Things like this are always alarming until you learn the base rate. Unfortunately, I cannot find a quick reference for this, but many many flights take off with some anomaly noted in the technical log book.
Commercial aviation would come to a halt if every aircraft had to be in 100% perfect condition for every flight. There are many systems that have redundant backups or are not essential for safe flight.
Every single plane has things that are broken, things that are inoperative, things that behave slightly out of tolerance. This is true for commercial aircraft the whole way down to single engine trainers. 172s are notorious for having fuel gauges that are basically only good for telling you how many fuel tanks you have and not anything related to quantity of fuel on the plane.
The illusion of control doesn't change your odds much.
Some would argue the per-trip comparison is invalid, but often the travel distance is not fixed, such as if you were weighing between vacation options of flying to NYC vs camping at a local campsite.
On a danger-per-hour-in-vehicle basis, airplanes of course still come out ahead, although not quite as overwhelmingly. NYC to LA is about a 5.5 hour flight; an equivalent drive would be about 350 km, and it will be very hard to match the safety of that flight even with defensive driving. You'd need to drive 70x better than average, even with the fatigue of a 5.5 hour drive.
> In 2007, the National Transportation Safety Board estimated a total of nearly 24 million flight hours. Of these 24 million hours, 6.84 of every 100,000 flight hours yielded an airplane crash, and 1.19 of every 100,000 yielded a fatal crash. https://www.psbr.law/aviation_accident_statistics.html
So we have 330M people in the US, of which let's say 100M are driving regularly. How regularly? Let's assume 2 hours a day for 52x5 = 260 working days in a year. So given that we have 43K traffic fatalities per year let's compute fatalities per hour of driving. 100M * 2 * 260 / 43K = 1.2M So we have 1 fatality per 1.2M hours of driving. At the same time we have roughly 1 fatality per 100K hours of flying. Oops!
Of course one should consider that:
(a) it's 2007 data, it's probably lower now (10 times lower?),
(b) we definitely cover longer distances per hour of flying (by the way not that much, 60 mph vs 600 mph is within 10x difference),
(c) it's probably all flying, including private, but I'm not considering just public buses either.
Add defensive driving though, and it's not that obvious which is safer.
Is it safer on average to do a long road trip, or fly? Historical crash data on long road trips (excluding commutes, local errands, etc.) probably doesn't exist, but if it did, that would be very preferable. Perhaps people crash more when driving unfamiliar roads, with additional fatigue of long durations, with additional distraction of kids, etc. Or perhaps routine drives are worse because one lets their guard down!
Of course I have to admit that flying commercial airlines is safer by average numbers, in the US and for now. But if we estimate total flying hours as 1.3B/year (http://web.mit.edu/airlinedata/www/2018%2012%20Month%20Docum... times 100 passengers per aircraft) it only takes 1300 deaths per year to make it even with average traffic fatalities. If that flight had been unlucky enough to go down we would have had 177 deaths, already not "orders of magnitude safer" than driving. And the trend is not good.
But again, we are comparing apples to oranges. Driving is a very different experience, both long and short trips. Nobody chooses to drive from Boston to LA just out of fear of flying (well, maybe there are exceptions, but "nobody" is still a very accurate word). As for short trips, changes of getting into an accident in urban area driving to the airport is probably higher than driving in the other direction towards your destination. Again, it depends.
https://www.aopa.org/-/media/files/aopa/home/training-and-sa...
The last passenger death in a US Commercial Air Carrier was in 2019: https://en.wikipedia.org/wiki/List_of_fatal_accidents_and_in...
Even assuming your 24-million flight hour number, that would mean 1 death over 100-million flight hours from 2020-2023.
If that door had hit horizontal stabilizer though we would have had a completely different statistics even with 1B hours. Fortunately it didn't happen, but with the current trend the idea that flying is always safer may become not so obvious, and "orders of magnitude" thing may disappear pretty fast.
If you look at just airlines, they’re in turn 10x _safer_ than driving if I remember correctly. There’s this anecdote that after 9/11 people were afraid to fly and died on the highways in much higher numbers. There’s also the fact that there there was a very small number of passenger deaths involving airliners in the US in over a decade (meaning no major crashes). Compared to thousands and thousands of traffic deaths a year that should drive the point home, even when you have to adjust for base rates.
