Why is everything wacky centered in Wasilla? First a disgraced former politician and now this? Color me baffled and dismayed.
Wasillallabout?
According to an FAA statement, the plane dove after a seat in the cockpit jolted forward and disconnected the autopilot system, causing a rapid descent. Since the March incident, the FAA received four additional reports of similar issues, with the latest occurring in June. The agency identified loose rocker switch caps on the back of cockpit seats as the cause in three of these incidents. The remaining two cases remain under investigation. As a result, the FAA issued an airworthiness directive, citing the “unsafe condition” of these seats, and required the inspection of 158 U.S.-registered airplanes and 737 airplanes worldwide. These inspections must be completed within 30 days.
Am I reading that right; Boeing can’t even build a chair anymore?
(Yeah, I know the pilot and co-pilot seats on a 787 aren’t just plain-old chairs, but it can’t be that hard to make a good one.)
Boeing didn’t manufacture the seating.
Probably these folks.
Of course, component subcontractors design and manufacturer to the prime’s specifications, so it’s ultimately Boeing’s fault.
I was thinking they probably didn’t, but it was a little more flippant to phrase it the way I did. I guess it would have been more accurate to say that Boeing can’t even buy a good chair, anymore.
And as you say, it’s ultimately their responsibility either way.
They has two opportunities. At the front end of the development process, designing and specifying the seats and the elements of the cockpit the seats interacted with.
And then as the products from the subcontractors were being received and integrated into prototypes or early production, to catch this kind of integration error.
It was a failure of system engineering at both ends.
it seems that the problem was the half-moon shaped lid at the rear of the seat (under the headrest):
IIRC, it opened (or would not stay shut) and a FA serving food touched the rocker-switch with the tray and had the seat move forward inadvertedly, forcing/pinning the pilot against the “stick”.
I’ve seen a video of a B787 seat that would move with the switch cover in place if you leaned on the switch cover itself. I think it was on the Blancalirio YouTube video on the incident flight. So the cover is in place and there’s no indication of anything amiss until you put some pressure on it and it pushes on the rocker switch. Basically does the exact thing it is supposed to prevent.
Here it is. It should be cued to the correct part of the video. https://youtu.be/cRF1YTVJ1Q4?si=s6-LX0zTBZ3-nM49&t=64
Presumably this can only happen if the switch has popped out of the retention mechanism (probably just a rectangular cutout in the plastic) and is dangling free in a way so that the plastic cover can jam it. Still–how would this happen without anyone noticing in advance? Is the switch not used that frequently? I can maybe imagine that it isn’t, since presumably the seat will already be in the far-back position when a pilot gets in, unless the previous pilot had somehow gotten out without moving the seat back.
The seat cutout for the yoke should allow a full forward seat position without interfering with the yoke.
And a memo isn’t going to suffice for a long term solution to a bad switch assembly design.
Or the plastic cover is sufficiently thin and flexible that leaning on it compresses it enough to still press the rocker switch underneath.
Could be, yeah, but in the case of the video Richard_Pearce linked to, he says that the rocker switch is loose, and it does appear that the cover doesn’t close all the way.
Ish…and depending on contracts…
Boeing has the responsibility for the installation and interface, but aircraft seats (crew and passenger) are generally approved under Technical Standard Orders (TSO), meaning that the supplier company (IPECO, in this case) has an approval certificate themselves for the seat design, components, and compliance to certain minimum performance standards (such as flammability or dynamic load qualification).
If the seat can track too far forward and can cause a problem, that is an interface problem and would lie in Boeing’s court. If the seat has a design feature that doesn’t perform, but it’s part of the TSO assembly, that would be IPECO’s responsibility. At least, from a certification perspective, but I’d assume supplier contracts would cover that liability.
Not every element of a TSO is fully investigated and reviewed by the OEM that purchases it, precisely because that same level of regulatory oversight is assumed to have already been done and is acceptable. Youll do certain acceptance tests and review that, but you won’t redo the full risk and hazard assessment and reliability testing of each feature and subcomponent etc that’s already been done. A TSO is more like a black box; you have to use it correctly, but you should be able to trust that the contents function without examining the inner contents and each line of code.
Seats are super expensive to design and certify, which is why they all look the same with variations on upholstery (ok, slight exaggeration!)
I guess the failure is in the grey area where it’s not obviously broken but still doesn’t work as designed and the failure is only apparent in a limit set of circumstances. Under normal use the pilot would lift the cover and operate the switch, it would be rare that they would just lean on the cover.
It would without the pilot in the seat. With a pilot in the seat and maybe a meal tray on his lap is easy enough to see how the column can get moved by the seat moving forwards.
Shouldn’t the failure at least be reported in that case? Maybe the switch has popped out slightly and still works, but it might be jammed by the cover. That’s a potential problem waiting to happen, but if it’s reported to maintenance quickly there wouldn’t be very much exposure. Unless maintenance didn’t view it as a serious problem and didn’t bother repairing it. Maybe there’s a sticky note with INOP under the cover…
The first indication it was faulty might have been the incident. Looks good, works under normal circumstances, but if you lean on it just so it operates when you don’t want it to. First to find out, the crew of the flight.
Or maybe they’d known it was dodgy for a while but didn’t comprehend the implications. I don’t know. Someone has to be the first to find a fault.
I can see that happening but at some point the pilot would just flip the tray up. And it’s moving at a snails pace so that’s plenty of time to dealy with it.
The seats move pretty quickly. And none of this is expected. You’re sitting there bored shitless mid-ocean in the middle of the god-damned night having dinner and suddenly the airplane goes apeshit. Your instinct is to grasp the yoke and try to fly. Not recognize that your tray is the problem.
There are a few bumps, screws, and notches on the column of at least earlier Boeings. Such that the tray edge might catch on one and be unflippable.
The key thing overall with this and many other emergencies and abnormalities, and even 9/11, is that anticipated problems are readily handled whereas unanticipated ones are often flubbed. NOW, every single 787 pilot on Earth knows about the risk of the seat moving due to random actions by FAs or other cockpit occupants behind you. When that happened, exactly zero 787 pilots were aware of the existence of the possibility. Much less had a plan to deal with it.
I expect I have a couple thousand canned scenarios in my head of shit that might go wrong and how to watch for each of them and how generally to react to them should they occur. But there’s still shit I never have considered, and I’m more imaginative than maybe 75% of my peers.
Once the airplane is going apeshit, sorting out root causes is properly prioritized to “much fucking later, assuming we live that long.” Regaining and maintaining aircraft control is job 1, 2, 3, and 4.
The affected pilot probably figured out it was the seat + tray about 10 seconds later. Which is 9.95 seconds too late to use that knowledge to circumvent the ensuing gyration. Shit happens very very slowly in this biz until it happens exceedingly, in fact blindingly, quickly.