I’ll acknowledge that it’s way, way too early to reach serious conclusions, but it seems that the 737-Max was heavily reliant on engineering and woefully insufficient in the area of instructional design/training/user experience…which has fatal consequences.
I don’t want to go all Trump “airplanes are too complicated”, but shit like the above gives me the heebie-jeebies. Airplanes should be commanded and configured by their pilots. We’re never going to give this job to robots, so just let the pilots do their jobs.
I’m curious to know how many flight hours you have logged, and in what aircraft.
If the Ethiopean Airlines crash turns out to be caused by MCAS, then it actually is Trump’s fault because it is a direct result of Trump’s government shutdown.
Actually, haven’t we already done so, partially?
A friend who flies for Delta mentioned that the auto-pilot flies the plane for more time than he does. It’s on most of the time during normal flights. He said he only flies during takeoffs & landings, and during bad weather or unscheduled route changes. And I believe that they are working on automating those. Many big airliners now have the technical capability for auto-landing built in.
Good. The Frenchies are going to be conduct an absolutely AIR tight investigation once the box’s BUS arrives.
I don’t agree with this. The high level of engineering of modern airplanes is precisely why air travel is safer than it’s ever been and continuing to get even better. If one makes the assumption for the sake of argument that in this instance, as with Lion Air, the flight control computer was incorrectly trimming the nose down as a response to the perceived conditions, the problem was that the perceived conditions were wrong due to bad data from the AOA sensor, not that the flight computer was faulty. On Lion Air this was compounded by apparently defective air speed indications. If the pilots don’t know what’s really going on, they are probably as likely to do the wrong thing as the flight computer – maybe more likely. Which is exactly what happened on AF447. It wasn’t the flight computer that did them in, it was bad information coming in to the cockpit and resultant human confusion. A jet airliner doing 630 mph at 35,000 feet over the Atlantic ocean in the darkest night in the middle of nowhere is not exactly amenable to the “seat of the pants” flying practiced by 1920s pilots in their faithful biplanes.
Here’s another counterexample. If Colgan Air 3407 had automatically taken control to push the nose down instead of what the idiot pilot actually did – which was the opposite – it may well have landed safely instead of killing everyone on board. In fact that’s exactly what the flight computer tried to do, but the pilot overrode it.
I think you’re mistaken in your assumptions about the scope of their responsibilities. They’re doing raw data extraction. If you want to be paranoid and conspiratorial, consider that Boeing has a whole team now involved in the investigative effort. Boeing – who declared the MAX 8 perfectly safe right up until today, when they suddenly agreed that it should be grounded “out of an abundance of caution”, but only after every major country in the world had already done so.
Thank god nobody has ever died because the way pilots are trained to do their jobs was faulty.
At that altitude I’m not sure it’s possible to land an airliner in a survivable manner anywhere other than a runway or equivalent paved surface. They were either at the same height or lower than when I used to start landing procedures in small Cessnas and Pipers. It was low, damn low, and regardless of what it was that low they might have all been doomed at that point.
I’m OK with the French BEA looking at the black boxes. What’s important is that that information is looked at as soon as possible because that will tell us more about what happened than eye witness accounts and kneejerk “ground them all”. And why did the Ethiopians drag their feet over getting the black box data out? WTF?
We’ve already given it to robots.
In ordinary circumstances with routine operations and uneventful flights machines really do tend to fly better than humans. It’s when the shit hits the fan that humans become the better pilots…IF they have the training and experience to do that.
And while your “heebie-jeebie” paragraph is full of jargon, it wasn’t opaque to me and I’m just a piddling single-engine piston-engine pilot.
^ This.
Now this is interesting …
Pilots repeatedly voiced safety concerns about the Boeing 737 Max 8 to federal authorities, with one captain calling the flight manual “inadequate and almost criminally insufficient” several months before Sunday’s Ethiopian Air crash that killed 157 people, an investigation by The Dallas Morning News found. The News found five complaints about the Boeing model in a federal database where pilots can voluntarily report about aviation incidents without fear of repercussions. The complaints are about the safety mechanism cited in preliminary reports about an October Boeing 737 Max 8 crash in Indonesia that killed 189.
The disclosures found by The News reference problems with an autopilot system, and they all occurred during the ascent after takeoff. Many mentioned the plane suddenly nosing down. While records show these flights occurred in October and November, the airlines the pilots were flying for is redacted from the database.
Records show that a captain who flies the Max 8 complained in November that it was “unconscionable” that the company and federal authorities allowed pilots to fly the planes without adequate training or fully disclosing information about how its systems were different from those on previous 737 models.
(More at the link …)
I’m happy to have every expert in the world get access to that data. I certainly want Boeing engineers to have a team working on it.
The two incidents quoted (I didn’t see links to the rest) have nothing to do with MCAS. The auto thrust failing to come up is not MCAS and MCAS is not active when the autopilot is on.
As others have mentioned, we’ve already reached a point where most phases of flight are handled by computer. Even when flown “manually”, fly-by-wire computer systems process and interpret the pilot’s inputs in order to generate appropriate control surface movements; extreme control inputs are modulated to ensure that the aircraft remains within the safe flight envelope.
Allowing manual override sometimes saves an aircraft in unusual circumstances; but manual override sometimes causes accidents. When we reach a point, as we surely will, where allowing manual override is a net safety negative, the rational course is to go to full automation. For psychological reasons, this will probably happen much later than it should, and we will continue to employ people to sit in a cockpit even when they have minimal authority to override the flight computers.
To throw in some skepticism, though… it is unusual for safety concerns to be voiced about a new airplane? Is five a lot? Above average, below average? I would fully expect pilots to formally lodge safety concerns; I imagine there is, in fact, a system for doing exactly that, and that pilots are ENCOURAGED to register safety concerns, as that sort of thing is how you improve safety. So maybe this is a bad sign, but maybe it’s absolutely normal. Without context and comparison, none of us know.
Absolutely. I was just pointing out how silly it was to be concerned that the organization extracting the flight recorder data was French. Data is data, and I’m sure that BEA is very competent at what they do.
Regarding the various comments about the story on 787 MAX pilots’ safety concerns, I have no idea of its relevance to this accident, I just thought it was an interesting data point that others may have missed, although CNN has now picked up the story. One would certainly hope that comments about the flight manual being “inadequate and almost criminally insufficient” would not be the norm! And it seems self-evident that Boeing’s Nov 6 2018 notice on the subject of potential “Ucommanded nose down stabilizer trim” problems, issued after the Lion Air accident, is something that should have been in the flight manual from the start.
You want me to list major airline crashes caused by pilot error?
Here’s a few:
Tenerife
Northwest Airlines Flight 255
Varig Flight 254
Aeroflot Flight 593
Those are just the ones off the top of my head. There are other disasters that had another root cause but the crew made bad decisions and the plane went down. Throw in near disasters like the Gimli Glider and Air transat 236, in both of which the crew made fuel calculation errors and flamed out at altitude, but landed safely.
So yea, pilots are infallible. Who needs instruments anyway? Real men ™ fly by the seat of their pants. And always chew Beemans.
Data may be data but extracting it is a technical process.
Germany declined to do it:
“This is a new type of aircraft with a new black box, with new software. We can’t do it,” BFU spokesman Germout Freitag said.
Huh, as the product of “stubborn Krauts”, I’m surprised they’d give up so easily. I’d expect the Germans to work tirelessly until they cracked it: “You will tell us vat ve vant to know!” (whack black box with Hugo Boss riding crop)
It seems to me that what Boeing did was fixing (or rather not fixing) a hardware problem with a software patch. The design of the airplane itself is flawed if throttling up can lead to an uncontrollable nose up and stall condition, that’s the first problem and it’s baked in.
An analogy I used elsewhere was that it’s as if Boeing designed a boat with a hole in the hull and out of expediency “solved” that problem by installing a pump to keep it from sinking instead of producing a boat with no holes in the hull to begin with.
The problem as I see it is that the design is flawed and that leads to two failure modes, one the MCAS doesn’t work when it should and the plane can crash because of uncontrollable pitch-up, the other MCAS works when it shouldn’t and it can (and has!) cause the plane to pitch down and crash.