Hacker News Reader

Comment by Someone1234 | Hacker News Reader

I'm glad the FAA does what they do. While it is easy to hand wave away their concerns, it is better to over-spec it now and to test the design than to just let manufacturers put "alpha" products into commercial usage and only fix them after a body-count.

If driving/vehicles/roads were held to the same or a similar standards to aircraft there might be fewer accidents and deaths today.

Ironically self-driving is held to a much higher standard than regular driving; which is appreciated but further makes one question why regular "dumb" vehicles are so under-regulated (the analogy that springs to my mind is when HTTPS with a self-signed certificate used to be treated much worse than regular HTTP, it took years to treat HTTP as the insecure connection it is, manual vehicles are the same way, something that has been around so long we ignore its safety issues).

View on HN

Bringing the same rules to light aircraft as airliners has actually made them less safe. Bringing a new four place plane to market has been cost prohibitive. The majority of new planes built today were designed in the 50’s on slide rules. Imagine if we were all still driving 50’s era automobiles. Luckily, the FAA loosened these rules this week.

That's assuming that over those years, the market would select for safer flights and not cheaper ones at the expense of safety.

The issue is that non-commercial aircraft are being held to the same standard as a 777, not that the 777 is held to an extremely high standard. There's nothing safe about a Cessna 172, period, and it's far too expensive for them to redesign it. So it's going to continue to be a 60 year old plane with updated avionics.

> There's nothing safe about a Cessna 172

Genuine question: is the 172 really considered unsafe, googling "cessna 172 safety record" brings up a lot of articles suggesting it is as safe or safer than comparable aircraft - or are you saying that light private aircraft aren't very safe in general?

Not specific to the 172, but a sizeable part of the training when getting your license is about what do when the engine stops (gliding, finding a suitable field/road/etc.) - and if you fly long enough there's a fairly good chance you'll have to put what you learned into practise.

You do realize most airplane crashes are small planes.

And most automotive crashes are non-commercial vehicles as well.

>There's nothing safe about a Cessna 172

I assume you have nothing to do with aviation. The Cessna 172 is extremely safe. In fact I'd argue that they're safer than commercial flights, because accidents are much more survivable. Safety depends purely on the pilot and the extremely well-documented maintenance routines. All accidents I've ever heard of Cessnas are caused by pilot error with minor exceptions.

Planes crashing due to pilot error is unsafe. Cars brake themselves now, Pipers have parachutes, B777s have redundancy on redundancy. The Cessna 172 is a very old design. Surviving an accident due to going too slow to die isn't "safe".

>Cars brake themselves now

Uber clearly doesn't follow this trend. Neither does Volvo, apparently https://www.youtube.com/watch?v=_47utWAoupo

>Pipers have parachutes

Only because they're prone to a certain unrecoverable spin in some configurations. You can get a ballistic chute on a Cessna too if you like.

>The Cessna 172 is a very old design

That's because it's a great design, and one that has been tested for 50 years. I'd take it over modern ones.

>Surviving an accident due to going too slow to die isn't "safe".

It is if you don't die.

No. It's assuming that modern tech is safer and more reliable. (A debatable assumption, but also a pretty fair one.)

Anyone that's ever looked to buy private planes knows that, except very small instructional aircraft, most are 40+ years old. My father-in-law just put an offer on an 8-seater built in the 1960s. And that's pretty much just the way it is for buying airplanes. It requires old tech and frequent maintenance, but it's still cheaper than the alternative.

But the problem with that is not so much that it creates a market for old airplanes but rather that airplane certifications don’t expire – if the 50’s planes you allude to had to obey the same standards as new planes (after some grace interval of e.g. 10 years), the problem could be solved?

> But the problem with that is not so much that it creates a market for old airplanes but rather that airplane certifications don’t expire

It seems to me that the problem here is that new airplanes are cost prohibitive. Your solution to light aircraft safety is to make the majority of light aviation illegal or too expensive to function, like solving the problem of unstylish watches by outlawing anything cheaper than a Rolex.

People seem to have the impression that GA has to be safer than driving, or safer than some other benchmark, before the safety level is acceptable. I am perfectly willing to accept some risk (the raw rate of fatality per ground mile is about 20x higher for GA vs. driving) even as a passenger, and I don't see why the FAA should be in the business of determining what risk is acceptable to passengers. I think the FAA's primary responsibility should be to the broader public, who experience the same risks on the ground regardless of how much or little GA they buy, rely on, or associate with.

