6 ms·

It will eventually happen, but in order to do so, you first need a catalyst - a major metropolitan area such as Manhattan that converts to 100% robocars within

It will eventually happen, but in order to do so, you first need a catalyst - a major metropolitan area such as Manhattan that converts to 100% robocars within a geographic area. The problem is that if you mix human and robocar drivers, any car company making a robocar is going to get sued into oblivion when they collide with a human driver, regardless of who is at fault. And, there are definitely ways that humans can cause robocars to have unavoidable accidents (swerving across the double yellow line at the last split second, etc). I think that's a really good solution, honestly. Just deploy 100,000 robotaxis the size of smart cars in Midtown and ban all other vehicles. On a side note, NYC pedestrians are already pretty aggressive at jaywalking and the fact that robocars stop on a dime when they sense a pedestrian in their way will create all kinds of new ways for pedestrians to harass robotaxis.

toomuchtodo 14y ago
Doesn't even have to be a major metro area. Start small. Someplace without snow (Google cars still have problems with snow). Puerto Rico. The Bahamas. Hawaii. These are all small places where you can shake out the major issues before expanding. Also, other places to start would be Nevada (as Google is doing), Texas, and other large states where there is a benefit to having something automated do the driving for 3-6 hours across large stretches of highway. You're never going to have a major metro just ban human-driven cars overnight; you win by attrition, small bites very quickly.
- illumin8 14y ago
  Good point - better to start small. Vacation islands would be perfect as most people don't really want to drive on vacation anyway, and you don't have to worry about getting a shady taxi driver.
- callmeed 14y ago
  I don't think tourist-heavy locations like Hawaii and the Bahamas would be good. Lots rental car business and airport traffic relative to the size.
  - toomuchtodo 14y ago
    You'd be replacing those rental cars with your self-driving cars.
bokonist 14y ago
I would assume that any robocar would keep a record of its visual interface, like the airline blackbox. It would probably become much easier to judge who is at fault.
brc 14y ago
I disagree. When cars started to take over from horse and cart, it still took 20+ years for them to become commonplace. It takes time for reliability to go up, and cost to come down. It takes time for the laws to catch up. What seems like a blink in time to us now (1880s-1900) was 20 years for the people living it at the time. Cars were, at first, a freak sighting - then gradually, little by little, they caught on, and millions of horses were, ahem, retired. A big step change isn't needed at all. Just gradual introduction and refinement until they are a compelling proposition. As for the legal issue - the first generation of cars will require a human driver behind the wheel. They'll remain repsonsible for the foreseeable future. Likely the answer to long-term compensation for injured victims will be a shared insurance scheme - you pay the registration on your self-drive vehicle, and a portion of that goes into an insurance fund. If there is an accident and property/people are adversely affected, they get paid out from that fund according to a pre-determined formula ($x,000 for a limb on a 25 year old, $y,000,000 for a 45 paraplegic, etc). This isn't greatly different from schemes in place around the world today, where pinning fault on someone comes second to making sure those injured in traffic accidents are compensated + rehabilitated properly.
- kkowalczyk 14y ago
  This a very context-free analogy. The biggest reason why it took a long time for cars to replace horses was high price (simply stated: most people couldn't afford a car even if they wanted one) and lack of infrastructure (it takes years to build roads). We already have the infrastructure and the price will favor self-driving cars. The laws heavily favor safety in cars (not everyone likes seat-belts and yet we have laws requiring people to use them; we have laws banning using a cell-phone while driving, driving while drunk; cars have to build in a way that withstand crashes etc.). As long as it can be demonstrated that self-driving cars are safer (and they are), there will not only be pressure to allow them but to require them and eventually ban human drivers altogether. Once the self-driving cars are allowed, the lower price and greater convenience will make it a blockbuster. I don't want to own a car, I don't want to operate it, I just want to get from Mission to SOMA fast and cheaply and self-driving cars will be able to offer that.
  - 14y ago
philwelch 14y ago
Maybe Singapore's a good bet--combination of a major metropolitan area and a technocratic dictatorship that might go for it.
- jasonlingx 14y ago
  Hah. Unlikely. We still tax diesel cars several times more then petrol cars because we think they are dirtier. This is only just beginning to change. Tesla gave up here because of the frosty reception from our government. The authorities here tend to adopt a wait and see attitude rather than pushing the boundaries on things like this.
  - brc 14y ago
    Diesel cars are dirtier for particulate and Nox emissions.
    - jasonlingx 14y ago
      exactly the kind of misinformation I was talking about...
ryannielsen 14y ago
any car company making a robocar is going to get sued into oblivion when they collide with a human driver, regardless of who is at fault I'd actually be somewhat surprised if this were the case. Self-driving cars will gather an enormous amount of sensor data and I bet they'll store that and decision tree logs in a "black box" of sorts. If this isn't being done now, I guarantee it'll be done after the first few lawsuits. With that info in hand, it should be rather clear – at least in the majority of cases – who's at fault. The human driver can claim they swerved over the double line to avoid an obstacle, but the sensor information can probably confirm whether that's actually the case. This isn't to say there won't be lawsuits, or that self-driving cars will never be at fault. I just expect there will be a rash of initial lawsuits. Then, as it becomes clear that the data recorded by self-driving cars makes them a difficult target to exploit (and assuming self-driving cars are in fact less error prone than human drivers), I expect lawsuits to taper off.
- hyperbovine 14y ago
  You are exactly right -- no company with half an ounce of brains is going to make a robocar that does anything except obey every traffic law on the books. Barring a software failure it's almost inconceivable to me that said cars even could cause an accident! But the flip side is much more damning: congratulations, you're now driving the speed limit everywhere you go. You're following at five car lengths. You stop for yellows and anybody within a quarter mile of a crosswalk. Basically, you drive like a Florida octogenarian, everyone else on the road hates you and it takes longer to go places. I can see a few die-hards going for that proposition, but not many.
  - humbledrone 14y ago
    Actually, a software-driven car could follow cars much closer than (mph / 10) car lengths. That particular rule of thumb is based off of human reaction times. A safety buffer is still a good thing, but software will react several orders of magnitude faster than a person. Also, if software-driven cars become common enough, you will probably get places a lot faster, at least in heavy traffic. Traffic flow could be much better than it is with humans driving. There have been several studies on freeway gridlock traffic that show that small driving mistakes (e.g. slowing down more abruptly than necessary) can butterfly-effect out into massive bottlenecks. And of course if software-driven cars cause fewer accidents, there will be less gridlock due to those as well.
maratd 14y ago
> the fact that robocars stop on a dime when they sense a pedestrian in their way will create all kinds of new ways for pedestrians to harass robotaxis. Meh. I can solve this with one line of code. Every programming language has a random function =) As a native, I can tell you that this is the solution employed by most human driven taxis already.
jasonlingx 14y ago
Why would robocar companies be sued if they are not at fault?

