Quote:
Originally Posted by Sutripta I have problems with 'assist'. You don't? |
No, not particularly. Truth is many system on a modern aircraft are there to assist the pilot. Pilots are often unaware what those system are doing exactly or even when they are doing it.
Here is an example from a Boeing 747-400 manual
Quote:
The control system includes an automatic speed trim function which improves the airspeed stability in certain flight phases. Typically, when the airspeed rises, the aircraft will increase pitch on its own and will climb, which in return will decrease the airspeed and thus restabilize it. Under certain conditions, however, the aircaft is not able to induce sufficient nose up momentum on its own. To compensate this problem, the speed trim function will—to a certain degree— move the stabilizer in the nose up direction when the airspeed rises. Speed trim is active when all of the following conditions are true:
+ 20 seconds have passed since liftoff.
+ Pilot is not trimming the stabilizer.
+ Autopilots are disengaged.
+ Flap lever is not set to 25 and not set to 30. + IAS is between 120 and 220 kt.
+ SRM L and hydraulic system 2 are operative; or SRM R and hydraulic system 3 are operative.
|
How different does that sound from MCAS really? It’s been in operations for decades “assisting” pilots during climb outs on every 747-400 out there.
There is this notion that pilots need to know every little detail that goes on, and must be able to resort to manual control at the touch of a button. Pilots are no engineers and have a functional level of understanding of some of the aircraft system.
When you disengage the auto-pilot, how much automation needs to disengage?
I don’t think anybody would like to see the hydraulic power assist on the flight controls disengage? If you disengage the autopilot, that would disengage the auto trim (but not the function described above!). But the trim function is speed specific. At higher speed less trim is required, so there is automation that controls how much trim is dialled in at the pilots (or auto-pilot) input depending on the speed. Do we leave it on, or should that be disabled too if you disengage the the auto-pilot. How much “assistance” do you take away?
Dis-engaging the auto-pilot does not disengage the auto throttle either. Two complete (set of) systems with a very complex set of interactions. And yes, I believe the auto throttle too is there to assist the pilot. If you disable the auto-throttle you still do not have full manual control over the engines either. There are still some electronic boxes involved, chief amongst them is the EEC (Electronic Engine Controls). You can disable those two, separate actions though, on the overhead panel.
Although it would be nice to think you press a button and everything would be down to manual control, in practice it does not work that and it might not even be possible or desired. What gets disengaged and what not via the auto-pilot is a carefully designed solution.
Quote:
Originally Posted by Sutripta And what is your personal opinion now? (If possible without putting on your lawyer hat please. ) |
Boeing got it wrong somehow. I can understand their overal initial approach to MCAS, but with what we know now, they got it wrong on multiple fronts with the actual implementation.
I still like to think that this is a case of primarily a bad design process, rather than an intentional/deliberate choice. (E.g. like the cheat devices on Diesel Engine, that was an intentional/Deliberate choice of a group of engineers and managers)
It might turn out to be wishful thinking, although we will probably never know. It appears Boeing really underestimated what the ramifications of a single sensor fault could be. It appears they never even tested such scenario’s. Neither desk simulation or real flight testing. Which is remarkable to say the least. Either, they felt it was just not worth the trouble (and they were operating within all legal constraints) or some persons took a conscious call not to go that way.
What is also interesting is to see how difficult it appears to be to recover these 737 MAX whilst nose down. Of course, these two tragic events where caused by the MCAS misbehaving, but MCAS is just some additional hardware and software on top of the existing (auto) trim systems. To put it differently, if a crew would encounter a run away stabiliser for other than MCAS problems, they might find themselves in the same situation. difficult to recover from, especially if you are low to the ground.
Quote:
Originally Posted by Sutripta And we come to the issue at the heart of it all - why did the importance of the MCAS, in fact the existence of the MCAS itself have to be downplayed/ kept under wraps? |
Again, it comes down to what importance is attached to a system during the design phase. Not every system is known to pilot, not every system gets its own dedicated alarm and or check list. It almost sounds they had come up with this neat solution (quick, simple, neat) and convinced themselves it ticked all the other boxes (e.g. legal, safety and certification requirements). They just went on with it. No opposition, no real oversight.
Everybody is screaming to have AoA descrepancy alarms / MCAS alarms installed. There is no evidence that suggest, even remotely, the outcome of these two flights would have been different.
Both crews found themselves in a situation with multiple alarm warnings, stick shaker going continuously, (huge racket) and they were barreling towards the grounds. Adding another red LED into the equation is not necessarily going to get you better situational awareness and decision making. Only so much information a pilot can absorb and process.
In that respect demanding additional training, instruction and simulation is at least as relevant as another red LED.
What I have always found puzzling: We have seen quite a few fatal accidents, where pilots, for a variety of reasons, find themselves with a multitude of alarms, erroneous readings of various instruments etc. For some reason the aviation industry likes to rely on humans to make sense of it all. That’s remarkable to say the least, because humans are not that great at it. Why do we not have advanced automation assisting pilots to figure out what the heck is going on?
Jeroen