7 ms·

Problems are easy to see in retrospect. You could have made similar arguments after every air crash that was caused by a design flaw. These problems are found,

by throwaway201103 6y ago

Problems are easy to see in retrospect. You could have made similar arguments after every air crash that was caused by a design flaw. These problems are found, fixed, lessons learned, new rules enacted, and we move on.

redis_mlc 6y ago
> Problems are easy to see in retrospect. Excuse me, that's nonsense in the case of the 737 MAX. MCAS problems were known in advance by Boeing factory pilots. They sent each other sms messages about it, for which one Boeing manager was fired. The certification paperwork filed with the FAA misrepresented the MCAS behavior, under-reporting the rate and number of pitch downs. MCAS was designed to rely on a single AoA sensor, which was external and subject to damage from jetways, birds and weather. And the software was outsourced to a low bidder paying programmers $15/hour. We'll see how that works for safety-critical aerospace software. But the most sinister thing is that this all resulted from a contractual item - that Boeing would have to pay Southwest $1 million per plane that required re-training. So any concerns about safety were side-stepped to meet that perverse incentive.
salawat 6y ago
No, no they weren't. The problems we have is an organization tasked with transporting millions over the timespan of a year at speeds approaching Mach 1 ended up in tge position whereby financial growth outpaced it's primary reason for existence. To manufacture and support the distribution, operation, and uptake of safe passenger airplanes. Until you get rid of that modus operandi from the upper management, you'll not see a significant culture change.
- Gibbon1 6y ago
  One thing I noticed about VW, their management culture changed a bit after the German government tossed a couple of managers in prison. I think the key thing is a fine is always just another line on the spreadsheet.
bronson 6y ago
Iterating works well with software but not so well with airplanes. And, you must admit that Boeing has shown utter disregard for rules through this process.
adriancr 6y ago
Thing is boeing made incredibly bad decisions that should have been avoided given current rules and procedures. That was only done due to concern for profits and not due to any engineering flaws. (MCAS existing in the first place, it didnt have reliable sensors, pilots were not instructed on it) How would you react if a loved one died due to boeing or any company greed? Cynic in me is hoping this is only allowed to fly in the US until it proves safe and boeing execs and employees are forced on every damn flight.
caconym_ 6y ago
> You could have made similar arguments after every air crash that was caused by a design flaw. I'm not really convinced this is true. For instance, in the classic case of de Havilland Comets brought down by square windows, I think it was an honest mistake that happened because pressurized cabins were new-ish tech. They worked to understand the problem and fixed the design. Nobody was trying to cut corners AFAIK (no pun intended). I guess we could make another category where an iffy design is made to fail by corner-cutting on the part of the airlines, as in the MD-80 pitch trim jackscrew issues. The MAX is a case of the actual aircraft manufacturer explicitly compromising the design of a safety feature because they knew it would help them sell more planes (due to avoiding training/recertification requirements). IIRC, the investigation revealed a consistent pattern of negligence and outright malfeasance on Boeing's part to this effect. I wouldn't say this kind of thing has never happened before, but off the top of my head I have to assume it's pretty rare.
- WalterBright 6y ago
  > The MAX is a case of the actual aircraft manufacturer explicitly compromising the design of a safety feature because they knew it would help them sell more planes The error was not in the concept of the MAX nor the concept of the MCAS. The problem was twofold: 1. MCAS should have used inputs from both AOA sensors, rather than just one. MCAS had too much authority over the travel, and it should have deactivated itself if the pilots repeatedly countermanded it. 2. The pilots were not trained properly in emergency procedures with the stab trim. Boeing put out an Emergency Airworthiness Directive after the first crash with explicit instructions on how to deal with it, but the EA pilots did not follow those instructions. https://theaircurrent.com/wp-content/uploads/2018/11/B737-MAX-AD-1107.pdf https://theaircurrent.com/wp-content/uploads/2018/11/B737-MA...
  - dTal 6y ago
    >The pilots were not trained properly in emergency procedures with the stab trim. Boeing put out an Emergency Airworthiness Directive after the first crash with explicit instructions on how to deal with it, but the EA pilots did not follow those instructions. If the planes are still crashing after the "problem" was fixed, then that wasn't the problem. (Specifically - the "emergency procedures" in question are "what to do if your plane randomly decides to fly you into the ground". The only way to save your skin is to take a specific emergency action within a time window of a few seconds. This is not a reasonable design.)

