You're not Boeing to believe this:Yet another software bug found in ill-fated 737 Max airplanes

a reply to: smurfy

Yeah, I've made no secret that Boeing made a series of mistakes that has left them open to liabilities, but the root of the accidents was low-hour pilots with poor training. Dozens of reports of MCAS misbehavior, and thousands of flights across the West -- but the only fatalities occurred with inexperienced pilots at the helm.

Bit of a tempest in a teapot. Media has to sell fear.

a reply to: smurfy

No it wasn't. Knowing about fly by wire doesn't mean people understand how much control the computer has.

From the article-

Overall, the sensors are reliable -- they last an average of 70,000 flight hours and the number of failures is small compared to the hundreds of millions of flights.

Boeing assumed, with the Max, that the pilots would take certain steps, in line with their training, and be able to recover the aircraft. The aircraft they were talking about, with the exception of AF447, the crews all took those steps and the aircraft landed safely even with the erroneous data in the cockpit. All the incidents of the Max showing nose down inputs where the crews followed those steps also landed just fine. The two flights where the crews didn't, crashed. Both flights had extremely low hour pilots in one seat, and one pilot who was known to have trouble with basic airmanship and memory item checklists.

originally posted by: RadioRobert
a reply to: smurfy

if Boeing wants autonomous aircraft, that's one thing, but 

It's not autonomous. Every new airliner has a computer flight laws controlling the flight (fly-by-wire). Not just Boeing. You're running out of options if you want to fly computer-free.

"But Steve Nordlund, a vice president at Boeing, said autonomous technology that would allow for a reduction in on-board crew was being developed at a “good speed”."

www.independent.co.uk...

Seems like people have already run out of options beyond their control...Oh! the irony. Night Night.

a reply to: smurfy

Developing it, even at a good pace doesn't mean that we're going to see it soon. Self driving trucks are being developed at a good pace, but we're years away from seeing them driving around regularly. Automation is coming, but a lot will have to happen first.

originally posted by: Argen
You know, my small brain is beginning to think that this is accepted practice in the computer/coding world. Get it out there and fix it later. Who is teaching these people? Or is it industry unteaching them? Somebody somewhere knows this plane was rushed, like everything else, and many, many people died.

It is accept practice in all of the manufacturing industry; not just computers and coding.

"Test" is considered a "No value added" process task; it is also very expensive. From a program management point of view reducing test is a "great" way to reduce costs.

The push these days is to do model based engineering. If you simulate and model more the less you have to do real hardware testing. The inherent flaw in that way of thinking is that your models can only be as good as your understanding of the system and the reason you do hardware testing to find the flaws you didn't know you don't know about.

Mil-Aerospace has another problem. They are increasingly trying to use commercial electric design and manufacturing processes but don't have the volume to do it effectively and efficiently. They also don't have the luxury of having their customers return a defective product to the store to exchange for another one; they have to get it right every time and commercial electric processes are just not designed for that. .... but try explaining that to dollars and cents program management. If I had a nickel for every time I had to defend the need for adequate testing to management I'd have a big bag of nickels. I once had a quality directory try to explain to me that we could eliminate test all together through continuous process improvement.

originally posted by: Zaphod58
a reply to: smurfy

Developing it, even at a good pace doesn't mean that we're going to see it soon. Self driving trucks are being developed at a good pace, but we're years away from seeing them driving around regularly. Automation is coming, but a lot will have to happen first.

Already said that, but it's a digression from this subject anyway.

Total smartaleky here with dripping sarcasm, lest the public were to misunderstand that the captain is a mythical figure of stature much as a royal is, but to actually fly....................

It should be disclosed prior to boarding passengers that your pilot/captain can only make "suggestions" to the flight control system and that under all circumstances including disagreements the "Flight Law" is actually in charge of the aircraft operation.

There that ought to fix it!

They failed a rollup audit that is pretty much standard before something goes back into production.

The link to the other system is new, but was probably required, however a bad time to be adding a component not originally in the system and needed more debug.

It is comm, but guessing something like a mutex/semaphore type handshake that broke down, but serious enough for a mandatory fix.

Looks bad, but it is a plus that it was discovered. IMO

If you read the full 2009 Turkish 737 accident report including the human factors, previous 737s also have a single point of failure built into them. The 737 Classic was designed so that the left seat was prioritized, since that was the Captain.

During the descent into Amsterdam, the left side radio altimeter was masked, causing the autothrottle to reduce power, and eventually stall. The crew was using the right side systems, and thought they were protecting the aircraft against something like that, but were unaware that the left side would basically override the other systems. It wasn't something that was taught or well documented.