I think it had bought it a couple of seconds.
I don't think afterburners really work when you're at flight idle though.
No, not really.
Yeah.
So I guess then we are down to Jato bottles.
Yeah, I don't know, how many Jato bottles would you need to take a 777 that's flying
114 knots at 100 feet above the water and push it upward?

(01:23:19):
You know, probably a lot.
At least three.
Couldn't they try and do that in Iran that one time with a C-130?
Yeah, they tried to get a special operation C-130 to land and take off in a soccer field
and they had the engineering right.
But when they did a test flight, it triggered something like a half a second too early and

(01:23:42):
basically thrust the aircraft straight down into the ground.
It broke up and they did not try again.
And we had Operation Eagle Claw, which was a disaster.
Maybe that'll be a different episode.
Yeah, so I don't know, Jato Rockets questionable.
We'll put a pin in that.
Look, I am absolutely on board with anything invented by Jack Parsons.

(01:24:05):
You know that.
True.
Yeah, no, I think rocket science and cult sex are really the two bedrock foundations
of this podcast.
You know, you don't know if sex magic can get you to the moon or not.
You haven't tried.
Jack Parsons was doing that science.

(01:24:25):
And we thank him.
He is a hero to us.
All right.
So when the cockpit voice recording says, sounds similar to impact, and that is, it
is similar to impact because it is in fact the sound of the plane impacting.
So the tail first grazed the water, which sent, you know, a lot of passengers realized

(01:24:48):
they were in trouble even before this, but they're really in trouble now.
Tail grazed the water for a few seconds and then plowed directly into the three meter
high stone seawall at a speed of 106 knots.
So that's like, what's that?
That's like 130 miles an hour or something.
Yeah, there thereabouts.

(01:25:09):
Basically hitting a brick wall.
Yeah, concrete wall in this case, literally.
Yeah, well, very big bricks.
It's actually it's actually a concrete cap on top of enormous volcanic rocks.
Yeah.
And so this totally obliterated the concrete cap and actually dug those big boulders up
and threw them several hundred meters down the runway.

(01:25:33):
So the tail section broke off.
The plane slammed down onto its landing gear and engines.
Landing gear and engines failed and ripped away.
And then the plane started sliding on its belly with no tail section.
It did a nearly 360 degree pirouette off of the nose and one wingtip.
As you can see in this second still from this security camera footage, the plane is sort

(01:25:57):
of back in the air.
Do we do we know what not exactly go around capable?
No.
What what what was digging it at this point?
The the left wingtip.
OK, yeah.
So I think it's safe to say at this point, this is too late for a go around.
This is, yeah, possibly too late.
So yeah, the the the plane sort of did this dramatic violent pirouette up into the air

(01:26:24):
and then spun nearly completely around and then slammed back down very hard.
This was a lot of jeez and then slid almost immediately to a halt in the grass beside
the runway.
Ari, I'm I'm I'm not an aeronautical engineer.
I'm a different kind of engineer, but correct me if I'm wrong.

(01:26:46):
This is going to void the warranty on that airframe.
I do believe so.
Yeah.
Yeah, I think this Magnus and Vox does not cover something of this of this magnitude.
I believe this would be operator error.
Yeah.
Yeah, I don't know.
This may be beyond the capabilities of speed tape to do anything about.

(01:27:07):
They'll have to use the same jacks they used to get the Alaska plane off the ground.
I think they actually ended up using a crane and they put it onto a very large flatbed
truck of some kind.
Yeah, I seem to remember them doing that.
Actually didn't they use a didn't they use a barge?
Oh, no, you're right.
They did use a barge.
Yeah, they floated it out on a barge.

(01:27:28):
That makes more sense because the water's right there.
Yeah, I think what we're doing is incredible.
And we'll see this in the next couple of slides.
This aircraft stayed mostly intact through an insane amount of kinetic energy being displaced.
Yeah.
So, you know, I so I personally heard from an you know, this is like third hand information,

(01:27:49):
right?
I know an air crash investigator who knows the NTSB investigators who were on duty when
this happened.
And basically, the NTSB started getting a ton of calls from hysterical witnesses that
are like, I just saw a widebody plane cartwheel into the runway at San Francisco.
And like, the people at the NTSB were like, is this legit?
You know, are these people are these people for real?

(01:28:10):
And it took it took like a few minutes to figure out actually, yeah, something has happened.
Now, now on paper, this is sort of the best case scenario.
If you have to have a widebody crash, this is the best case scenario.
You are landing at one of the biggest and most advanced airports in the world.

(01:28:31):
In the middle of the day in great weather.
Yeah.
And as the biggest stroke of luck at all, of all, except for the tail section, the plane
stayed in one piece and the fuel tanks were not breached.
So a fire did start.
But actually, this is this is from the one of the security camera on top of the towers,
right?
Yeah, that captured this photo.

(01:28:52):
This photo.
The so the number two or the right engine had slid off and then become lodged against
the right side of the forward fuselage and an oil tank burst and caught fire and probably
so it ignited the oil and ignited ignited grass.
It probably ignited leaking hydraulic fluid.
And so a fire was starting, but it was not the sort of, you know, raging inferno that

(01:29:13):
consumes the plane in moments type of thing you have when the fuel tanks are involved.
But in the inside of the plane was a mess.
You know, there were slides that inflated inside the plane because the impact exceeded
their G limitations and they malfunctioned.
And so flight attendants were pinned in their positions by slides that inflated and people

(01:29:37):
had to go find knives to try to stab these and deflate them.
And you had I believe the floor actually buckled.
So you had a lot of seats had come out of their rails and were basically pressing up
against each other.
Isn't that right?
The seats actually stay almost universally stayed attached to the floor, which is basically
a triumph of design after previous accidents where the seats just kind of flew everywhere.

(01:29:58):
So that was actually a big success and reduced injuries and fatalities probably quite substantially.
There were a couple of seats that did collapse, but they stayed attached to the floor.
Yeah.
So virtually everyone survived this impact and most were able to evacuate under their

(01:30:20):
own power.
The evacuation didn't start particularly quickly.
In fact, the pilots were sort of, you know, trying they were asking air traffic control,
you know, what is the condition of our plane?
But crashed.
So that is the condition of your plane is crashed.
But before they could actually figure that out, the flight attendants initiated the evacuation
themselves about a minute after the crash because they saw the fire.

(01:30:44):
Right.
So they did that with absolute king shit.
Yeah.
No, as they should.
If you see fire, if you're a flight attendant and you see fire outside the plane and the
pilot has not called for an evacuation, just evacuate.
You know, that's actually what they're trying to do.
And if you are a passenger and a flight attendant says fire, you should also evacuate.
You just take them out their word.

(01:31:05):
They used all the exits on the left side here and and some passengers also walked out the
hole at the tail section and the evacuation was fairly orderly.
However, there were some people who were seriously injured and trapped, especially in the very
back where the G forces have been a lot higher.

(01:31:26):
And you can see the cabin was sort of crushed down a bit there.
And so some people were trapped in that area, especially between collapsed seats.
And one of the flight attendants back there in the by the four L and four R doors was
incapacitated.
The other managed to help was staying on board, trying to free these passengers for something

(01:31:47):
like a considerable amount of time after the crash until firefighters managed to get in
there and pry them out.
The last last person was pulled out of the plane 19 minutes after the crash, by which
time the fire had actually spread up inside the fuselage near the front of the plane already
by that point.
But the worst the people who fared the worst in this crash were the anyone who wasn't wearing

(01:32:11):
a seatbelt in the very back of the plane about to get to that and also the flight attendants
in the very back of the plane, not the four L and four R flight attendants, but the ones
behind even them or the not the three L and three R flight attendants, but the ones at
the very the very back behind the aft galley seated in the tail section that actually broke

(01:32:31):
off.
So the entire area where they were seating was no longer attached to the airplane and
it was thrown some distance down the runway.
These flight attendants were all ejected from the airplane, still strapped into their seats.
And by some miracle, they all survived.
In fact, nobody even knew they were where they were for like 20 minutes until somebody

(01:32:53):
one of them actually got up and walked toward the plane and somebody saw them.
And that's when they realized, you know, oh, shit, we left some people behind on the runway.
But those four all made it.
So I'm presumably did they come out of this bulkhead here where the rear bulkhead?
Yeah, they were they were seated back there.
However, not everyone was so lucky.

(01:33:19):
So in the end, three people died in this accident, 15 or 16 year old schoolgirls from China who
were traveling as part of the same group to attend a Christian summer camp in the US.
They were all seated in the very back.
And one of them had simply had the misfortune of being beamed by the four left exit door,

(01:33:41):
which came loose inside the plane during the crash.
So basically she was put into a coma by the impact from the door and she never woke up,
which is very unfortunate.
And the other two fatalities were believed to have not been wearing their seatbelts and
they were likely thrown bodily from the plane out through the hole in the back as it pirouetted

(01:34:05):
through the air, which is unfortunately not a survivable event.
Now, here, should we talk about the fire response in just a sec?
Yeah, so one of those one of those victims was found way back down the runway, deceased,
and the other was found just forward of the left wing next to that that paved path.

(01:34:28):
And that's where the fire response thing starts to get interesting.
Jay, did you want to talk about that?
Yeah, so the fire response thing was kind of complicated.
So the thing about SFO is that it isn't actually in San Francisco.

(01:34:48):
It's in San Mateo County.
However, the San Francisco Fire Department is actually responsible for the ARFF, the
Aircraft Rescue and Firefighting Team.
They have these kind of cool looking vehicles that you can see they have a sort of robot

(01:35:09):
arm thing that they can use to spray foam into inaccessible bits of planes.
It also has you can see there right at the top of the sort of end manipulator, a pointy
thing and that is actually a pneumatically operated stabby bit.
I think that is the technical term.

(01:35:31):
Stabby?
Yeah, that's correct.
That can be used to actually open a hole in an aircraft fuselage so that the other bit,
which is I believe technically known as the foam squirty bit, can actually then fill the
inside of that fuselage with foam.

(01:35:52):
They had a policy of not doing this until everyone had been evacuated.
This policy existed for no reason.
It shouldn't have existed.
It would have potentially made it easier to rescue some of the people who were trapped
in the plane.
Yeah, the smoke was getting fairly heavy by the time the last people were being pulled

(01:36:16):
out and it was not great conditions in there.
It wasn't at all necessary for it to be this way.
This was a relatively small fire and it was burning not particularly energetically.
So actually deluging the cabin with some foam would actually have made that situation much

(01:36:40):
better for the people who were still trapped in there.
Could have potentially avoided further injuring some of these people, some of whom were quite
seriously injured.
Yeah, there were 49 people were seriously injured in this accident and something like
over 100 had minor injuries.
So there are some people who suffered injuries in this crash severe enough that they're going

(01:37:04):
to be dealing with them for the rest of their lives.
The next thing of course is that because the SF fire department is not actually the San
Mateo County emergency services, they had a different radio system to the surrounding

(01:37:25):
agencies.
At the time they had a thing called Project 25 because San Francisco is kind of a big
city.
I mean, it's not huge, but it's a much bigger city than the small cities on the peninsula
that comprise San Mateo County.
So it had upgraded its trunked radio system, its public safety radio system to this digital

(01:37:49):
standard called Project 25.
Hey Jay, I'm sorry.
What is a trunked radio?
So they want to be able to do sort of radio things, you know, walkie talkie type stuff,
but walkie talkie type stuff means small radios and small radios don't have a lot of range.
And obviously you don't want to have like a satellite or something because that would

(01:38:12):
be really expensive and you don't want to have to build basically a cellular network
because again, that's really expensive.
Cellular networks use microwaves, but these trunked radios use UHF and VHF, which is much
lower frequency.

(01:38:32):
So basically they have these sort of towers that are dotted around the coverage area that
can actually receive these signals and broadcast them out to their immediate neighborhood in

(01:38:53):
what's called a talk group.
But also they can fire it into a cable.
It used to be a coaxial cable that they would run all over the place, but nowadays I believe
it's mostly over the internet, you know, in much the same way that we are currently having
this conversation.

(01:39:14):
They have these sort of points of presence of this trunked radio system and it gives
them push to talk capability that can span dozens of miles of area.
You know, if you have a large metropolitan area, you might have one trunked radio system,

(01:39:36):
but then it's broken up into these regional talk groups where, you know, a particular
police precinct and a particular sort of set of fire stations would all be on this one
talk group.
So because the San Francisco Fire Department is actually 15 miles away from the airport

(01:40:03):
pretty much, ARFF had an agreement, a mutual aid agreement with these agencies in San Mateo
County that if they had a major casualty event such as this...
A plane crash from a widebody?

(01:40:25):
Exactly.
Exactly.
Where a hundred people are injured of which probably 70 of them need ambulances.
This is sort of your worst case scenario, right?
If you're during a disaster response, this is up there with a building collapse in number
of casualties that you're going to have to treat with that are all going to be fairly

(01:40:46):
severe.
Yeah, yeah.
It's very much so.
I mean, this was not the worst case because the worst case would have been that it burned
with a bunch of people who were only barely getting out, in which case, you know, obviously
the number of fatalities would have been much higher and the amount of hospital assistance

(01:41:06):
that they would need would be much higher.
And I don't know if you've noticed over the last few years what with a certain strain
of human malware that's been affecting everyone, but US cities don't actually have a lot of
slack in their ability to provide intensive care to people who might have, you know, lung

(01:41:31):
injuries or severe spinal injuries or, you know, these kinds of things.
So they had to call in mutual aid from all of these agencies, but because the San Francisco
Fire Department and therefore SFOs, ARFF, had moved to this next generation radio standard,

(01:41:54):
but none of the agencies in the area surrounding it had, when these ambulances and fire trucks
and things started arriving, they couldn't talk to each other.
They actually could not have a conversation.
And this lack of communication caused a significant delay in some of these people getting the

(01:42:19):
care that they actually needed.
It was actually a huge thing.
And even the NTSB, which doesn't normally comment on such things, the NTSB was pretty
angry about how badly this went.
Yeah, you know, there were like whole sections in the accident report and so on about how,

(01:42:41):
you know, they had these, you know, mass casualty buses that are supposed to go out to the scene
and, you know, bring all sorts of triage equipment and then, you know, bring injured people back
to an ambulance staging point and how just like for ages, nobody called them because
nobody could figure out how, you know, among many other things.
Yeah, and so it was basically only a few people died, but it was not through good preparation

(01:43:12):
or competent response on the part of the emergency services at the airport.
Now there is sort of an asterisk on this, right?
Isn't there?
Right.
The girl who got run over, which is one of the things this accident is most famous for.
Exactly.
So she was first spotted minutes after the crash lying on the ground.

