6 ms·

Isn't it going to be easier to be safe on the sky as there's much more space, i.e. collision rate will be much much lower?

jacobolus 7y ago
Small planes flown by amateurs are extremely unsafe compared to most kinds of transportation.
- A2017U1 7y ago
  It's about the same if not less dangerous than riding a motorcycle. https://inspire.eaa.org/2017/05/11/how-safe-is-it/ https://inspire.eaa.org/2017/05/11/how-safe-is-it/
  - jacobolus 7y ago
    Riding a motorcycle is extremely unsafe.
- dreamcompiler 7y ago
  It's highly correlated with pilot experience, and it's not monotonic. Pilots with 20 hours are pretty safe because they are scared of everything. Pilots with 2000 hours are extremely safe because of their experience. Pilots with 200 hours are quite dangerous because they're complacent and overconfident. https://www.amazon.com/Killing-Zone-Second-How-Pilots/dp/0071798404/ref=mp_s_a_1_2?keywords=killing+zone https://www.amazon.com/Killing-Zone-Second-How-Pilots/dp/007...
crispyambulance 7y ago
> collision rate will be much much lower? I agree that colliding with other flying cars is going to be rare. The issue is that when there's a failure of any kind, you can't just "pull-over" in a flying car. You need to land. Flying cars just aren't going to be a big market, IMHO, in the foreseeable future. Pilots need to be qualified, and even then, they need to be in a place that can truly accommodate flying vehicles-- the middle of nowhere. And if you're going to that, you might as well buy an actual plane rather than an all-in-one hot mess of a flying car.
- ghaff 7y ago
  >The issue is that when there's a failure of any kind, you can't just "pull-over" in a flying car. You need to land. Honestly, this (and noise/general nuisance factor) would seem to be the bigger issue than reliable autopilots, which presumably can be gotten to over time. If you scale up personal flying vehicles to mainstream use, even with a mandatory costly and rigorous maintenance schedule, it's hard not to believe that you'd have lots of these dropping out of the sky and causing fatalities and damage on a daily basis.
brownbat 7y ago
There's a lot of open space available, but transportation doesn't evenly distribute itself among the available space. There will be clustering points because people want to go the same places or use the same thoroughfares, and those areas will have higher risks of collisions. Before the 60s, collisions were rare per mile traveled, but a roughly yearly occurrence across major commercial airlines. It took a lot of policy and engineering work to reduce their frequency: https://en.wikipedia.org/wiki/Mid-air_collision https://en.wikipedia.org/wiki/Mid-air_collision Automated systems might be able to fix this? They wouldn't have to deal with the paper bag problem, just avoiding every obstacle, so that's a plus. But sudden braking isn't the near universal solution or mitigation that it is with driving, and predicting the movement of other airborne objects is really hard, so those are minuses. Beyond collision prevention, there's an interesting general question in this thread about whether automated flying or automated driving is a more difficult engineering problem. I can see arguments for both sides, would love to hear a really thorough analysis, preferably by someone who has worked on either or both problems.

jacobolus 7y ago

Small planes flown by amateurs are extremely unsafe compared to most kinds of transportation.

A2017U1 7y ago
It's about the same if not less dangerous than riding a motorcycle. https://inspire.eaa.org/2017/05/11/how-safe-is-it/ https://inspire.eaa.org/2017/05/11/how-safe-is-it/
- jacobolus 7y ago
  Riding a motorcycle is extremely unsafe.
dreamcompiler 7y ago
It's highly correlated with pilot experience, and it's not monotonic. Pilots with 20 hours are pretty safe because they are scared of everything. Pilots with 2000 hours are extremely safe because of their experience. Pilots with 200 hours are quite dangerous because they're complacent and overconfident. https://www.amazon.com/Killing-Zone-Second-How-Pilots/dp/0071798404/ref=mp_s_a_1_2?keywords=killing+zone https://www.amazon.com/Killing-Zone-Second-How-Pilots/dp/007...

crispyambulance 7y ago

> collision rate will be much much lower? I agree that colliding with other flying cars is going to be rare. The issue is that when there's a failure of any kind, you can't just "pull-over" in a flying car. You need to land. Flying cars just aren't going to be a big market, IMHO, in the foreseeable future. Pilots need to be qualified, and even then, they need to be in a place that can truly accommodate flying vehicles-- the middle of nowhere. And if you're going to that, you might as well buy an actual plane rather than an all-in-one hot mess of a flying car.

ghaff 7y ago
>The issue is that when there's a failure of any kind, you can't just "pull-over" in a flying car. You need to land. Honestly, this (and noise/general nuisance factor) would seem to be the bigger issue than reliable autopilots, which presumably can be gotten to over time. If you scale up personal flying vehicles to mainstream use, even with a mandatory costly and rigorous maintenance schedule, it's hard not to believe that you'd have lots of these dropping out of the sky and causing fatalities and damage on a daily basis.

brownbat 7y ago

There's a lot of open space available, but transportation doesn't evenly distribute itself among the available space. There will be clustering points because people want to go the same places or use the same thoroughfares, and those areas will have higher risks of collisions. Before the 60s, collisions were rare per mile traveled, but a roughly yearly occurrence across major commercial airlines. It took a lot of policy and engineering work to reduce their frequency: https://en.wikipedia.org/wiki/Mid-air_collision https://en.wikipedia.org/wiki/Mid-air_collision Automated systems might be able to fix this? They wouldn't have to deal with the paper bag problem, just avoiding every obstacle, so that's a plus. But sudden braking isn't the near universal solution or mitigation that it is with driving, and predicting the movement of other airborne objects is really hard, so those are minuses. Beyond collision prevention, there's an interesting general question in this thread about whether automated flying or automated driving is a more difficult engineering problem. I can see arguments for both sides, would love to hear a really thorough analysis, preferably by someone who has worked on either or both problems.