toomuchtodo 14y ago

Doesn't even have to be a major metro area. Start small. Someplace without snow (Google cars still have problems with snow). Puerto Rico. The Bahamas. Hawaii. These are all small places where you can shake out the major issues before expanding. Also, other places to start would be Nevada (as Google is doing), Texas, and other large states where there is a benefit to having something automated do the driving for 3-6 hours across large stretches of highway. You're never going to have a major metro just ban human-driven cars overnight; you win by attrition, small bites very quickly.

illumin8 14y ago
Good point - better to start small. Vacation islands would be perfect as most people don't really want to drive on vacation anyway, and you don't have to worry about getting a shady taxi driver.
callmeed 14y ago
I don't think tourist-heavy locations like Hawaii and the Bahamas would be good. Lots rental car business and airport traffic relative to the size.
- toomuchtodo 14y ago
  You'd be replacing those rental cars with your self-driving cars.

bokonist 14y ago

I would assume that any robocar would keep a record of its visual interface, like the airline blackbox. It would probably become much easier to judge who is at fault.

brc 14y ago

I disagree. When cars started to take over from horse and cart, it still took 20+ years for them to become commonplace. It takes time for reliability to go up, and cost to come down. It takes time for the laws to catch up. What seems like a blink in time to us now (1880s-1900) was 20 years for the people living it at the time. Cars were, at first, a freak sighting - then gradually, little by little, they caught on, and millions of horses were, ahem, retired. A big step change isn't needed at all. Just gradual introduction and refinement until they are a compelling proposition. As for the legal issue - the first generation of cars will require a human driver behind the wheel. They'll remain repsonsible for the foreseeable future. Likely the answer to long-term compensation for injured victims will be a shared insurance scheme - you pay the registration on your self-drive vehicle, and a portion of that goes into an insurance fund. If there is an accident and property/people are adversely affected, they get paid out from that fund according to a pre-determined formula ($x,000 for a limb on a 25 year old, $y,000,000 for a 45 paraplegic, etc). This isn't greatly different from schemes in place around the world today, where pinning fault on someone comes second to making sure those injured in traffic accidents are compensated + rehabilitated properly.

