A Boeing contact suggested I read Bjorn’s Corner: Cutting corners in aerospace costs a fortune.
It seems more and more likely the 737 MAX grounding will go well beyond six months and it can approach nine months to a year depending on developments in the next months.
The costs to Boeing for the MAX debacle are now approaching the costs of a new aircraft development.
The end result of the management culture which produced this engineering shortcut is horrendous:
End Result
- Two aircraft and 346 lives lost.
- Boeing in eight months transformed from an admired civil aviation leader to a distrusted brand, subject to several criminal investigations.
- The economic losses are not yet clear but they will approach the costs of a new aircraft development.
Must the management which pushed for lower costs and higher profits now learn the hard way: there simply is no way past thorough and sound engineering in aerospace. Any shortcuts will cost the company many times more than what was saved in the first place. In the extreme, it can challenge the existence of the company.
Bjorn Fehrm, the author of the article is a fighter pilot and an aviation consultant.
Comments to the article are interesting. Please consider a set of comments by Fehrm in reply to a question.
- Research: MCAS was implemented with repetitive nose down trim commands if AoA stayed high. The research for the update found the necessary augmentation only needed one nose down cycle. Research for MCAS was not correctly done.
- Implementation: MCAS was implemented with a single sensor trigger and without global limitation on nose down trim. The implementation of the update uses dual sensors and deactivation of MCAS if they don’t agree. It also has a global limitation of nose down to leave 1.5G nose up authority to the pilot via the elevator. The implementation of MCAS was not correctly done.
- Testing: The testing of MCAS was done by Boeing on behalf of FAA. It did not judge the single sensor triggered repetitive MCAS as dangerous. It judged the Pilot would easily identify an incorrectly functioning MCAS despite not knowing of the function and how to distinguish it from the very similar and ever-present Speed Trim System function. The testing of MCAS was not correctly done.
Not Learning the Hard Way
Boeing saved nothing and lost its reputation and 346 lives by cutting corners.
But did Boeing really learn anything?
I suspect this will all be swept under the rug with a small set of fines and a promise to not do this again.
Ten years from now such promises will be long forgotten.
Mike “Mish” Shedlock
While the debate “rages” here in the comments about BA management versus $9/hr programmers… I note the FAA approved all the nonsense, and accepted zero blame.
If taxpayers can get exactly the same results (eg the FAA is just going to rubber stamp the manufacturer) — why can’t taxpayers save some money by eliminating the FAA? They aren’t actually checking anything.
Ditto the FDA. All these drugs coming to market, tested by their manufacturers. The FDA adds millions to the costs (and those millions in costs get passed on to customers) not to mention uncertainty. Yet, there is no reason to believe the FDA actually prevents bad drugs from making it to market. So why does the FDA exist? Consumers have to take their chances either way, the FDA just makes everything cost more.
Go ahead and tell us about all how the SEC saved public investors from getting fleeced… aaaand its gone.
For those keeping score, the FAA, FDA and SEC cost taxpayers billions each and every year (plus bureaucrats get a pension, guaranteed by taxpayers who don’t get a pension).
Its really frightening how much “public servants” get paid to fleece the public. Perhaps more frightening that there are so many quasi socialists who want to be fleeced even more
Remember the previous Dreamliner battery issues/fires?
What else has Boeing messed up and hidden in the 737 Max design?
Why did 2+? sensors fail so early in working life? – I’d have a hunch there is something more to it like a systemic wiring fault or production issue
Hope they do a comprehensive design audit of everything.
Personally I now have low trust in them to the point of really not wanting to fly on it.
So if the highest 346 executives and employees of Boeing took the seats on those doomed planes would they still design it the same way? Like, you know, cost-savings are very important.
“So if the highest 346 executives and employees of Boeing took the seats on those doomed planes would they still design it the same way?”
They would. And that is exactly the problem.
Obviously, there are some serious issues with regards to the certification of the 737 max, but this isn’t a new debacle. Airbus has had it’s fair share of mistakes and lives were lost because of this plane makers mistakes. The fact is this article along with so many others is full of misinformation that leads people to believe all the details without fully understanding the processes that goes into engineering and then building planes. Yes, people died and that’s unfortunate, but the simple fact is the pilots that flew those planes, didn’t have enough flight time have any business being inside the cockpit of those planes.
“Yes, people died and that’s unfortunate, but the simple fact is the pilots that flew those planes, didn’t have enough flight time have any business being inside the cockpit of those planes.”
It seems to be more than simply the amount of flight time of the pilots.
The crashes had nothing to do with the pilots. Do you follow the news at all?
Is Boeing paying you to make comments that blame the pilots?
“At the hearing into the MAX crashes… Sullenberger strongly rejected the suggestion that pilot error was to blame for the accidents.” “according to Sullenberger, the original version of the MCAS, “was fatally flawed and should never have been approved”.
Today’s “career” means moving often to keep the promotions and salary increases coming. With that observation, I’ve got to ask, “How many managers who made design decision on the 737-Max moved on before the problems surfaced?” Those who moved on will not be held accountable unless there was criminal intent or a complete disregard for human safety. Most likely a committee made the final approval, which hides individual responsibility. Such is the game inside a corporation.
It’s not just the mistakes that were made. It’s also the fraud committed in getting the thing approved. Nor is the FAA free of blame. But revising the authority of MCAS to 4× the original without resubmission of the plans is simply black letter fraud.
Would be nice if significant costs for this could be charged to the senior management fo making these bad decisions. Unfortunately, as we know the largest losses will fall on everyone else. Although hopefully,
it’s no more unfair than the GFC bank bailouts and zero management accountability..
Nah, why would Senior Management take the blame? They’ll pass the blame down to the $9/hr coders, and maybe some to the guy that hired them. The buck passing has already started.
