Boeing Admits It Knew of a Faulty Safety Alert in 737 Max | Hacker News Reader

Boeing Admits It Knew of a Faulty Safety Alert in 737 Max

digit.fyi

76 pointsdigighoul30 comments

As far as I can tell, this is exactly the same story as 4 days ago: https://news.ycombinator.com/item?id=19835608

This post is of very low quality. "Boeing admits..." but no source, no date, no quote.

I suppose this is just a mashup of previous articles published in newspapers, which were all discussed here. The WSJ was the first to reveal (11 days ago) that Boeing knew for a whole year that the alert device was faulty[1]. CNBC added some details[2]. A NYT link was published here and much commented[3], where a comment pointed to the official Boeing statement[4].

[1]: WSJ https://news.ycombinator.com/item?id=19772220

[2]: CNBC https://news.ycombinator.com/item?id=19780032

[3]: NYT https://news.ycombinator.com/item?id=19835608

[4]: Boeing https://news.ycombinator.com/item?id=19835901

[edit: typo]

Always looks worse when you keep denying, denying just to turn around and admit it later.

Why is this such a common way to handle these situations?

Likely because it’s not worse. Facebook does this and I’m betting the know what they’re doing beter than your impression that it “always looks worse”

Ha, if only you were really betting, I'd love to bet the opposite side of this. Nothing about how Facebook has been handling its recent crises would lead me to conclude that they 'know what they're doing' in regards to public relations.

This is a guess, but maybe because CEOs are just trying to stall until they can secure a previously agreed upon exit package? Since prosecuting CEOs is so rare these days (here is a great book on that subject https://www.amazon.com/Chickenshit-Club-Department-Prosecute...) it is in their best interest to just deny to avoid culpability.

Not sure why this is getting downvoted. We see again and again that in spite of these kinds of disasters and callous disregard and fraud, CEOs don't actually get punished, instead are rewarded with massive exit packages.

Whilst the lack of prosecution of CEOs is disconcerting, I don't think it is fair to say that CEOs come out of a scandal better because of their exit package.

Sure, its a nice sum of money, but I think most CEOs still prefer not having the money and the scandal not happening. That is not to say that current repercussions are enough, just that it is currently not net beneficial to have a scandal. Notably, with the current low levels of repercussions, it might still be net beneficial to risk a scandal though.

We're not discussing whether or not a scandal is good. We're discussing how a CEO behaves after a scandal has already been unveiled.

Correct. I assure you, at no point in any of these board rooms, are CEOs saying "let's not do that, it might land me in prison" – they're probably just saying "let's not do that because it'll depress the share price by 5%" – but in the case of Boeing and other similar companies, the "let's not do that" part is something that can result in deaths.

Good public relations (PR) is not intuitive. The best practice when it comes to bad news is Robert Dilenschneider's advice to "tell it all and tell it fast", but companies are often hesitant to do so for legal and political reasons.

EDIT: To elaborate, getting ahead of an unfolding story is difficult even for experienced PR people. One misplaced word may be taken out of context or even used as admission of guilt in an ongoing investigation. This is generally why PR folks default to true generalities, e.g. "we care about our customers' safety" or "what happened is terrible and should be prevented from happening again".

My knowledge of PR is limited to the book "When the Headline Is You" which I highly recommend. It gives a great perspective on how companies manage themselves in the public light.

>The best practice when it comes to bad news is Robert Dilenschneider's advice to "tell it all and tell it fast"

When it comes to disclosing hacks, standard practice is to disclose a small hack first then disclose that the hack was "worse than previously thought".

Examples:

https://www.slashgear.com/outlook-com-hacks-is-worse-than-mi...

https://news.vice.com/en_us/article/a3pqqb/the-facebook-hack...

https://www.scmp.com/news/hong-kong/law-and-crime/article/21...

https://www.theregister.co.uk/2018/02/13/equifax_security_br...

https://www.trustedreviews.com/news/ba-hack-investigation-36...

https://www.thedailybeast.com/iran-hack-worse-than-reported

Corruption is widespread.