kkowalczyk 14y ago
This a very context-free analogy. The biggest reason why it took a long time for cars to replace horses was high price (simply stated: most people couldn't afford a car even if they wanted one) and lack of infrastructure (it takes years to build roads). We already have the infrastructure and the price will favor self-driving cars. The laws heavily favor safety in cars (not everyone likes seat-belts and yet we have laws requiring people to use them; we have laws banning using a cell-phone while driving, driving while drunk; cars have to build in a way that withstand crashes etc.). As long as it can be demonstrated that self-driving cars are safer (and they are), there will not only be pressure to allow them but to require them and eventually ban human drivers altogether. Once the self-driving cars are allowed, the lower price and greater convenience will make it a blockbuster. I don't want to own a car, I don't want to operate it, I just want to get from Mission to SOMA fast and cheaply and self-driving cars will be able to offer that.
- 14y ago

philwelch 14y ago

Maybe Singapore's a good bet--combination of a major metropolitan area and a technocratic dictatorship that might go for it.

jasonlingx 14y ago
Hah. Unlikely. We still tax diesel cars several times more then petrol cars because we think they are dirtier. This is only just beginning to change. Tesla gave up here because of the frosty reception from our government. The authorities here tend to adopt a wait and see attitude rather than pushing the boundaries on things like this.
- brc 14y ago
  Diesel cars are dirtier for particulate and Nox emissions.
  - jasonlingx 14y ago
    exactly the kind of misinformation I was talking about...

ryannielsen 14y ago

any car company making a robocar is going to get sued into oblivion when they collide with a human driver, regardless of who is at fault I'd actually be somewhat surprised if this were the case. Self-driving cars will gather an enormous amount of sensor data and I bet they'll store that and decision tree logs in a "black box" of sorts. If this isn't being done now, I guarantee it'll be done after the first few lawsuits. With that info in hand, it should be rather clear – at least in the majority of cases – who's at fault. The human driver can claim they swerved over the double line to avoid an obstacle, but the sensor information can probably confirm whether that's actually the case. This isn't to say there won't be lawsuits, or that self-driving cars will never be at fault. I just expect there will be a rash of initial lawsuits. Then, as it becomes clear that the data recorded by self-driving cars makes them a difficult target to exploit (and assuming self-driving cars are in fact less error prone than human drivers), I expect lawsuits to taper off.

hyperbovine 14y ago
You are exactly right -- no company with half an ounce of brains is going to make a robocar that does anything except obey every traffic law on the books. Barring a software failure it's almost inconceivable to me that said cars even could cause an accident! But the flip side is much more damning: congratulations, you're now driving the speed limit everywhere you go. You're following at five car lengths. You stop for yellows and anybody within a quarter mile of a crosswalk. Basically, you drive like a Florida octogenarian, everyone else on the road hates you and it takes longer to go places. I can see a few die-hards going for that proposition, but not many.
- humbledrone 14y ago
  Actually, a software-driven car could follow cars much closer than (mph / 10) car lengths. That particular rule of thumb is based off of human reaction times. A safety buffer is still a good thing, but software will react several orders of magnitude faster than a person. Also, if software-driven cars become common enough, you will probably get places a lot faster, at least in heavy traffic. Traffic flow could be much better than it is with humans driving. There have been several studies on freeway gridlock traffic that show that small driving mistakes (e.g. slowing down more abruptly than necessary) can butterfly-effect out into massive bottlenecks. And of course if software-driven cars cause fewer accidents, there will be less gridlock due to those as well.

maratd 14y ago

> the fact that robocars stop on a dime when they sense a pedestrian in their way will create all kinds of new ways for pedestrians to harass robotaxis. Meh. I can solve this with one line of code. Every programming language has a random function =) As a native, I can tell you that this is the solution employed by most human driven taxis already.

jasonlingx 14y ago

Why would robocar companies be sued if they are not at fault?