(01:43:36):
She was not moving.
She was in a fetal position.
It had no outward signs of life, but also no obviously fatal injuries when seen from
the exterior.
But for some reason, the first firefighters on the scene looked at her and they just said,
okay, she is obviously deceased.
And the NTSB said in this report actually was not really clear what they made that judgment

(01:43:57):
based on because there was, again, she didn't have any injuries that would qualify as obviously
deceased despite having been thrown from the plane.
So they just, they were just like, okay, we'll make a mental note of her position basically.
And this was, they didn't put a yellow blanket over her as they would normally, they really

(01:44:18):
should do over confirmed casualties.
They didn't check her vital signs.
So firefighting foam is beginning to build up over this area and not everyone has really
gotten the message that there's this victim there.
And so about how long after the crash did this happen?

(01:44:39):
It was like...
I think it was about 20 minutes.
Yeah.
Yeah.
Okay, 22 minutes after the crash, the girl was run over at slow speed by an airport fire
truck rescue 10 as it was approaching the fuselage with that specialized penetration
device to penetrate it and apply foam to the fire inside the cabin.

(01:45:03):
And they did not realize that they had run over this person until sometime later.
And when they did, they were just like, sort of like, you know, their radio chatter was
like, well, shit happens.
And, you know, that came out and made a lot of people very angry.
Rightly so.
Notably it made several Far Eastern governments very angry was not ideal.

(01:45:27):
The thing is, as far as we know, she was probably already dead.
And this is something, this might be news to a lot of people because the story is often
repeated that she was alive and yet there's quite a lot of evidence that was not the case.
So basically the reason there has been a lot of reporting that she was alive is because

(01:45:48):
the San Mateo County coroner came out and said, you know, based on my autopsy report,
she was still alive.
But his own autopsy report didn't really have any analysis supporting that conclusion.
It just, it just sort of listed what her injuries were.
And one of her injuries was a lacerated aorta consistent with massive deceleration, which
is typically a fatal injury and was probably associated with her being flung from the plane

(01:46:12):
at 130 miles an hour and then hitting the ground.
Yeah, that's that injury is called an internal decapitation, unfortunately.
Yeah, you don't survive that.
I don't think that's, I don't think it's an internal decapitation, but yeah, not to mention
that, you know, she had all these other injuries, basically the same as the other girl who was

(01:46:33):
ejected and found dead.
And furthermore, if you know, she was buried in dust and firefighting foam before she was
run over, if she was alive during that period, she should have inhaled some of that into
her if she was breathing, but there was no trace of any of that found in her trachea.
Yeah.
And also if you haven't been breathing for 22 minutes, you're probably already dead.

(01:46:56):
Yes, exactly.
And you know, there was a, for them, you know, there was a whole lawsuit brought by the the
girl's family against the city of San Francisco over running her over and that eventually
ended with no settlement.
They just dropped the case.

(01:47:16):
They dropped or they dropped the lawsuit.
They never received any kind of settlement because and we don't really know why, you
know, this has never been unsealed, but we think it's probably because there was pretty,
pretty much irrefutable, conclusive evidence, she was not alive when she was run over.
Even though, you know, I think there's probably something, some value to be had in, you know,

(01:47:40):
damages for disrespect of a deceased person.
But like, you know, it was, you know, they really should have done better.
They shouldn't have run her over even if she was already dead, is my opinion.
I don't think that's a controversial hot take.
No, I mean, I mean, they shouldn't, they shouldn't have done a lot of things.

(01:48:02):
Well, yes, it's not quite as bad as, you know, running someone over who is still alive.
You say actually it's quite a lot less bad.
Yeah, there is a large difference there.
Yeah, but still bad.
Let's, let's.
All right.
Okay.

(01:48:22):
So I think if we talk about CRM, so this is crew resource management, right?
And this is a training, it's a training item, but it's also a cultural shift.
And what it essentially presupposes is that there is no hierarchy in a cockpit.
There's a final decision.
Well, there is a, there is a hierarchy, but there's a hierarchy as far as decision making,

(01:48:44):
but nobody's opinion matters more than anybody else.
Everybody's view is considered and problems are, are solved collectively.
Right.
And it's not, you know, there's whether it's you fly the aircraft while I try to diagnose
this problem or, you know, you work the radios, right?
And I want to get out in front of this and say the problem here wasn't that there was
a hierarchy where there shouldn't have been.