Some of the FARs make a lot of sense in this regard, for example the rules for single-occupant ultralight aircraft (including powered paragliders), and ideally a similarly pragmatic approach can be applied to other questions in aviation regulation.

When planes crash they might fall on houses. That seems like a good reason to me to make them a lot safer than cars.

Well, also consider that you don't make GA safer than driving by regulating it. A lot of GA can be compared with the kind of unconventional driving done on remote or logging roads. GA is pretty safe where driving is also safe.

Luckily, when cars crash (or just dride) they never hit anything like houses or people.

Oh wait ...

You say that, but the same can be said about cars - you can still drive a car from the 50's, even if it no longer meets current-day safety standards (notably, rollover protection and environmental standards). Older vehicles, be it cars or planes, will retire naturally because maintenance eventually becomes prohibitively expensive.

This ignores that, unlike cars, 1950s planes are built from scratch today.

Meet the most popular aircraft in history: https://en.wikipedia.org/wiki/Cessna_172

The only updates are new engines, the aircraft's control is not just analog, it's literally "planks moved by pulling on wires".

> Bringing a new four place plane to market has been cost prohibitive. ... the FAA loosened these rules this week

Is there any independent research that supports this? I've often heard the argument that regulations are too expensive as an excuse for lowering standards.

> The majority of new planes built today were designed in the 50’s on slide rules.

That's not true.

Of the 15 most popular light aircraft models built in 2015, only one was introduced in the 1950s: the Cessna 172. And even that one underwent a pretty substantial overhaul in 1998. The most popular light plane sold in 2015 was the Cirrus SR22 with 270 units sold. The SR22 debuted in 2001.

Source: http://www.fi-aeroweb.com/General-Aviation.html

And costs upwards of 600 thousand new.

>Of the 15 most popular light aircraft models built in 2015, only one was introduced in the 1950s: the Cessna 172.

It's disingenuous to say that and not talk about numbers. There were over 44 thousand of them built, and they are literally the most produced aircraft ever. Cirrus is a blip on the radar compared to Cessna.

At 300 planes/year they will not catch up with Cessna in a century even if Cessna ceased to exist today!

In fact, the top six most produced civilian aircraft ever that are still in production are[1]:

Cessna 172 (1950's)

Piper PA-28 (1960's)

Cessna 182 (1950's)

Cessna 152 (1950's)

Antonov AN-2 (1950's)

Beechcraft Bonanza (1950's)

Do you have the statistics on the market share of the "slide-rule" planes today? Not as a percentage of planes being built or sold, but of the ones in service. I'd bet on them being the majority.

[1] https://en.wikipedia.org/wiki/List_of_most-produced_aircraft

No, it is not disingenuous because the topic at hand is "the majority of new planes built today."

Obviously if you look at production through all of aviation history you are going end up overweight in older models. But that is just a reflection of the fact that planes are generally designed and built to last. It has nothing to do with the current state of the art, which is what the OP was lamenting.

OK. You are correct, you have refuted the point that the majority of planes built today are of older designs.

The point we were discussing is that excessive regulation on light aircraft prevents innovation by driving the cost prohibitively up and thus leading to a decrease in sales of new designs.

It is not clear to me that the argument that the industry overall is less safer is invalid if most of the planes flown are old, and if over-regulation of light aircraft is the cause of that.

We do need the statistics on the numbers of old-design planes in the air, together with the pricing data, statistics on sales which include old planes, etc.

Wow, new planes cost more than old planes. Who knew?

New planes today cost more than new planes used to cost back in the day (accounting for inflation), which creates more demand for older planes.

Cessna 172 cost under $9000 in 1956 when it was introduced, which is about $80K today when accounted for inflation.

Do you know of any similar new aircraft in the same ballpark? Last time I checked, $50K still gets a Cessna 172 built in the 70's. A new Cessna 172 seems to be north of $250K, depending on configuration.

Cirrus SR22 is north of $550K.

Civil aviation has definitely become a richer man's game over the past 5 decades; the question is why this happened.

