SoylentNews is people
SoylentNews
[CEO Dennis] Muilenburg said Boeing is making "steady progress" on a fix to the MCAS flight control system that's at the center of crash investigations in Ethiopia and Indonesia, but he stopped short of faulting the software's basic design.
"We've confirmed that it was designed for our standards, certified for our standards and we're confident in that process," he said. "It operated according to those design and certification standards. We haven't seen a technical slip or gap."
Preliminary reports from both crashes suggest that the MCAS system, which is designed to push the Max's nose down under certain flight conditions, was receiving erroneous data from faulty sensors. In both accidents, flight crews struggled unsuccessfully to take control as the airplanes continually dove just after takeoff.
In his remarks, Muilenburg said the incorrect data was a common link in a chain of events that led to both crashes. It's a link Boeing owns and that the software update will fix.
"[The update] will make the aircraft safer going forward," he said. "I'm confident with that change it will be one of the safest airplanes ever to fly."
Without elaborating Muilenburg also said that in some cases pilots did not "completely" follow the procedures that Boeing had outlined to prevent a crash in the case of a MCAS malfunction.
(Score: 4, Insightful) by The Shire on Tuesday April 30 2019, @04:05PM (6 children)
Every time this topic comes up I have to wonder why MCAS didn't simply tie into the stall warning system with a stick shaker. I don't understand the logic behind deciding that this system must completely override and even fight against the pilot at all. Any pilot worth his salt can properly evaluate the situation and decide if the nose needs to come down or not.
And it was inexcusable to have a system that takes control of the aircraft based on a single sensor. The costs involved in having at least one redundant sensor pale in comparison to the loss of life and aircraft. Any idiot can tell you that sensors sometimes fail. Combine a single sensor with a control system that can override the flight characteristics of the aircraft - well they should have seen disaster coming.
(Score: 3, Informative) by All Your Lawn Are Belong To Us on Tuesday April 30 2019, @05:02PM
The source isn't entirely clean but apparently MCAS did tie into the stick shaker [beasleyallen.com]. If I get it right (and I am only a simulator pilot) the concern was that a vertical stall of the type the MCAS was trying to prevent would be an unrecoverable event (because of the forward COG of the engines), which is why the system would take action to prevent it. I'd swear I'd read accounts of other pilots actually arriving at the correct solution maybe one of the crashed flights experienced the issue the day before and the pilots successfully shut the autotrim down? I liked this AOPA article [aopa.org]. Now why their system relied on only one sensor to indicate a position where pilot control would be taken away and why pilots were not explicitly trained about the system change (beyond risking type acceptance loss?) and to treat it as an out-of-control trim.... that one is a mystery.
This sig for rent.
(Score: 3, Interesting) by bob_super on Tuesday April 30 2019, @07:05PM (2 children)
The lack of input redundancy was a criminal oversight, especially when there are two sensors available ...
But the override can be explained easily: The plane was supposed to silently hide the dynamic flight characteristics changes caused by putting the bigger, more forward, engines. "The pilot should get no costly training or certification, because the plane behaves the same, see?" If you properly diagnose a problem, and shake the stick in response, a highly experienced old-737 pilot will wonder why the plane is not behaving as expected in a familiar situation, braking the "training-free" concept. So you don't, and just nudge the controls to make it seem like the old behavior.
At least, that's the decision Boeing appears to have made.
(Score: 2) by driverless on Wednesday May 01 2019, @07:11AM (1 child)
It wasn't a criminal oversight, as Boeing have pointed out it was available as an optional, paid-for extra. I hope this teaches all the other airlines a lesson and they'll now cough up the dough to Boeing to pay for that optional extra.
(Score: 1, Informative) by Anonymous Coward on Wednesday May 01 2019, @08:45AM
Nope. The optional, paid-for extra was only an indicator light that showed if the two AoA sensors reported different data. It did not suddenly feed two sensor streams into the MCAS system.
(Score: 2) by choose another one on Wednesday May 01 2019, @10:45AM (1 child)
MCAS doesn't "tie into" stick shaker because it is not an anit-stall system. Period.
Yes the media have repeatedly painted it that way, because they can just about get their heads around that, but that doesn't make it true.
Both MCAS and stall warning system operate at high AOA, but that is where the similarity ends - they don't even trigger at the same AOA threshold (MCAS can kick at lower AOA).
MCAS is a (very) blunt tool way of altering the "feel" of the stick because on the MAX the stick can get lighter at high AOA due to lift from the engine nacelles (because they are very forward, which is because they are bigger and because bigger didn't fit under 737 wing). The stick "feel" is a certification requirement, the MAX couldn't meet that requirement without _something_ to make the stick heavier in certain corners of the flight envelope. In technical terms we are talking about the Cm-alpha curve - for a far better explanation than I could ever give see Bjorn's corner over at Leeham:
https://leehamnews.com/2019/02/15/bjorns-corner-pitch-stability-part-10-wrap-up [leehamnews.com]
(or read the entire series if you have time).
(Score: 2) by The Shire on Wednesday May 01 2019, @08:06PM
Except in both instances the pilots were unable to override the MCAS nose down. That doesn't sound like a simple change to how the stick feels - thats a full on override.