Hacker News Reader

Comment by alexk307 | Hacker News Reader

You're upset with the FAA for the Boeing fiasco, but simultaneously upset that they are going to be more strict with certifying a new class of aircraft?

View on HN

Why not? One of the main failure modes of bureaucracy is escalating attempts to deflect blame for both individuals and the organization as a whole, where all that leads to a nightmare of red tape that pulls against the original intent of the organization. So you have a bunch of rules and strictness plus a bunch of snafus.

“Be indiscriminately more strict” obviously isn’t the thing to do. But I don’t think it’s too much to ask that a federal oversight organization be actually-effective at overseeing the industry it’s tasked with overseeing. Be strict enough in smart ways to catch real problems before they happen, might be one way to think of the ideal mandate.

If I had my way, organizations like the FAA would do a good job, they'd do it efficiently, and they wouldn't be subject to compromising incentives.

I think this is basically what you're saying, too

Regulation is not just on a spectrum with not enough on one end and too much on the other end.

There is an orthogonal dimension that comprises the quality of regulation (good and poor).

Would that more understand this idea.

The FAA allowed Boeing to self certify, so they have blame in that. Strict regulations are not necessarily a bad thing if they are actually needed, but regulations for regulation's sake to keep the agency with a purpose is definitely bad. At some point the bueracracy takes over, and it's hard to avoid the useless regulations that make it look like things are being done but ignore the true things needing regulation.

They're taking the wrong approach to safety...

The correct approach is to reward new things, reward changes and innovation, and punish massively anything that causes an accident putting lives at risk.

For example, make a fund that any aircraft manufacturer pays fines into. Fines for engine failures. Fines for oxygen mask deployments. Fines for crash landings. Massive fines for deaths. Aim to fine about 50% of the value of all aircraft sold.

Then give the pot of those fines back to aircraft designers and operators per passenger mile safely flown.

Overall, the industry gets the same amount of cash. But manufacturers and operators who manage to do it more safely will end up more profitable.

You also need a system of watchdogs who try to find 'coverups' - ie. times where a safety procedure is skipped to avoid the fine. A combination of whistleblower rewards and automatic data reporting from the plane should help with that issue.

That sounds a lot like testing in production. Move fast and break things when the stake are human life. I will accept a slower pace of aircraft development, thank you.

We could move fast and quickly figure out a new lead free fuel... Or we could keep using our reliable leaded fuel that kills hundreds of people annually.

Sometimes, 'move fast and break things' actually saves lives. Especially when you're moving fast on the development of new safety systems.

As a pilot, I can't see any reason to fine for an oxygen mask deployment at all and very little reason to fine for an engine failure resulting a safe landing (which is the incredibly overwhelming majority of them in airline transport category aircraft). That would be like fining a driver for a hard braking incident or triggering their ABS.

Not everything needs to have an economy created out of it.

An engine failure resulting in a safe landing was only one or two more failures away from a deadly incident.

When you have multiple layers of safety, having one or two fail is a 'close call' that should be engineered to happen less frequently, and ideally never.

That's the thing. Having one engine shutdown is not a "close call" in a mathematical or engineering sense. The failure rate for revenue service transport category aircraft is well studied. The failure rates are in the 10⁻⁷ to 10⁻⁶ per flight order of magnitude.

In multi-engine turbine transport-category aircraft, an engine shutdown rarely results in a mishap. Extremely rarely. If the goal is safety, shutting down a possibly misbehaving engine is safety enhancing over an alternate system where an engine shutdown results in meaningful fines and so there would be pressure to not shutdown an engine that should be shutdown (or a delay in making the decision to do so).

I'll have to double-check this later, but I think if you took a 50 mile drive to the airport, got on a flight that had an engine failure at V1, and drove 50 miles back to your house that you were at a greater risk during the 100 miles of driving than during that worst-case single-engine shutdown flight. (They're order of magnitude the same I'm pretty sure. If safety is the goal, how much should each of the people who commuted to that flight be fined for their risk-assumption?)

FAA studies on the topic:

https://www.faa.gov/aircraft/air_cert/design_approvals/engin...

Many insurance companies put a box in your car and do fine you for that. It is disgusting, but real.

Reply on news.ycombinator.com