To be clear: a middle-class enthusiast really can't afford to buy the best-selling plane of the year new today, while that was a realistic option in the 1950's.

1. Do you not think this is bad?

2. What do you think are the causes of that?

These are very different topics from the original one under discussion, and much more complex. I don't know why it should matter to you what I think about this, but since you asked, no, I don't think the increased cost of aircraft is bad. Aviation is a very environmentally destructive pastime, so high barriers to entry are a good thing.

As to your second question, liability concerns and government regulations certainly add a lot to the cost of an airplane. But the flip side of that is that modern airplanes tend to be pretty darn safe. As a pilot, I think that's a good thing too.

> If driving/vehicles/roads were held to the same or a similar standards to aircraft there might be fewer accidents and deaths today.

Vehicles are already regulated with regards to safety features; seat-belts, airbags, anti-lock braking systems etc. These have and do prevent many deaths and serious injuries.

But holding cars to the same mechanical rigors as airplanes would make them neigh unaffordable for most, to say nothing of yearly inspections many states have which is often considered nothing more than a government subsidy to mechanic shops. It is also entirely unnecessary, as when a car has a mechanical failure, it can just roll to a stop.

As for preventing accidents in the first place, car crashes, as with plane crashes, are caused overwhelmingly by operator error.

> But holding cars to the same mechanical rigors as airplanes would make them neigh unaffordable for most, to say nothing of yearly inspections many states have which is often considered nothing more than a government subsidy to mechanic shops. It is also entirely unnecessary, as when a car has a mechanical failure, it can just roll to a stop.

For engines, yes, but it's scary how many people don't maintain their brakes properly. Or otherwise drive vehicles that are unsafe to be on the road.

https://www.reddit.com/r/Justrolledintotheshop/

There's a reason why many states, particularly northern and road salt states, have mandatory inspections.

There's also very good reasons why state patrols and the DOT do mandatory inspections of larger cargo trucks.

Obviously a car does not need to be maintained to the same standard and expense as a light aircraft. If you fuck up your own clutch it's nobody's problem but your own wallet.

Brakes and tires are two legitimate inspection points that often get neglected by drivers. Still, brake inspections especially, are very often abused by mechanic shop and/or unreasonable regulation terms to force car owners to replace far before is actually necessary.

At least in my experience, it seems every single year the shop states they won't pass the car unless some of the brakes are replaced. I suspect in this case they are just bald-face lying, as recently I balked at yet another ~$600 maintenance charge to pass a routine inspection, took the car to a friends shop out of town, and was able to conclude that the original shop never even checked the brakes, as the wheels took _a lot_ of force to remove once the lug nuts were off.

Similarly with emissions checks. It often turns into nothing more than a yearly tax on your vehicle and does nothing to improve anyone's safety.

Some states require the testers and the repair shops to be separate businesses.

Costs less than $200 to setup an LLC around here. Not a major obstacle.

But holding cars to the same mechanical rigors as airplanes would make them neigh unaffordable for most

That's why most states have an emissions test (to address the negative externality that individual drivers don't feel a cost to emit pollutants, but society does pay that cost) and a safety test. Most states do not have a general mechanical test. (The definition of what part of 'mechanical' is also 'safety' is not a bright line, so if you're arguing the safety tests sometimes go too far, maybe that's right -- we could ask an actuary who has the data I suppose.)

Well consumer cars do not have 5 point harnesses, and most cars don't even have roll cage or roll resistant frame. We use a 3 point seat belt out of convenience, and we let people drive on freeways with said cars.

Also don't install a 5 point harness without a proper roll cage you will be asking for trouble.

Not related to autonomous vehicles, but related to the risk of death for light aircraft flight vs. NASA spacecraft human-rating...

The FAA seems to be relatively OK with the current annual fatality rate for light-sport aircraft, homebuilt experimental licensed aircraft, and light aircraft in the size range of the Cessna 172. For pilots with less than 1000 hours, flying a light single engine aircraft is a relatively dangerous activity on a per-hour basis.

If human spaceflight had the same number of annual fatalities as the current number deaths in the US 48 states while flying homebuilt/kit/light sport aircraft, both NASA and the news media would be freaking out.