That’s part of WHY air travel is safer.
Naturally that tends to push aviation towards avoidance of mechanical issues and on cars we are much more tolerant. I've seen people driving cars with their door duck taped on!
Southwest Airlines Flight 1380 was a Boeing 737-700 that experienced an uncontained engine failure[a] in the left CFM56-7B engine after departing from New York–LaGuardia Airport en route to Dallas Love Field on April 17, 2018. […] One passenger was partially ejected from the aircraft and sustained fatal injuries[…]
https://en.m.wikipedia.org/wiki/Southwest_Airlines_Flight_13...
https://en.m.wikipedia.org/wiki/List_of_fatal_accidents_and_...
The numbers would go up quite a bit if it included private and military. The numbers you linked to seem to have a very tight definition of which flights were considered, as the Wikipedia list showcases several more in-flight deaths involving air carrier class airplanes than just two.
That list includes a bunch of incidents that are not really relevant for assessing risk level when flying on a commercial airline:
- Someone committing suicide by getting sucked into a plane engine while the plane was on the ground.
- Someone sneaking onto a runway and getting struck by a plane that was landing.
- Another person stealing a plane and intentionally crashing it into the ground.
- The Kobe Bryant helicopter crash.
Looking through the list I would conclude the parent comment was correct. The only incidents with passenger fatalities on US airlines since 2009 were Southwest 1380 and PenAir 3296.
We are seeing more and more that these systems and processes have been breaking down, from regulation to manufacturing to pilots to systems to maintenance.
It really is hard to compare, but for an airplane, a passenger has to have faith that literally hundreds of thousands of people and things have done their job correctly. When I drive my car, I am much less reliant on people, systems, and processes, as cars are just plain simple. Most leople barely change their oil or check tire pressure. I even once had a complete engine failure and was able to just roll to a stop from highway speeds. Furthermore, I am in control. If I am too tired, I don't drive. I can pay attention to other drivers. I am directly responsible for maintenance. Etc.
I think it's hard to compare airplanes to cars by numbers alone. There are subtleties that are not exposed by numbers.
For an airplane passenger, it is absolutely a risk. You rely on so much happening correctly, and you are not in control of any of it. As little bits in that chain of things that need to happen don't happen correctly, percentages of failure and death go way up, and fast.
And we haven't even discussed in-flight medical emergencies, as there are actually quite a few in-flight deaths every year that would likely not yield a death if the medical emergency happened on the ground.
> There has not been a single US airline fatality since 2009.
Wikipedia says there has been 51, not counting private or military. And that list doesnt include American aircraft flying overseas, to which the MAX planes would add hundreds.
https://en.m.wikipedia.org/wiki/List_of_fatal_accidents_and_...
Yes it will be when you loose all power while going 80mph on a motorway because rodents ate the wiring 10 years later.
"tons of electronics" in modern cars is not required for the most basic functionality such as steering and braking. Yes, you loose power steering and power breaking if it goes off, but you can still drive (unless maybe you've never driven without it and it surprises you during a high speed takeover manouver etc).
This "drive by wire" takes away the most essential security feature present in "all" cars up to now. A direct mechanical connection between the steering wheel and the wheels. Even manufacturers of small airplanes (like "executive" 8 seater jets) say "no way" to this tech unless it controls an auxiliary control surface.
Putting it in a production car that will get crashes, it will corrode (yes copper wiring and insulation corrodes too), it will get attacked by rodents and it might get driven in 50 years from now as a "classic" is sheer, unabridged, stupidity of the highest order.
You’re leaving out the biggest risk: other people. Most deaths, in planes or cars, are caused by human error. In a car you’re dependent on everyone going down the road (and there may be thousands in one trip) not drifting across the median. You’re dependent on the person coming the other way at an intersection to stop. Etc.
Traffic deaths have been climbing again after decades of decline, probably due to distracted driving. Driving is much more dangerous.
A weird metric. I'd prefer travel time, not per mile.
Still far safer.
https://observer.com/1998/03/driving-versus-flying-the-debat...
But, I disagree with the article, I think the debate is hardly settled.