This is huge organization, assuming that something coherent is going on is unwarranted. Most likely people who made decision to deny didn’t even know about indicator not being installed By default in the first place.

Because it's normally better: "most" people only see the first story, then it's done, ticked off.

Because it's hard to tell the truth when you know the first reaction will be highly negative.

This comes to mind as an example of how things might have gone differently for Boeing if they'd handled this better: https://www.ou.edu/deptcomm/dodjcc/groups/02C2/Johnson%20&%2...

You want to make it appear as though something was being hidden from the executive team so they can deny any culpability.

Who wrote the software?

Even though there are two sensors, the signal from the faulty sensor was used. (Why no comparison checks or consensus checks?)

Also, there was no limit on how far the MCAS could move the stabilizer. MCAS was limited to 2.5 degrees of change 'per cycle' but each time a pilot tried to pull up the nose, MCAS reset enabling another 2.5 degree of movement. Ultimately the stabilizer was pinned in the max-down position.

Boeing and perhaps the FAA let the software into the field but it was bad/lazy software engineers that killed people.

Like all 737s, the MAX actually has two of the sensors, one on each side of the fuselage near the cockpit. But the MCAS was designed to take a reading from only one of them.

Lemme said Boeing could have designed the system to compare the readings from the two vanes, which would have indicated if one of them was way off.

Alternatively, the system could have been designed to check that the angle-of-attack reading was accurate while the plane was taxiing on the ground before takeoff, when the angle of attack should read zero.

“They could have designed a two-channel system. Or they could have tested the value of angle of attack on the ground,” said Lemme. “I don’t know why they didn’t.”

The black box data provided in the preliminary investigation report shows that readings from the two sensors differed by some 20 degrees not only throughout the flight but also while the airplane taxied on the ground before takeoff.

https://www.seattletimes.com/business/boeing-aerospace/faile...

If you have been following, in one of the pieces over the last week or so reported that they outsourced the software.

Which begs the question of who signed off on the acceptance tests.

People need to be in jail.

That wouldn’t fix anything that’s happened in this situation or help prevent it happening again.

I mean, their negligence literally killed 2 planes full of people. It’s hard to view this in a way that ISN’T a jailable offense, at the very least.

>It’s hard to view this in a way that ISN’T a jailable offense, at the very least.

It's highly likely that no one jail-able person (you can't jail a corporate entity) had enough information and knowledge of the situation to meet the criteria of criminal negligence. Nobody knows the full picture. That's just the nature of building a large system like this. I know what my code does. I have a pretty good idea of how it affects the whole system. I can reason able what edge cases I don't know about might cause my portion of the system to do. I can't (with any reasonable degree of confidence) reason about the end result of that edge case being fed through the entire system.

If the corporation has a person then we can jail a person for crimes. The "person" is not allowed to do business or communicate with other persons, corporate or otherwise, without oversight.

Instead of jailing the corporation or fining them, the U.S. government should instead forcibly dilute the company's shares. Obviously you can't put a price on killing two full planes worth of people, but you can instead hit the investors where it would likely force their hands the most. The fine would be that some large (double digit) percentage of Boeing becomes nationalized and all the shareholders have their shares diluted and the U.S. government now has corporate voting rights, etc.

I call BS on that. The person who designed the software knew about testing for mission critical code and if they didn't know, they shouldn't have been hired. The investigation into Toyota's unintended acceleration problems found a widespread corporate culture of ignoring known protocols in the testing of mission critical software. Is this the same problem Boeing has?

Even this story manages to confuse the people responsible with the company where they work. It says "Boeing didn't know the feature was optional." What a crock. Surely someone knew, like the person who decided to make it optional, for starters!

Just jail them for revenge, then? The public isn’t at risk if you just ban them from engineering again.

>Just jail them for revenge, then?

That's the cornerstone of the American justice system, yes.

It might make CEOs care more about safety of their product. Specifically, it might balance out the commercial pressures to take risk for profit.

If calculated risks by coorperations regarinding safety come down on the less risky side, that is a good thing.

No, no. Jail time would definitely make this problem, and many similar corporate malfeasance problems, go away.

Reply on news.ycombinator.com