NASA is requiring spacex to put the Falcon9 and Dragon2 through a very expensive and rigorous period of qualification before it can carry humans on commercial launch contracts. The likelihood that there will be a totally catastrophic disaster that the launch escape system can't recover from is low, but the standard of safety (in terms of flight-hours per human per year) seems to be much, much higher than the FAA's.

The Space Shuttle suffered 2 total losses in 135 missions. Apollo had 1 total loss from a dozen or so flights (plus several heart-stopping near misses). If privately flown Cessna 172s had that kind of fatality rate, I (and many other posters on this board) would long since be a smoking hole in the ground.

Space flight is spectacularly dangerous, and US launch vehicles have a particularly gnarly safety record. Some paranoia is probably warranted.

The space shuttle, Apollo and early Soyuz were also very different spacecraft than current designs and technology.

Challenger and Columbia were lost due to institutional ineptitude. The engineers that were the manufacturer of the SRBs on Challenger specifically warned against the cold temperatures and were overruled. People noticed the foam shedding issue with the space shuttle many years before the Columbia incident. The space shuttle was a catastrophic clusterfuck of a pork barrel political design. There was actually no launch escape system or possibility of recovery from early launch phase anomaly once the SRBs were lit, until they burned out (unlike the launch escape system on Soyuz, Apollo, Dragon2, etc).

The space shuttle had a ridiculous number of moving parts and degree of complexity compared to a Dragon2 on a Falcon9 (or a Soyuz on top of its booster)

The soviet union and russia have had zero fatalities since Soyuz 11 in 1971. I would actually venture that at this point it might be statistically safer to fly along as a passenger in a modern Soyuz mission than to fly as a second passenger in some sub-500-hour pilot's kitbuilt piper cub clone.

Granted that current Soyuz is the most reliable human-carrying spacecraft today (not that there's much competition). And granted that the Shuttle was deeply weird and a bad idea.

But Wikipedia tells me that the next Soyuz to go up will be #142, counting from Gagarin. In the hypothetical world where (say) the Mir crisis had been fatal, Soyuz would have a lifetime success rate almost exactly the same as the Shuttle's.

Even if the new ones are much better, we don't have the statistical power to say. Gimme the kitbuilt Piper any day.

Considering the Soyuz-U has over 700 launches and the R7 family has over 1000 (of which all Suyuz launch vehicles are part), only taking the manned versions is a somewhat distorted comparison (though not unfavorable in some aspects).

The Soyuz-success/failure manned rate is about the same as the space shuttle, though a bit higher but with successful escapes and launch aborts. If you start comparing casualties the space shuttle loses massively as only 4 cosmonauts have died, all in the decent stage (either crashed or died on entry) with the R7 and derivatives as launch vehicles and Suyuz or Voskhod as name.

To be fair, the total of the R7 family (which is skewed since there are way more failures in the early days than after the 1990's for all kinds of rockets) is 137 vs 1755 or about 7 % vs the 1.5% of the space shuttle.

> If human spaceflight had the same number of annual fatalities as the current number deaths in the US 48 states while flying homebuilt/kit/light sport aircraft, both NASA and the news media would be freaking out.

What a nonsensical statement. If light sport aircraft had as many fatalities as the current number of deaths in the US 48 states while driving/walking/biking, both the FAA and the news media would be freaking out.

And, as pointed out, spaceflight is actually significantly more dangerous than light sport aircraft, so I'm not so sure that the "standard of safety [...] seems to be much, much higher than the FAA's."

It's supposed to be a nonsensical statement: I guess what I was trying to say is that we have come to culturally "accept" a certain number of fatalities each year in commonplace activities, like walking down the street, using a bathtub, driving a car, etc.

These are not remarked upon outside of local news. Three people die in an experimental category airplane crash, makes the local news. If three people were to die in a spacecraft accident, national media would freak the fuck out.

> we have come to culturally "accept" a certain number of fatalities each year in commonplace activities, like walking down the street, using a bathtub, driving a car, etc.

I disagree; there are ongoing, successful efforts to make driving, walking down the street (see Vision Zero for example), and many more activities safer. We haven't accepted the risk levels at all.

> I disagree; there are ongoing, successful efforts to make driving, walking down the street (see Vision Zero for example), and many more activities safer. We haven't accepted the risk levels at all.