(01:49:05):
It's more that simply there was no communication at all, or at least the communication that
took place is not about the right things.
And yeah.
And so I think that this, this, this crash will run into this in a few accidents in this
past, but I think this crash has a sort of undercurrent of racism for a lot of people
where there is a sort of, I don't want to call it a myth or an urban legend, but one

(01:49:30):
of the reasons that Lee Kang Cook did not speak up was that he felt that he was, he
was not allowed to speak up to a superior officer.
And I don't know.
That's definitely, yeah, that misunderstands what actually happened.
You know, the person who didn't speak up was the instructor captain, Lee Jung Min.
You know, he was the one who, who knew who figured out what was going on and should have
said, Hey, wait a minute, you know, my controls going around.

(01:49:54):
And he should have done that a lot earlier than he actually did because the situation
was dangerous from, you know, at least 30 seconds before the crash.
And he didn't do anything until seven seconds before the crash, even though he knew, you
know, our speed is too low.
You know, the auto throttle is not maintaining the MCP speed.
You know, he, he should have known they were in a dangerous situation and he did know they

(01:50:16):
were in a dangerous situation.
And he was, he was sort of handicapped by the fact that he was, this was his first training
flight and Lee Kang Cook had done very well throughout the flight up until that point.
And he was not totally clear on, you know, when does he need to, when does he need to

(01:50:36):
intervene versus letting the trainee captain mess up?
You know, at what point does it go from, he's not doing a good job to he's being dangerous.
And the joke answer is impact, but the real answer probably would have been somewhere
around the do yet way point.
Yeah.
Or at least at the, you know, the stabilized approach point or where they were not stabilized

(01:50:57):
or even a little bit later when he looked down to his airspeed indicator and he saw
120 knots and they were in three reds on the pappies.
Any of these points he could have said, you know, this isn't, this isn't going well.
You know, he had the power to say something.
I mean, he was in the, he was in the, he outranked the trainee captain, but he let the trainee
captain just fly the plane into the ground basically.

(01:51:20):
And part of, part of the reason I think that this happened is because he also did not have
a good understanding of what the plane was doing on the automation level.
So we need to talk a bit about understanding of the automation on the Boeing 777 because
the, after this crash, the NTSB asked a bunch of Asiana airlines pilots and instructor captain

(01:51:42):
specifically, you know, when will the auto throttle not provide low speed protection?
You know, when will the auto throttle allow the speed to fall below the MCP selected airspeed?
Because there are conditions where it will do this, unlike on an Airbus where they're
normally are not.
And most of the instructor captains knew that if the auto throttle was in hold mode, there

(01:52:04):
would be no low speed protection.
So they asked five Asiana instructor captains.
One said he only learned that this could happen after the accident.
Another said he was completely unaware of this exception.
This is an instructor captain did not know that if the auto throttle is in hold mode,
it won't maintain the MCP airspeed.
You know, this, this touches on, you know, an issue that this automation is really complex

(01:52:28):
and it takes, you know, it takes a while to get used to it and understand it.
And so it's understandable that Lee Kang Cook didn't know how it worked.
Pilots say it typically takes at least 300 hours or so on a plane to sort of feel like
you actually understand the, um, its automation and he had only 43 hours.
So it's not that surprising that he didn't understand it, but the instructors didn't

(01:52:50):
understand it either.
And that was a big problem because during his training, Lee Kang Cook had been shown
how the auto throttle will wake up, even if it's turned off and you know, to, to prevent
the, um, the plane from decelerating below the MCP airspeed.
He was like, you know, wow, that's incredible.
It's just like the low speed protections on the Airbus A320 only it wasn't, there wasn't

(01:53:14):
really any discussion during training of the exceptions.
When will this not happen?
And so he came in fully believing with this completely, you know, unquestioned faith from
his, that the low speed protection on the 777 was just like the low speed protection
on the Airbus that he was used to flying and it just wasn't.
And so there was some criticism of Boeing over this because, um, the Boeing 787 has

(01:53:40):
the same auto throttle system.
And back when that airplane is being certified, the F, an FAA, um, test pilot found out about
this when he, um, you know, he accidentally got the auto throttle into, into hold mode
and discovered, wait a minute, it's not maintaining my MCP airspeed.
There's no kind of indication about this and it's barely mentioned in the manual.
And so he said to Boeing, Hey, you have to change your manuals to a very explicitly call

(01:54:03):
this out.
So they did, but they didn't change the manuals for the 777, even though it has the same auto
throttle system.
Wait, wait, wait, wait, you're saying-
Listeners, listeners, Boeing, Boeing not putting things in manuals will become a thing on this
podcast.
Oh yeah.
I was going to say, I mean, that, that's like their main thing, isn't it?
Yeah.

