Sadly it seems that Boeing has exchanged their culture of Engineering Safety , by engineers … for a penny pinching culture , by bean counters.
I think it will be difficult for Boeing to recover from the 737 Max disasters , and nearly impossible for them to reset their company culture.
Who would not have thought to incorporate redundancy into a system that had the potential to crash planes and kill people?
The sensor in question is activated by a simple mechanical tab that responds to airflow and sends a signal when it reaches a certain angle. It’s just a basic mechanical stall warning which is found on all airplanes.
A most reliable system would consist of 3 sensors with the system relying on the information relayed by the 2 with the closest readings.
Institutional memory wasn’t the only thing lacking in the lead up to this failure.
I couldn’t agree more.
Cheers,
Earl
You make it sound like a simple limit switch, but there’s a bit more to it than that: AoA sensor teardown by AvE on YT
You beat me to it. 
Since we’re talking about strange reasoning though, what on earth possessed them to have the software alternate between the two senders on successive flights? That’s just asking for problems with resolving complaints.
A reason I can think of is that when alternating systems after each flight you know that both are working properly. With only one connected you never know that the spare one is still working when needed.
Of course this sensor is a bit more complicated than an on/off switch but the concept is simple in that it translate a mechanical signal. It becomes problematic when it gets to tell the flight computer to wrestle control away from the pilots. It might work out fine in a world where computers are infallible but evidently we’re not there just yet. Open the pod bay doors, Hal.
there needs to be 3 of them, then it would work fine as Kalman filtering could drop one if it disagreed
DOH the basis of DP2 class rules
And this, the collapse of the Hard Rock cafe floors. This piece exonerates the engineers and throws shadow on the workers but not surprising given the source.
Either way, this stuff is only supposed to happen in third world countries.
I’m not very impressed by that blog entry. If you don’t take into account the skill level of the people fabricating your system when you engineer it, you’re not much of an engineer. IMHO.
Earl
Article about the last minutes of Lion Air flight 610 from cockpit recorder:
Agreed that the blog entry is crap but unless the architect is a rock star, is this kind of control over the workforce even feasible.
I meant the opposite. You don’t have control over the workforce, so your design has to reflect their abilities or lack thereof.
As I used to say in my software project management class, look around the room at the final design review. Those are the people who are going to pull this off for you. If you’re not comfortable with what you see, you’ve got some decisions to make.
Cheers,
Earl
Please explain. I’d like to hear how you propose to accomplish this.
Did it all the time when I was software team leader or CTO of my own company. Ideally secure software is built from the ground up but that may require a skill set that the organization doesn’t have, so you start by analyzing and modifying off the shelf software for the lower levels and as the team gains skill and confidence you move into more advanced areas. You see similar situations on the Boat Design forum. Some guy reads up on a more or less exotic technique (concrete and aluminum appear frequently) and asks how to get a boat built that way. Then the naval architects chime in and tell him there’s no boatyard in his country capable of building one like that. Or, to examine the converse, the ETNZ/Luna Rossa design team for this America’s Cup came up with a foil arm design that was clearly beyond the abilities of the fabrication team. We know that because the first one out of the shop broke under static test, requiring a new team, a new design, and several months of schedule slip. Production engineers used to make similar decisions when setting up a new factory in a new area, in this case deciding how much parallelism to put in the line with the corresponding risk that parts wouldn’t fit when put together at the end. Robots have pretty much solved that, although supposedly Tesla had some trouble in that area in their early days.
So to get back to the blog’s argument, if the chosen construction method overwhelmed the construction crew, then the real question is why the mismatch between the method and the crew’s capability wasn’t realized soon enough to do something about it. Of course, if the blog’s argument turns out to be wrong and some other factor or factors contributed to the collapse, then this discussion is irrelevant.
If you’re going to be a successful engineering manager you always have to take into account the “who” as well as the “what.” As Petroski said, “To Engineer is Human.”
Cheers,
Earl
Interesting essay on the development of innovative projects but there was no mention of the building in question incorporating radical new methods, designs or exotic materials.