Getting back to the click-bait discussion of the fact that the software work was done in part by $9/hr contract programmers, it would seem that the $9/hr programmers did an excellent job, and are not at fault here at all. The system worked exactly as it was designed to do. Unfortunately the problem was that it was not well designed at all.
Some might say that blindly following orders for $9/hr isn’t the proper measure of “excellent job”.
If the software engineers really were qualified (we don’t know) and if they noticed a shortcoming (again, we don’t know if they did)… an “excellent job” would involve alerting senior managers to the discovered problem. Delivering a high quality product is everyone’s job, at least in companies that last.
India’s government promised to buy $21 billion worth of aircraft (between military and civilian uses), and it appears that was the reason the $9/hr programmers were selected. Putting sales in charge of engineering was a really bad idea.
The $9/hr programmers did not speak up if they noticed a problem, or they weren’t experienced in aviation matters and didn’t know there was a problem. Either way… poor job on their part too.
Obviously the programmers were not aeronautic engineers. Obviously it wasn’t their position to say “this plane has multiple angle of attack indicators, why are we only using one of them?” Someone told them “when this indicator does A, make B happen.” , and that’s what they did.
By all accounts, the planes crash because the program does exactly what it was designed to do. When the angle of attack shows too high of an angle, it makes the plane dive. That’s great, if the angle of attack indicator is correct, but when it fails, and when the second angle of attack indicator isn’t even factored in, the planes crash. That’s a flaw that happened at a lot higher pay grade than the coder.
The implication that these $9/hr programmers take no personal responsibility for their product is a good reason not to hire them. Ever.
You have pride in your work, or you don’t. If you don’t think your work is any good, no one else should either. This lesson about “cheap labor” applies to your local tradesman just as much as to cheap Indian contract labor.
(not absolving management, they chose to hire these dopes in the first place. Not to mention, as@snow_dog does below, BA management actually thought it was a good idea to move INTO Chicago)
I’m not saying they don’t take pride in their work. I’m saying the exact opposite. I’m saying that they were given an assignment, and completed it as specified. The only negative thing that anyone has said about these programmers is that they were paid only $9/hr, and that they were from India. No one has alleged that their code didn’t work as designed.
It is ridiculous to believe that every person who did any work on the plane was responsible for understanding the entire plane design, all the sensors that were present, all the controls that were present, and how they all should work together in a proper aeronautic design.
In a large scale project, people at the top are in charge of the overall design, and knowing how everything works together, and how it is all supposed to interact. Those engineers down the tasks which much be completed into sub-tasks, which they assign to sub-groups. Some engineer at the top made, or should have made, the decision of whether to use one sensor, or both, and how much correction to make based on the signal, not to mention how much control the pilot should have, and what the pilot needed to do to override the automated system.
If they allowed programmers to decide those questions, that is crazy, regardless of whether they were $9/hour programmers or $200k/year programmers, and regardless of they the were contract programmers from India, or staff programmers. Those are engineering and design questions, not programming questions.
Taking pride in your work goes beyond just doing what you are told.
I mentioned the dangers of “cheap” local tradesman because a plumber I recently hired to fix the thing that was visible to me (the employer if you will). He took the initiative to tell me that wasn’t really the problem, showed me the root cause, and saved me several hundred dollars. And he mopped up after he finished working. He takes pride in his work. He wants to do the best job possible.
That guy (and his dad) have been in business for decades. The local Drain-o-rooter (name changed to protect the stupid) is much cheaper, but that franchise just filed for bankruptcy for the 4th time in recent memory.
When one of my neighbors asks me for a recommendation, guess who’s name I am going to give?
As for the quality of their coding, I have no idea what the $9/hr Indians wrote. But their mindless adherence to the specifications tells me they have issues way beyond coding skills
have you done any code reviews of code coming out of bangalore? i’ve done (too) many. too often, the code was fugly, unmaintainable spaghetti code clearly written by novices. but hey, if the unit tests pass, it flies…
You’re quite right. I have worked a fair bit with cheap Indian contract developers. They don’t ask any questions, you give them a spec detailing how you want the software to work and they go and code it. They will give you exactly what you ask for. No more and no less
It’s impossible to say how good of a job they performed without being closer to the problems. It could be the MCAS control software itself was done well, but a loop monitoring inputs to determine when to activate MCAS control was not.
I have seen a bad engineer’s work who made several checks for problems within a loop and several calls to correct the problems immediately as they were detected. The application wasn’t one that required or could utilize such timing considering the loop took about one microsecond to execute, except when those calls were made repeatedly.
The same check and same issue was “corrected” numerous times within one iteration of the loop, which was a problem. It very well sounds like a similar problem could be happening here, and that would not be good quality work.
That said, there are situations where they could have done everything to spec, separately,and there was a flaw in documentation, testing, communication, overall design, etc., that led to this problem while the engineers themselves did excellent work with the material they had.
I want to reserve my judgment, but my experiences incline me to believe the work they performed was low quality. Given the number of software flaws that made it into the final product I am pinning most of the blame on Boeing’s poor acceptance testing and focus on making an easy buck.
So how does Boeing sell 200 more of these death ships?
Letter of intent is not an order…
Trump phone call.
I hold an MBA degree from a good school. More and more I am beginning to think I and my fellow MBA degree holders are largely responsible for so much of the world’s grief. Our rolls in profit maximization in Boeing and banking should be exhibit 1.
Most of the Federal regulators are CAPTIVE to their respective industries including FDA, SEC ++
It seems like more gender studies graduates and fewer engineers would solve this problem pretty quick!! /sarc
The sheer number of clueless twits that the education system pumps out it probably wouldn’t make much difference.
University is organized stupidity