(01:54:23):
So the manual did sort of say it was possible to determine from the manual that this would
happen, but it wasn't, it wasn't like called to the reader's attention in any significant
way.
And after this crash, they did change so that it was, but, um, the design of the system
actually didn't change.
And one of the NTSB members dissented and said, actually, we should have recommended

(01:54:46):
that they change the design because the auto throttle not maintaining the MCP airspeed
in hold mode is stupid.
But the NTSB as a whole did not endorse that position.
Just one guy.
It is, it is my take and maybe even the take of this podcast, that there should be a policy
that if a plane has a kill all humans remote, you remove that boat.

(01:55:10):
And it does require on the pilot not paying attention, but this sort of, it's sort of
conducive to happening in moments when the pilot is not paying attention because the
easiest way to get into the auto throttle into hold mode in a situation like this is
by, is by accident when the situation is already becoming overly complex, which is exactly

(01:55:30):
what happened here.
And so the chances that the pilot will notice this are reduced simply because of the fact
that the situation is most likely to occur in a sit in the midst of another situation
that's already complicated.
And that's, that's potentially a safety hazard.
I mean, it was a safety hazard in this case.
Yeah.
It was a safety hazard right at the point that this plane hit the sea wall.

(01:55:52):
Right.
So to be, to be absolutely, absolutely clear, there are aviation accidents that have happened
because of company culture or even national culture with a very strong authority gradient
in the cockpit.

(01:56:13):
And these have happened, but this wasn't one of them.
Yeah, this is not one of them.
It was not.
And if, if we can blame something on corporate culture, it would be...
Corporate culture did, did play a role.
Absolutely.
The corporate, the corporate policy basically of insisting that Asiana pilots use as much
automation as possible and barely ever giving them any manual flying training, which left

(01:56:36):
basically really everyone involved unprepared to fly what was really, you know, which is
a complicated approach, you know, from a layman's perspective, you know, trying to do all of
that energy management, but also something that should be a fundamental of flying for
anyone who is a captain on the Boeing 777.
And actually the fact that Lee Cancuk didn't really know how to do this while being a captain

(01:56:58):
on the Boeing 777 may have made him really self-conscious.
I think he didn't ask for help doing this and he didn't, you know, he didn't say, I
can't do it.
You know, we need to go around.
And part of the reason for that might've simply been that, you know, the expectation was that
he'd be able to do this and everybody else approaching San Francisco that day was able
to do it.
And, you know, so why shouldn't he be able to do it too?

(01:57:23):
And that's, that can be, that can be a really insidious mindset.
Interestingly actually on that sort of note, before they started the approach, when they
were still actually above 10,000 feet and so they didn't have the sterile cockpit rule,
they actually had this long conversation about the medical requirements, whether they wanted

(01:57:47):
to wear sunglasses or not because of the sunlight dazzling them while they're trying to land
the plane and whether they, whether they had depth perception when they're wearing sunglasses.
They also had a conversation about Jeju Air, which is a low-cost carrier in South Korea

(01:58:10):
and that they knew someone else who was at Air Busan, which is a smaller low-cost carrier
in South Korea.
They mentioned that Jeju's physicals are more lenient, but that the benefits of working
for this company aren't as good.
And it's possible that, again, they didn't want to step outside of the chai ball that

(01:58:32):
they'd been working for for their entire careers at this point.
Yeah, I mean that may be getting a little bit abstract from the situation at hand, but
there was also the fact, again, you have to remember Lee Kang-kook was on his initial
operating experience, so if he made a serious mistake, that would be reported because it

(01:58:53):
may indicate, and it should be reported because it may indicate that he needs more training.
And you know, I mean that's just embarrassing.
It's inherently human to feel that that's embarrassing.
That also could have contributed to him wanting to salvage this approach, even after repeated
signs that it was going wrong.
And yeah, you know, sometimes there are more important things than...

(01:59:20):
Your pride.
Yeah.
We don't know for sure, you know, was pride a factor here?
But you know, actually, I remember one of the NTSB investigators, I think it was Bill
Bramble, who was interviewed...
Or maybe it was Bill English.
I don't know, it was one of the Bills, NTSB investigators, interviewed on the...

(01:59:41):
Curism with these guys.
She's got him on the first name basis.
Yeah, I was interviewed in the Mayday episode of this crash and said, you know, why didn't
Lee Kang-kook say this approach is not salvageable?
And his quote was, I think he just didn't want to admit the weakness.

(02:00:03):
And that's kind of harsh, but also possibly true, because there were so many signs that
this wasn't going well, and he saw a lot of them.
And he just kept going.
This is sort of an abstract thought, but it feels like this is a crash that was caused
by too much trust.
Too much trust in the automation systems, too much trust in each other, right?

