Waiting..
Auto Scroll
Sync
Top
Bottom
Select text to annotate, Click play in YouTube to begin
NICKOLAS: Oh, my goodness, everybody.  Thanks so much for being here. It is so,   so good to be on stage in person again and I have  missed this so much. And I think we have all.   Like Penelope said, my name is Nick Means, I  work at SIM, empowering security and engineering   teams to make workflows that fit the way  they work. We're hiring. Please say hi.  
And thanks to everyone at SIM for holding  down the fort so I could be here. And   I host a podcast, managing up. If you're a leader,  official title or not, this show is for you.   I learn something literally every time I  record, and I have stickers. Come get one.  And I want to start off with a brief content  warning. This talk contains stories of plane   crashes. One pretty vivid. If you're a nervous  flyer, this might not be the talk for you.  
Now, if you follow me on Twitter, you might  see I'm occasionally in a plane spotting.   Especially when travelling. I love new plane like   that one weird one at London city airport. It has  four engines, but only carries a hundred people.   And the one hanging out on London Heathrow. But I  like identifying planes in the you are a, and one  
of the first I learned to spot was the Boeing 737.  It's the best selling commercial aircraft of all   time, it's everywhere. Once you know the trick,  it's incredibly easy to identify in the air.   It has no doors over the landing gear, if it's too  big to be a regional, it's a 737. You can see how  
the gears swing out from the wheel shaped  cavities in the center of the fuselage.   Why am I talking about plane spotting? Well, there's an interesting reason that the   737 doesn't have rear landing gear doors and  that's everything to do with the problems with   the 737 MAX. We got the first hint that there  was a problem with this plane with the crash of   flight 610 off the coast of Indonesia, October  29th, 2018. This was PK 7QP that would operate  
in 2019. This plane was at Boeing's  delivery facility, August 2018,   just before it left on the final delivery to  Indonesia. That means this plane was just over 2   months old on the day of the accident. Basically  brand new. The captain was Bhavye Suneja,   an experienced pilot with over 6,000 hours of  flying experience, most in the cockpit of a 737.  
And then Harvino, just one name. Fairly common  practice in Indonesia. Almost as many hours as the   pilot. And scheduled domestic service from Jakarta  to Indonesia. The easiest way to tell the story   of Lion Air is through data. This is from the  flight data recorder from the accident report.  
And it tells a really clear story. And let me walk you through it.   First, get you oriented, the timestamps are 1  minute and 19 seconds apart. For some reason,   they divided this into 10 equal parts instead of  the 12 minutes that the flight roughly lasted.   Now more legible labels so we can November gate  this chart. Start with the moment the plane   gets airborne. A couple things immediately  indicate that the plane's got a problem.  
First, as soon as the plane is in the air,  the pilot's stick shaker starts shaking.   This is what it sounds like. It literally  shakes the yolk that the pilot is hanging on to.   It is the most urgent warning indicator on  the flight deck. It's there to get the pilot's   attention when the plane is about to stall. In  this context, it's nothing to do with the engines,   it's the wings and the ability to generate  lift. When it's too slow or too steep, there are  
vortices, disrupting the air flow. This is bad in  the airplane, and when the plane sense this is,   it kicks on the stick shaker. But the  plane has just lifted off the ground and   it's climbing smoothly. Why does the plane  think it's about to stall? Even stranger,   why is it only the pilot's stick shaker going  off? You would expect both to be shaking.  
And the answer to both questions  is a few lines down the smart.  The angle of attack indicated on the pilot's  instruments is 20 degrees later than the co pilots   instruments. What is angle of attack? That's the  angle at which the plane is moving through the air   and doesn't always follow the nose. Approaching  a stall, the plane might be moving more   perpendicular parallel to the ground. Now, the  plane senses angle of attack using a vein on  
each side of the aircraft with each vein feeding  the instruments on that side of the flight deck.   On flight 610, the pilot's was malfunctioning,  reading 20 degrees steeper than the copilot.   This agreement starts on the ground  and lasts the entire flight. And so,   when it does take off, the plane  immediately thinks that it's stalling.   About 2 minutes into the flight, Harvino is asking  for clearance to a holding point. They ask for a  
reason, to which the response, we have a flight  control problem. But doesn't declare an emergency.   A minute later, having reached their hold point,  the captain is worried about the extension of   the wing to generate at lower speeds. He had been  ignoring them because of everything else going on.   But they're flying too fast to have the flaps  extended and retracts them. Almost immediately   the plane plunges 700 feet un commanded. Now, if you're a rollercoaster fan,  
this is three times the main drop of a modern  hypercoaster. He finds himself pulling back on   a very suddenly heavy yolk and has an idea  what's happened. The plane is out of trim.   Sure enough, there's a nose down trim reflected  in the data. What is trim? Let's look at tail   of the 737. At the back of the plane is the mini  wing, the horizontal stabilizer and -- the back is  
the elevator. That's what responds when the pilot  pulls or pushes on the yolk to make changes to the   pitch. It would be exhausting to use it for the  entire flight. That's where trim comes in. Looking   at the front of the horizontal stabilizer you can  see this metal track, this entire mini wing can   angle up or down. And this is what trim adjusts.  It serves as sort of cruise control for climbs  
and descents. For some reason, the auto trim had  made a dramatic adjustment to pitch, the nose of   the plane down, seen here. And there's a trim  adjustment to pull it down. But the auto trim   pulls the nose back down. There's an unwritten  rule in aviation when you make a change to the   configuration of the plane and the plane doesn't  do something you understand, you undo the change.  
That's what he does. He doesn't  understand why auto trim kicked in,   but extends the flaps again and  hopes that the auto trim will stop.   And sure enough, it does. There are a few  routine adjustments, but no more dramatic drops.   But he's worried about the flaps. He's  planning on completing this flight. So,   he retracts the flaps again. And almost  immediately, he's fighting auto trim again.  
About this time Harvino asks air traffic control  for a return to Jakarta. There's a few things that   are odd. The captain is proceeding as if this is a  normal flight. Second, the controller doesn't ask   flight 610 if it wants to declare an emergency.  That's a question a controller should ask you.   Instead of getting airplanes out of the way,  air traffic controller keeps giving them  
turn around other traffic that complicates  the work of keeping the plane in the air.   He fought the plane for 6 minutes while Harvino  scoured the flight manual for something, anything,   that what was going on and fix it. And he  would counter with an equal burst of trim.   It would average out mostly to even over  6 minutes and the with altitude. It would  
have been terrifying to be in the back  of the plane as it went up and down,   but he's keeping the plane more or less at 5,000  fleet. Not sure what else to do, he gives the   controls over to Harvino so he can look through  the flight manual and see if he can find anything   that will help. But he neglects to tell Harvino  what he's been doing to keep the plane level.   The auto trim continues to activate, he  counters, but not nearly enough to counter  
the auto trim. And less than 30 seconds later,  on the cockpit voice controller, he's reciting   from the Qur'an. Flight 610 plunged 5,000 feet  in 15 seconds, killing all 189 souls on board.  
An absolute tragedy. The world understandably   wanted to know what happened. But the other  airlines flying the 737 MAX needed to know   what happened and if it could possibly  happen to their brand new 737 MAXs as well.  So, 8 days later, Boeing would send the  first message to the plane's operators.   Doesn't really say much, though. Basically,  just that the early information about lion 610   indicated there was an un commanded nose down trim  as the result of a malfunctioning attack sensor.  
One problem, there's no documented system on the  737 MAX by which a malfunctioning attack sensor   could have nose down trim. Doesn't exist.  This bulletin doesn't clear that up at all.   Boeing got so many questions that  four days later they sent out another   operator message. Now, this message contained the  first public acknowledgment of the new infamous   Maneuvering Characteristics Augmentation System,  MCAS. This was responsible for the air on 610.  
If you have followed the story at  all, you have probably heard of it.   So, what is MCAS? To answer that, we need  to know about the history of the 737.   737 was launched by Boeing in 1967, 54 years ago.  It's an old plane. Commercial aviation was young,   and they wanted to market beyond the bigger  airports. They needed to fly into smaller  
fields with no jet bridges. They were built  low to the ground to load, and you been load   baggage from the ground without a conveyor or a  stepladder. As you may have figured out, this is   why the 737 doesn't have rear landing gear doors.  There just isn't room. It's too low to the ground.   This was a problem as engine technology evolved.  In the early 1980s to upgrade the engines,  
they had to find a way to fit the engines under  the wings. This strange looking engine inlet was   the result. The only way to make it fit. They  used it in the '90s as well. The 737 next gen   squeezes a flightily more efficient engine  under the wing, using similar but not as   dramatically shaped inlets. But it went from  engineering challenge to real problem in 2011. 
A year prior Airbus introduced the A320neo. It's  the 737's closest competitor, carrying the same   187 passengers along the same routes. It was the  first major revision of the A320 since the launch.   Had a much higher bypass ratio that resulted  in nearly a 20% fuel savings over the older  
A320 and critically, the 737 next gen. Boeing  long disregarded the threat from the Airbus.   Boeing was planning on designing a new  plane from scratch for the 180 seat market   and thought that they would buy the  737 next gen until it was ready.   The CEO got a rude awakening from American CEO.  American was going to buy 400 planes to replace  
the McDonnell Douglas A380s. The modern fleet  was all Boeing. That was until he was called to   let him know that the first 200 would be  a mix of A320s and A320neos from Airbus.   It was a kick in the pants. But Boeing could  compete for the other half of the order.   They wanted the same fuel economy and on the same  5 year timeline that Airbus promised. To fully  
understand Boeing's respondent, we have to go  back to 1997 when they acquired McDonnell Douglas.   Boeing had been an engineering driven company.  McDonnell Douglas, on the other hand, was led by   Harry Stonecipher, a graduate of Jack Welch's GE,  maximizing shareholder value. Came into Boeing as   President and CEO shortly after the merger and  was quickly . They were much more focused on  
margins and stock price. And moved from Seattle to  Chicago. He did this as a means of culture change.   He wanted to give execs more insulation from the  engineers that might push back on those decisions.   When he resigned in 2003, Stonecipher had  this to say in an interview. When people   say I changed the culture of Boeing, that was  the intent. It's run like a business rather  
than a great engineering firm. Now, that sounds  rather innocuous on the service. But this was a   growing win at all costs environment inside of  Boeing. His resignation was a result of this.   Forced out by stolen documents and procurement  government issues on his watch. After the ethical   lapse, the next was cut from the same GE cloth. Suffering from manufacturing delays and battery  
problems, the 737 was the only plane that Boeing  was actually delivering in the early 2010s.   Now, Boeing's stock price was down at this  point because of the problems with the 787.   American's threatened defection to Airbus might  inspire other airlines to buy from Airbus as well.   He was not going to let this happen no matter what  it took. And so, 3 months later, the 737 MAX was   born. Designing it was a frantic project with  engineers working at double their normal space.  
Made some aerodynamic improvements and spec’d  the A320 neo, matching it on fuel efficiency.   American ordered 737 MAXs to fill the rest of the  order but with one condition. The Max had to be   type rated with the 737 next gen. Pilots wouldn't  need expensive simulator to fly the max. Southwest  
insisted on a $1 million rebate per plane if  pilots needed simulator training to fly it.  They get the efficiency from a higher bypass  ratio. There's more space around the core of   the jet engine allowing more air to flow through  it, generating more thrust. But they are bigger.   A lot bigger. Common type rating wasn't a  challenge for Airbus, swap in the new engine  
with few other major changes. But because  of how low the 737 sits to the ground,   it wasn't nearly so straightforward for  Boeing. Even with the sculpted inlet,   it was too large to fit under the 737 wing. Boeing  worked around it. Instead of fitting it below   the wing, they moved it in front of the wing. It  was a great work around, but it had drawbacks.  
And discovered in early wind tunnel testing,  climbing a steep banked turn, the pressure the   pilot felt on the yolk didn't match earlier 737  models because of the extra lift generated bit   placement and the power of the engines. Now, this  is a situation that a passenger plane would almost   never be in, and it has to match with the previous  generation of the plane for a common type rating.  
Some looked for an aerodynamic fix, physical  changes to the plane. That would delay by months.   They added a software solution, the KC46, MCAS. On  the KC46, it is used to keep handling consistent.   On the Max, it would adjust 0.6 degrees nose  down in that one specific scenario just enough  
to mimic the control feel of prior 737s. It was at  this point that Boeing made a critical decision.   The FAA long delegated non critical parts of  the certification to employee manufacturer's   itself. Focused on digging into critical  safety systems. Because of how rarely MCAS   was expected to kick in, it was hazardous, but  not catastrophic. Now, this is important. Because  
that let Boeing self certify the feature. And it  also let Boeing drive the operation with a single   angle of attack sensor rather than redundancy  for systems with catastrophic failure potential.  But Boeing's test pilots found another problem  in the air. We talked about stalls when the wing   loses lift. In previous versions, low speed stalls  are dramatic with lots of buffeting vibration  
and then a quick 30 degree nose down drop once the  stall happens. When the Max stalled because of the   power and the placement of the engines, it was  less dramatic. The nose only gently dropped by 10   degrees. This handling difference would absolutely  keep the Max from a common type rating. This is   common in a plane. Boeing reached for MCAS  again, instead of 0.6 degrees of activation,  
they would quickly add 2.5 degrees of nose down  trim to get the desired 30 degree drop. Again,   the solution worked great. Making the Max behave  almost like previous generations of the plane.   Because MCAS was declared non critical and only  at low speeds when pilots had plenty of time to   react, Boeing didn't update the analysis to  the FAA as part of the plane's certification  
despite the major new use case of MCAS. The chief  pilots was behind one of the major decisions,   he suggested that MCAS be intentionally  omitted from the 737 flight manual   that it operated in the background and pilots  would never interact with it. And the FAA agreed   with the logic and allowed the omission. But  the FAA's agreement didn't take into account   the new low speed stall activation scenario  because Boeing hadn't updated the analysis.  
And something seemingly nobody took into account  was the possibility of MCAS activating repeatedly.   The maximum adjustment of the trim is  4.7 degrees. It would take just two   MCAS activations to get maximum nose down  trim if the pilot didn't counter it. And it   would pause for 5 seconds between the  manual trim adjustment before kicking in.   It was assuming that a pilot would be able to spot  and address the erroneous activation in 3 seconds.  
The reason is because MCAS is remarkably similar  to what pilots do drill for, runaway trim.   The auto trim of the plane starts adjusting to  one extreme and doesn't stop and continuously   goes until it hits activation. The wheels start  trimming. They're located on the console between   the pilot and co pilot in the 737. They have the  white marking on them specifically so you can see  
them spinning. They would spin about 40 times in  a 2.5 degree MCAS activation And even more in a   runaway trim. These wheels would be sitting here  spinning. And when this happens, recovery should   be relatively straightforward. First the pilot  turns off the two stab trim cut out switches,   disabling the trim all together and  reverting to complete manual control.  
Second, the pilot would pull back on the  yolk to get the plane flying level again.   And third, the pilot uses the manual trim wheels  to set back in for flight and gradually release   the pressure on the yolk. We knew there was a  trim problem because he repeatedly countered the   adjustments with the manual nose up trim. Why  didn't they hit the stab trim cutout switches?   Pilot and journalist argues in his New York  Times Magazine article that it's pilot training.  
There are more flight decks to fill than pilots  to fill them. Lion Air had to scale up and train   pilots in a hurry to keep them in the sky. The  assertion was that pilots trained this quickly   learned by rote rather than experience and  graduated knowing the steps to fly a plane,   but not out of certain scenarios. The  stunning 97% graduation rate of the academy  
seems to back up the training. That's an absurdly  high graduation rate from a pilot training. As did   the fact that Indonesia with one of the greatest  aviation sectors in the world has more fatalities   than the global average. The situation didn't  precisely match anything they had been trained in,   because MCAS pulsed on and off instead of  being continuous like a runaway trim event,  
they didn't have the base aviation  knowledge instincts to know what to do.   It was repeated on Ethiopian airlines flight  302. This was after the published on MCAS. The   pilot would have been aware of the situation.  It meant that the stick shaker was operating   the whole flight, and MCAS kicked in as  soon as the flaps were retracted. The   Ethiopian pilots hit the switches, but they had  to turn it back on because the aerodynamic load  
was so great they couldn't manually adjust  the trim. They were going to turn it on,   get the plane trimmed and turn it off. But in the  time they tried to do that, MCAS activated again,   and they crashed shortly thereafter,  killing everybody on board.   The second crash led to the worldwide  grounding of the 737 MAX. And scenes like this,   the planes parked at Boeing's plant in  Washington. Why did this happen? They added  
MCAS to the 737 MAX? Or the FAA's fault for not  more closely certifying it? Or the MCAS failure?   Or the culture change at Boeing? In her fantastic  book, thinking in systems, we are given great   tools to pick apart the situation using systems  thinking. Meadows introduces the stocks and flows.  
Stocks are the foundational units of a system, the  thing you can count and measure. They move slowly,   taking time to change. And flows is increase  or decrease. Analogy of the water in the   bathtub. The water is the stock, and the spigot is  the increase of the amount of the stock, and the   drain is the influence that can reduce the amount  of that stock. And there are feedback loops as  
well in system that affect the rate of flows. Now,  this modeling rabbit hole goes deep, I will keep   this simple. You will laugh about it in a minute. The stock is safety.   It's hard to quantify. But the bathtub analogy  makes sense here. If you think about it,   there's things that make safety in the system go  up and down. It's kind of an overall measurement.  The clouds are the boundaries of the  system, not think where it goes to, but  
what increases and decreases safety in the system.  Never get rid of everything that decreases it,   but keep the safety high enough to make up  for the decrease. Start with pilot training.   We know that the safety in the system  increases the quality of pilot training,   and that pilot training increases safety  in the system. It's a positive feedback   loop. But we also know that the increasing  safety of air travel has been a major factor  
in increasing the amount of travel. Just  the number of people flying every year.   And more travel means that more pilots are needed.  And that demand for pilots negatively impacts the   training quality resulting in a decrease in safety  in the system. The increased amount of travel also   increased the number of planes needed by airlines  every year. This has a couple of effects. First,   a positive influence on the rate of technological  advance. More planes sold is easier to recoup  
the investments. And by and large, those  advancements have a positive increase on safety.   It created an economic environment ripe  for consolidation, there were opportunities   for cost savings. This caused dilution of the  engineering excellence, and the safety culture   made it feel safe to dial back engineering  excellence just a bit in the name of profit.   We know this caused a decrease in safety as well.  And finally, the increase of planes needed in the  
market increased competitive pressure. Boeing had  to play catch up with Airbus. They had to increase   design speed and felt safe doing this because of  the safety in the system. But with we know this   reduced design quality and decreased the safety in  the system. And it made it possible for Congress   to tell the FAA to increase the certification  delegation. That meant missing the change to MCAS,  
clearly decreasing safety. Like I said, simple. Now, if you'll notice, every feedback loop in   this system is rooted in the safety already  in the system. So, what does that mean. That   safety caused the problems with the 737 MAX?  And actually, the answer in a way is yes.  In the intro to thinking and its, meadows  says everyone on everything in a system  
can act dutifully and rationally,  yet all of these well meaning actions   often lead up to a perfectly terrible  result. This is what happened here.   Every actor in the system borrowed a little  bit of safety to optimize something else.   And it is a balance if you think about it.  Because the only way to achieve perfect   safety in aviation is not to fly at all. Now,  most of the choices were reasonable given   their incentives and motivations. And yet the  cumulative effect of their actions in the system  
is tragedy. And this is the lesson for us. If  we only pay attention to the events around us,   to the action of individuals, not to the system  as a whole, we'll struggle to understand why   we're getting the result that is we're getting. And practical terms, if your team is struggling   to ship and you start trying to get individuals  on the team to work harder without understanding   what about the system is slowing them down,  you're just going it burn everybody out.  
Or if there's an incident and you focus on who  did what without paying attention to the why,   you're not likely to learn anything. You'll likely  repeat it again. And you don't need to be in a   leadership role to have this orientation. Anyone  can bring it to the table. All of life is system.   And learning to see and think in systems will  pay huge dividends in achieving what you want.   And a positive note about the 737 MAX.  Systems are interested this their survival.  
During the grounding of the max, a few things  happened. First, Congress and the FAA rethinking   the delegation. Manufacturers cannot no  longer self certify a novel design feature,   and they have to approve manufacturers. This  should increase the safety in the system.   The crashes had an impact on design speed and  quality at Boeing both directly and indirectly via  
changes to FAA delegation and certification. The  737 MAX was grounded for one year and 8 months,   revising the time to get it right. Taking the  time they probably should have in the initial   design phase. Other problems were found and  fixed in the course of re certification as well,   clearly improving safety. The version  of MCAS on the Max can only activate   once in a stall event and only if the pilot  and co pilot attack sensors are in agreement.  
And pilots do need simulator training to fly  the 737 MAX. A positive impact on training   quality further increasing safety in the  system. So, the system has balanced itself.   Corrected in the wake of these tragedies. Now, odds are at least one person in this   room is flying home on a 737 MAX. And I  don't want to leave you anxious about that.   So, if that's you, don't worry. The MAX has gotten  so much scrutiny at this point, it might actually  
be the safest plane in the air. Now, as you  go, good luck learning to see and influence the   systems operating all around you. It is your path  to achieve real change. Thanks so much for coming.  [ Applause ] ¶
End of transcript

This page is an adaptation of Dan Whaley's DropDoc web application.