But the existence of awareness campaigns proves the point: Someone has to connect the dots on the statistics and then advocate to get people to care. Nobody needs to advocate for more coverage of spacefaring fatalities.

> awareness campaigns

These aren't campaigns to build awareness, but to change laws, training, etc.

> Someone has to connect the dots on the statistics and then advocate to get people to care

People need to know about something in order to care about it. I'm not sure what that means.

> Nobody needs to advocate for more coverage of spacefaring fatalities.

Right; those are sensational and get lots of coverage.

I think the FAA is more concerned with 500+ people on a high-explosive ballistic missile we call a 777 than they are the one or two people on a plane they built themselves. I'm not sure how you begin to compare the two.

I'm not comparing the two. I'm comparing 4-seat spacecraft to 4-seat aircraft in the kitbuilt/homebuilt/experimental licensed/light sport category.

I'd be curious to see some actual numbers here...

I ride motorcycles, and am personally fully aware of the fact that this mode of transport makes me to a first approximation 10 times as likely to die on the road compared to driving a car, and I'm happy to accept that risk for myself.

(Probably because I have enough hubris to believe I can make a significant difference there by being more skillful and less reckless than all those _other_ motorcycle riders - which I believe psychologists refer to as "illusory superiority"...)

Put autonomous cars in the exact same conditions as all cars today rather than an easy mode Silicon Valley subset and the body count will be way higher.

I don't disagree with your sentiment about the FAA. I don't believe autonomous cars are being held to a higher standard needlessly.

The FAA is a common source of criticism and complaint among the aviation community. It's the driving force behind the continually rising costs in general aviation. Owning a small single engine plane with a hull value of $40-50k requires a financial commitment tantamount to purchasing a second home.

But, it's good to step back and recognize every now and then that every Federal Aviation Regulation (FAR, 14 CFR) is written in blood. The FAA can be a pain to deal with sometimes but every pilot I know is very, very glad that they do what they do.

How often is failure of the car itself the cause of a fatal automobile accident? Just looking at lack of seatbelt use and alcohol make up 50% of traffic fatalities. It seems that the "Human Choice" is the biggest thing to fix when it comes to auto accidents.

* https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/...

I have a relative who worked at the FHWA (Federal Highway Admin) for many years. The estimate I often heard is that improper pavement design and maintenance were a factor in 20% - 30% of fatalities (think: asphalt). That friction with your tires is all that's really keeping you on the road. Many states refused to perform testing of their pavement because (a) they would have to remediate, and (b) they would likely have to go with less politically connected companies for pavement that would work better.

According to USDOT and NHTSA human error is responsible for 94% of human fatalities involving traffic crashes (including cyclist and pedestrian deaths).

If the human drives right into a huge wheel-swallowing pothole, is that "human error" or is it "improper pavement design and maintenance"?

The human should watch out. On the other hand, the pothole should be fixed.

It is probably typical to have multiple factors, making it easy to bend the numbers to one's liking.

>If driving/vehicles/roads were held to the same or a similar standards to aircraft there might be fewer accidents and deaths today.

But at what cost? Ground transportation wouldn't be very useful if we couldn't afford it

Sure, "human life is sacred" and all that but that's kind of a first world problem you don't get to care about except when you already have infrastructure that's good enough for most use cases.

>Ironically self-driving is held to a much higher standard than regular driving; which is appreciated but further makes one question why regular "dumb" vehicles are so under-regulated

Design and construction of new vehicles is subject to a large and ever increasing amount of regulation. It just doesn't make the news when a FMVSS gets updated to require that brake hoses meet a later revision of some SAE spec which specifies a more rigorous QC procedure.

Regarding maintenance, some negligible fraction (I've heard 6%, "less than 1%" and 0.5%) is caused by mechanical failure. If mechanical failure caused a meaningful portion of crashes you'd see insurance companies harping on about it and offering elective inspection programs in non-inspection states.

While there are some mechanical failure caused crashes, aren’t the vast majority of car crashes/incidents caused by driver error?

Either drunk, speeding, not paying attention, deliberately being a bellend, etc?

If anything that just means the licensing process should be stricter - not specifically a change in the cars themselves.

With self-driving cars the car is the driver so of course it has to be more strictly controlled - it has to make decisions.

> Ironically self-driving is held to a much

Is it?

Reply on news.ycombinator.com