(02:00:25):
And I'm not saying that the guys shouldn't have trusted each other, but nobody ever questioned
the situation.
No, not really.
Until it was far too late.
Nobody said, guys, this is an unsalvageable approach, right?
Let's just go around, we'll figure it out.
Or nobody said we're too high.
Again, speed brakes, why hasn't anybody deployed the spoilers?

(02:00:46):
There should be, below 10,000 feet, there should be constant communication about the
landing.
Yeah.
And there was basically, the communication that took place was, you're too high, OK,
I will do something about it.
There was not the call-outs of what the, the pilots were changing modes and things without

(02:01:09):
calling it out.
Like, Lee Ken Cooke selected flight level change speed mode without calling it out,
without discussing this with the instructor.
If he had said, oh, you know, do you think I should, you know, I think I should select
flight level change speed mode to, you know, to help us decrease, decrease the speed, the

(02:01:31):
instructor captain probably would have said, no, that's a terrible idea.
But he just did it, you know, and you're supposed to call these things out.
It's, it's, it's company policy, it's company policy at pretty much every, every airline.
When you talk about CRM and collective problem solving, being consistently too high and being
behind the energy curve on a descent is a problem that they should have been working

(02:01:53):
to solve.
Yeah.
And there was, there was no indication that they, they saw this that way.
And so-
First they had too much energy and then they didn't have enough.
Yeah.
And that, it tends to go that way.
You know, a large proportion of flights that land short of the runway started out being

(02:02:13):
too high, which doesn't, it doesn't make sense unless you're a pilot.
If you are a pilot, it probably makes total sense that, you know, you have to, you're
too high.
You have to try to bleed off this, this height, you know, and then suddenly you're, you're
in a rapidly developing situation much more so than if you're sedately descending along

(02:02:36):
that glide path.
So it doesn't take all that much to, you know, let things go too far.
Yeah.
The irony is if they were too low, if they were below the glide slope, this probably
would have been easier to save.
Oh, undoubtedly.
Because if they were-
And that's why, that's why it's, it's always policy to intercept the glide path from below
and trying to do it from above is discouraged.

(02:02:58):
And probably at some airlines possibly not allowed.
I'm not totally sure about that, but yeah.
Yeah, don't they have a name for that?
Like they call it something like a slam dunk or something?
Yeah.
Slam dunk is when you approach the glide path from above and it's not incurred.
It's not incurred.
It's like, it's like dive bombing.
Yeah.
Yeah.

(02:03:19):
But at the end of the day, this is sort of a, sort of a cautionary tale about neglecting,
you know, fundamentals of piloting and how there were, there were simple, simple things
these pilots could have done to just fly the plane on, on the glide path.
And instead they just kept getting deeper and deeper bogged down into modes.

(02:03:41):
You know, the fact that the, the trainee captain's first thought when he was too fast is change
the, is try to change the autopilot mode instead of changing the configuration of the airplane
is kind of telling.
And this was, this was something that was encouraged by the, the company culture.

(02:04:03):
And I want to quote Robert Sumwalt, who is a, now the, he was the member of the investigation
team who said that this encouragement of overreliance and automation, quote, caused this highly
experienced pilot with an unblemished record to be uncomfortable about manually accomplishing
a very basic task landing on an 11,000 foot runway on a clear day with a very little wind.

(02:04:28):
Yeah.
And that was the dream approach to hand fly.
This would have been ludicrously easy to hand fly.
Well, I mean, if, yeah, when you're trained to hand fly, obviously I couldn't do it.
I've never flown a triple seven, but, but he should have, but he should have been able
to do it.
And the company was systematically on leaving these pilots unprepared to do that in the

(02:04:50):
situation where they, they had to, they had no choice, but to do that because the, again,
the glide slope was out of service.
It all started from that.
Yeah, you're, we have a couple of these accidents, obviously Air France 447, which we'll a hundred
percent be covering, falls in this category of pilots that get behind the, that get behind
the aircraft, that get behind the energy curve.
But it's also sort of a fundamental misunderstanding of what the aircraft is doing at a time when

(02:05:17):
the, when the industry had become over reliant on automation.
Yeah.
And since, and since this crash, there's, it's because this crash has become a major
training topic in, at airlines around the world.
As the industry has moved away from the idea that we should use as much automation as possible
100% of the time, there's been a, there's been a renaissance of basic, basic flying

(02:05:38):
skills because of this crash and others like Air France 447 that you mentioned.
And that's been a positive outcome of this.
And so most pilots who fly large airliners are probably quite familiar with this crash
simply because it's, it's such a big training topic.
Yeah, for sure.
And that's a good thing because it was part of a trend and it's a trend that has been

(02:05:59):
going away thanks to this recognition.
So shall we talk about some racism?
Yeah, let's talk about racism before we end.
Oh boy.
If you were on Reddit, you have seen this image.
KTVU actually broadcast this to millions of people in the Bay area.