An example? Travel to the moon would be the safest thing ever, even if 50% of the ships exploded, because of how far it is. I bet travel to Mars would the safest thing ever, based upon miles, even if 99% of the ships exploded.
Things break based upon two things. One is maintenance per trip. And each trip has riskier parts, of which start and end are parts. Planes have issues taking off and landing, a lot more than cruising. Same for space ships. Even cars have issues at start and end of trip, if you're driving very long distances.
Airliner passenger deaths are 1 per several billion miles.
It isn't though. Airliners have suffered in-flight engine explosions and decompressions multiple times since 2001 without fatalities. The last time a structural failure of an aircraft resulted in a crash was in 2000 and it was a design that entered service in 1959. Modern airliners don't just fall out of the sky. They feature robust designs and highly competent crews.
Cars regularly crash fatally without mechanical failures at all. And that says nothing of the dire state of car maintenance among the general population.
While I agree with your point, if the scope is scheduled airline flights in the USA, AA flight 587 crashed on 2001-11-12; globally, Egypt Air 804 on 2016-05-19. There have also been a few close calls, such as Qantas 32 on 2010-11-04, and collisions can also occur, such as that between BAL flight 2937 and DHL slight 611 on 2002-07-01, and Gol Transportes Aéreos flight 1907 with a business jet on 2006-09-29. And, given the topic is things going catastrophically wrong at altitude, there is AF447 on 2009-06-01.
How can that be? Very few serious in flight mishaps STAY in the news for more than one day, but proportionally MANY mishaps that lead to death stay in the news.
This incident has juice because it's a Boing 737 series aircraft.
It may be hard to believe what I wrote here, and it would be hard to verify if you just look at the general news. You'll need to look at specialized air transport reporting to see the baseline of major mishaps.
Yes, if you want to calculate a lifetime risk, you must know (or estimate) a lifetime usage.
>So deaths per hours travelled is actually better.
It really isn't. Does the average person spend as many times more time in cars as planes are faster than cars? If so, per mile risks would be comparable to lifetime risks. The average person probably spends more time in cars than that, though that still leaves per mile risk as closer to accurate (for the average person. Are you the average person?). But neither per mile nor per hour risk should be conflated with lifetime risk, and if you're not going to use personalized assumptions about usage, it's much better to just look at actual mortality data to avoid the issue entirely. For most comparisons though, risk/cost/pollution/whatever per passenger (or for goods, ton) mile is probably by far the more useful measure. If something needs to get from A to B and you need to know what that entails or what the best option is, those are the more directly relevant figures.
Slowly but surely we see more cutting corners in aviation, especially in the US.
This ranges from trying to evade certification for planes to crew hours, to more lax regulation on how air traffic is managed to increase movements at airports, to overworked and understaffed ATC.
I don’t think it has risen to levels that affect statistics in terms of death, but the statistics in terms of near catastrophic events has risen over the years.
I have never seen a 100% serviceable aircraft, as far as I'm concerned a aircraft where everything works to spec and the spec works to needs is a myth that we strive for but can never achieve.
You're far more likely to die in a private plane or in an executive jet. GA aircraft crash constantly.
You want people incentivized to participate in safety culture, not motivated to frame others or destroy evidence to save their jobs.
People will always make mistakes and have poor judgment sometimes. That will never go away, and punishing people will not stop it. A robust safety culture expects this to happen, and builds in mechanisms to catch problems.
Of which as mentioned previously by Miraste/flight 261 regarding Alaska were not implemented clearly and cleanly given that the person who raised the alarm of ghr situation was fired and the situation continued onwards
Right now regarding boeing there are also boeing whistleblowers in new NGOs set up to police boeing malpractice... If only laws work and this weren't necessary, as boeing also has bribed faa staff to achieve their own objectives risking public safety
Once policing fails to stop their trend economic pressure is really the only thing that is left. Especially if policing failed due to said company being able to use economic power to influence the policymakers and the people tasked with enforcing those policies.
As long as people buy stuff from companies with a negative safety culture that have political pull, they dont have an incentive to change. I dont see an alternative to boycotting these dangerous malfunctioning actors.
edit: Or to get to your initial wording, this isnt about how to deal with people making errors or having poor judgement but people acting with intent for economic reasons.