redis_mlc 6y ago

> Problems are easy to see in retrospect. Excuse me, that's nonsense in the case of the 737 MAX. MCAS problems were known in advance by Boeing factory pilots. They sent each other sms messages about it, for which one Boeing manager was fired. The certification paperwork filed with the FAA misrepresented the MCAS behavior, under-reporting the rate and number of pitch downs. MCAS was designed to rely on a single AoA sensor, which was external and subject to damage from jetways, birds and weather. And the software was outsourced to a low bidder paying programmers $15/hour. We'll see how that works for safety-critical aerospace software. But the most sinister thing is that this all resulted from a contractual item - that Boeing would have to pay Southwest $1 million per plane that required re-training. So any concerns about safety were side-stepped to meet that perverse incentive.

salawat 6y ago

No, no they weren't. The problems we have is an organization tasked with transporting millions over the timespan of a year at speeds approaching Mach 1 ended up in tge position whereby financial growth outpaced it's primary reason for existence. To manufacture and support the distribution, operation, and uptake of safe passenger airplanes. Until you get rid of that modus operandi from the upper management, you'll not see a significant culture change.

Gibbon1 6y ago
One thing I noticed about VW, their management culture changed a bit after the German government tossed a couple of managers in prison. I think the key thing is a fine is always just another line on the spreadsheet.

bronson 6y ago

Iterating works well with software but not so well with airplanes. And, you must admit that Boeing has shown utter disregard for rules through this process.

adriancr 6y ago

Thing is boeing made incredibly bad decisions that should have been avoided given current rules and procedures. That was only done due to concern for profits and not due to any engineering flaws. (MCAS existing in the first place, it didnt have reliable sensors, pilots were not instructed on it) How would you react if a loved one died due to boeing or any company greed? Cynic in me is hoping this is only allowed to fly in the US until it proves safe and boeing execs and employees are forced on every damn flight.

caconym_ 6y ago

> You could have made similar arguments after every air crash that was caused by a design flaw. I'm not really convinced this is true. For instance, in the classic case of de Havilland Comets brought down by square windows, I think it was an honest mistake that happened because pressurized cabins were new-ish tech. They worked to understand the problem and fixed the design. Nobody was trying to cut corners AFAIK (no pun intended). I guess we could make another category where an iffy design is made to fail by corner-cutting on the part of the airlines, as in the MD-80 pitch trim jackscrew issues. The MAX is a case of the actual aircraft manufacturer explicitly compromising the design of a safety feature because they knew it would help them sell more planes (due to avoiding training/recertification requirements). IIRC, the investigation revealed a consistent pattern of negligence and outright malfeasance on Boeing's part to this effect. I wouldn't say this kind of thing has never happened before, but off the top of my head I have to assume it's pretty rare.

WalterBright 6y ago
> The MAX is a case of the actual aircraft manufacturer explicitly compromising the design of a safety feature because they knew it would help them sell more planes The error was not in the concept of the MAX nor the concept of the MCAS. The problem was twofold: 1. MCAS should have used inputs from both AOA sensors, rather than just one. MCAS had too much authority over the travel, and it should have deactivated itself if the pilots repeatedly countermanded it. 2. The pilots were not trained properly in emergency procedures with the stab trim. Boeing put out an Emergency Airworthiness Directive after the first crash with explicit instructions on how to deal with it, but the EA pilots did not follow those instructions. https://theaircurrent.com/wp-content/uploads/2018/11/B737-MAX-AD-1107.pdf https://theaircurrent.com/wp-content/uploads/2018/11/B737-MA...
- dTal 6y ago
  >The pilots were not trained properly in emergency procedures with the stab trim. Boeing put out an Emergency Airworthiness Directive after the first crash with explicit instructions on how to deal with it, but the EA pilots did not follow those instructions. If the planes are still crashing after the "problem" was fixed, then that wasn't the problem. (Specifically - the "emergency procedures" in question are "what to do if your plane randomly decides to fly you into the ground". The only way to save your skin is to take a specific emergency action within a time window of a few seconds. This is not a reasonable design.)