(02:06:20):
Yeah, after allegedly receiving these names from an NTSB intern who thought he was very
funny.
At least that's the story.
We don't really know for sure, but that's, that's what they claim happened.
Because they called it, somebody gave them these names.
They called an NTSB intern to ask, are these the names of the pilots?
And the intern was like, that's hilarious internally.

(02:06:42):
And so yeah, they totally are.
And obviously I'm pretty sure that intern was fired.
But that's, that's allegedly what happened.
We don't know who came up with these originally.
I wouldn't be surprised if they originated on 4chan, to be honest.
Yeah that would track.
And I do just want to say, if you are the type of person that finds this funny, please
get fucked and please do not listen to our podcast.

(02:07:03):
We do not want you as a fan.
No, like it was, it was never, it was never really that funny.
It was, it was racist.
And then like people still repeating it 10 years later is just, it's just what are you
doing?
It's, it was, it was not funny then.
It's definitely not funny a decade later.
And yet you can go to any Reddit thread about ASEAN214 and sort by controversial to find

(02:07:25):
legions of brain dead troggs commenting something long.
Like it's the funniest thing they've ever heard.
I just don't understand.
If you find this funny, take a good long hard look in the mirror.
Or if you, I mean, if you, if you still find it funny to comment this under Reddit threads
especially.
I, I'm fairly sure that not only did that NTSB intern get fired, but so did whoever

(02:07:48):
it was, was the writer for KTVU because this should never have gone out on the air.
I mean, if they-
No, it was obvious that this is a joke.
If they'd, if they'd checked it before broadcasting it, this would not have happened.
It couldn't possibly have happened.

(02:08:08):
Who could think this was real?
What kind of idiot would you have to be?
I really don't know.
Have you been on 4chan?
That kind of idiot?
Well, yeah, but I mean, no, on 4chan they'd be completely aware that this is fake and
a joke.
It's why did KTVU broadcast this?
I mean, I don't know.
We'll never know.
Probably.

(02:08:29):
We'll never know.
Nobody's going to admit that they looked at this and saw it and thought, yeah, that's
totally real.
Yeah, and it's unfortunate that this is what the crash is famous for.
This is what people remember about this accident.
Yeah, it literally is.
That's what I said.
You can go to any Reddit thread.
There's just so many people underneath it just commenting this over and over.
Like, why?

(02:08:50):
You're not being original.
Some people see the dive to the bottom and they think, I can go lower.
I can, yeah.
I could be more offensive.
But guys, just, you're not being offensive.
You're not being edgy.
When name sounds different, it stops being funny when you're about seven.
Literally.
But of course, I mean, seven is probably the average age of people commenting and writing

(02:09:13):
threads anyway.
That's fair.
All right.
Do we have any final thoughts to wrap up ASEAN, or ASEANA 214?
Was there anything else that was learned from this that we didn't mention?
I don't think so.
There was no need to change anything physically on the aircraft, right?
Well, I mean, no.
I mean, I think it could have used some, I guess, a software update.
I should have said there was no airworthiness directives, right?

(02:09:34):
There was no mid-trick rounding.
No, there weren't.
Nothing was actually changed.
So the wording in the manual was changed to really explicitly call out that this can happen.
And obviously, every 777 pilot now knows this can happen.
You can't go through life as a 777 pilot and not hear about it because obviously people
died.
Yeah, we're not going to replicate this value mode.
Right.
But I would say, and Kara, maybe you can see if you guys agree with me, that what this

(02:09:55):
aircraft changed was the broader culture, right?
Yeah, I totally agree.
Like you said, this was up there with 447.
Yeah.
And it changed the way that airlines think about training their pilots.
That's absolutely true.
And that's a really positive change.
That's probably more positive than any modification to the 777 that could have been made.

(02:10:17):
Possibly more visible to residents of the Bay Area, or I don't know, maybe not, was
that the radio systems used by San Mateo County and San Francisco all got moved on to Project
25.
And all of those radios are now preloaded with mutual aid, talk groups and encryption

(02:10:39):
keys so that if they have a major casualty event, actually anywhere in the Bay Area,
the emergency services responding to that can actually talk to each other now, which
is something that actually came directly out of this.
That's awesome.
Yeah.
That's really cool.

(02:11:00):
So a fair amount was learned, which is, yeah, we're not going to be the ones who end the
podcast saying nothing was learned just when stuff was learned.
No, for the most part, almost every crash, I think we're going to be able to come away
and say that a lot of positive things came out of it.
At least some.
Some.
There are some crashes where I can't say that, but this is not one of them.

(02:11:21):
This one had a lot of positive change, which is a nice note to end on, I guess.
All right.
Everybody, the next episode will be on Malaysia Air 370.
Totally.
We're being completely honest about that.