737 Max grounded after second deadly crash by new airplane

[crazy canuck]

My limited knowledge of the industry (mostly helping a guy who *is* an expert write a paper about parts development in aerospace) is that this is done fairly often. Authorities allow large proven manufacturers to conduct certification themselves in order to expedite development, if they are dealing with mature processes. It's not shady or anything - whether is a good idea or not is another matter.

That is a fairly recent phenomenon.  But governments are beginning to understand that allowing industry to regulate itself is problematic.

Boeing as a company has huge incentives to ensure it's aircraft are safe however.  This incident is costing them billions and billions of dollars.  This is not some rogue criminal enterprise.

Yeah, that is exactly the argument that was made to allow industry to regulated itself.   

[Malthus]

I think it's important to differentiate what is normally done, and whether that is a good idea.  A lot of things may look bad in hindsight, and they may indeed be bad, but at the same time they wouldn't be excepional.  For all we know, the same exact regulatory process was applied to planes that never ever crashed, but no one does indepth investigations into why planes don't crash.

Regulation tends to be a one-way ratchet, always increasing over time - as any incident is likely to *increase* regulation for obvious reasons, but a lack of incidents cannot act to *decrease* regulation, because the argument can reasonably be made that it is the existing regulation that prevents incidents.

Problem is that regulation carries costs; having an infinite amount of it is not infinitely good, nor will it make whatever being regulated infinitely safe. Yet any attempt to quantify and limit costs is seen as ghoulish, particularly in hindsight after an incident. A rational decision that, as a society, we are willing to pay such-and-such for X degree of safety is generally treated with distain when someone is hurt or killed by something.   

[Zanza]

The article quoted by Tim does not sound like it needs more or tighter regulation.

The issue here according to that article is that Boeing apparently misstated some facts on the flight control system so that it would be classified less crucial than it actually is for flight safety and thus bypassed the more stringent regulation.

It will be interesting to see if there is a paper trail that shows that these misstatements were deliberate. It seems to be at least negligent.

[viper37]

The 737 MAX has some 300 planes that have been delivered, with only two accidents.  The two accidents appear to be in situations where the air crew did not think to de-activate the automatic pilot when this behaviour commenced.  Not to minimize the loss of two planes full of passengers, and certainly this problem needs to be fixed before the plane resumes flying, but I just don't see where tighter FAA regulation would have prevented these tragedies.

It is my understanding that the problem persists even with the autopilot set to off.

[Barrister]

The 737 MAX has some 300 planes that have been delivered, with only two accidents.  The two accidents appear to be in situations where the air crew did not think to de-activate the automatic pilot when this behaviour commenced.  Not to minimize the loss of two planes full of passengers, and certainly this problem needs to be fixed before the plane resumes flying, but I just don't see where tighter FAA regulation would have prevented these tragedies.

It is my understanding that the problem persists even with the autopilot set to off.

That is absolutely not my understanding.  The issue has happened a couple of times in north america, and each time resolved by turning auto pilot off.  Problem was turning it off is somehow different in the MAX than in older 737s.

[Iormlund]

From what I have read the system needs the autopilot to be off. Which might make sense, because it seems to be designed to compensate for manual error leading to a stall and both accidents took place immediately after take off, when autopilot might not have been switched on yet.

[crazy canuck]

This system is always on and operates in the background unless manually turned off.  Unlike normal autopilot systems, the pilot does not turn it on.

[crazy canuck]

I think it's important to differentiate what is normally done, and whether that is a good idea.  A lot of things may look bad in hindsight, and they may indeed be bad, but at the same time they wouldn't be excepional.  For all we know, the same exact regulatory process was applied to planes that never ever crashed, but no one does indepth investigations into why planes don't crash.

Regulation tends to be a one-way ratchet, always increasing over time - as any incident is likely to *increase* regulation for obvious reasons, but a lack of incidents cannot act to *decrease* regulation, because the argument can reasonably be made that it is the existing regulation that prevents incidents.

Problem is that regulation carries costs; having an infinite amount of it is not infinitely good, nor will it make whatever being regulated infinitely safe. Yet any attempt to quantify and limit costs is seen as ghoulish, particularly in hindsight after an incident. A rational decision that, as a society, we are willing to pay such-and-such for X degree of safety is generally treated with distain when someone is hurt or killed by something.

Yes, regulation incurs cost and to effect cost savings there has been a trend toward self regulation on the naive assumption that corporations can appropriately self regulate.

[alfred russel]

If we've turned over aircraft regulation to industry in the past decade or so, it seems the past few years would endorse self regulation as flying is so safe these days.

[DGuller]

The 737 MAX has some 300 planes that have been delivered, with only two accidents.  The two accidents appear to be in situations where the air crew did not think to de-activate the automatic pilot when this behaviour commenced.  Not to minimize the loss of two planes full of passengers, and certainly this problem needs to be fixed before the plane resumes flying, but I just don't see where tighter FAA regulation would have prevented these tragedies.

It is my understanding that the problem persists even with the autopilot set to off.

That is absolutely not my understanding.  The issue has happened a couple of times in north america, and each time resolved by turning auto pilot off.  Problem was turning it off is somehow different in the MAX than in older 737s.

You're assuming that the reports cited by the media are relevant, which is a big assumption.  MCAS in fact is supposed to function only when the auto-pilot is off.  That makes sense, because auto-pilot will never get the plane near a stall condition.

[Zanza]

If we've turned over aircraft regulation to industry in the past decade or so, it seems the past few years would endorse self regulation as flying is so safe these days.

This story is not about flying in general though, but about one specific aircraft design, which happens to have a very poor safety record with this latest crash. Unless you are assuming that all aircraft average each other out when it comes to safety, the safety record of other aircraft models has no influence on the safety of this type. For all we know, while aircraft in general have become much safer, this particular type might be the exception.

[dps]

If we've turned over aircraft regulation to industry in the past decade or so, it seems the past few years would endorse self regulation as flying is so safe these days.

This story is not about flying in general though, but about one specific aircraft design, which happens to have a very poor safety record with this latest crash. Unless you are assuming that all aircraft average each other out when it comes to safety, the safety record of other aircraft models has no influence on the safety of this type. For all we know, while aircraft in general have become much safer, this particular type might be the exception.

Even if we accept your assertion that this type has a poor safety record, we aren't sure that the crashes are due to an inherent flaw or pilot error.  I don't know enough from what I've read to tell if the MCAS itself is the cause, or that the pilots weren't adequately informed and trained in how it worked.  If the latter, I don't see how more or different regulation in the certification process could have made any difference.

[DGuller]

Pilot error does not automatically absolve Boeing or the design.  You have to design the planes to be operated by humans, and humans make mistakes.  If some design change is making pilots make fatal errors often enough, then it's still a design flaw that regulations exist to prevent or correct.  Mandating pilot education is another thing that regulations do.

[Monoriu]

Self-regulation happens in the university sector as well.  A start-up university requires external accreditation before it can offer a new degree level course.  The external party will examine the curriculum, the background of the teaching staff, send in audit teams etc before it gives the all-clear.  But more established universities are what we call "self-accrediting".  Essentially if it says the courses are ok, then they are ok.  This is because, say, the University of Cambridge's reputation is its most important asset.  It has incentive not to screw up. 

[Admiral Yi]

Only His Excellency, the Sublime Permanent Undersecretary of Hong Kong Educational Affairs would know something that arcane.