Hacker News new | past | comments | ask | show | jobs | submit login

> According to one estimate, for every 1 billion passenger miles travelled by car, 7.2 people die; by plane, it's 0.07 people.

Interesting to read concrete numbers for this comparison. It's well-known that air travel is much safer than cars per mile, but I didn't realize it was over 100 times safer.

I'd be curious to know what this number is for fully self-driving cars once they start being used for the commercial transport of passengers (as opposed to just testing/refining the technology).




How does this change if you look at it on number of trips?

If I get in a plane, I'm probably going hundreds or thousands of miles. But in a car it might be only a few miles.

If I flew everywhere I would go by car, is it still safer?


Wikipedia gives 22 deaths per million flights for 2002 to 2011, while england and wales have about 5 traffic deaths per day and about 15 million people commuting daily by car.

The car numbers are very rough, but flying will stay at least one order of magnitude more dangerous than driving by this measure, which makes sense: large heights are dangerous.


Also, "if you flew everywhere you would go by car", you wouldn't feasibly be flying in airliners for 5-mile trips, but in helicopters or GA craft, which are statistically way more dangerous.


A plane is also much faster than a car, so the same comparison in terms of hours instead of miles would probably add a couple orders of magnitude.


It’s also higher variance. When a plane goes down, it’s often the case that everyone on board dies. Traffic collisions can sometimes lead to multi-vehicle pileups but they’re unlikely to cause hundreds of deaths all at once like a plane crash.


>it’s often the case that everyone on board dies.

But not always, interestingly enough. The lede mentions the Tenerife disaster, which had 61 survivors out of 644, and Avianca Flight 52, which had 85 survivors out of 158. Most accidents occur during takeoff and landing, when altitude and relative velocity are low.

Disasters caused by pure incompetence, like when the pilot of Air France 447 flew a fully functional plane into the Atlantic, are rare. So rare that the United States hasn't had a fatal crash of an airliner since 2009.


I don't know if I would refer to that as "variance" in the safety. A person can only die once, which puts a hard upper-bound on that.

Variance in the death rate of events, maybe.


If you measure by distance traveled, flying is safer. ...but if you measure by time traveling, driving is safer.

Source: http://www.meretrix.com/~harry/flying/notes/safetyvsdriving....




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: