So far Google's self-driving cars have been designed to have the driver take over in certain situations. I wonder how that works without a steering wheel and pedals. Are the cars manually drivable at all?
This wouldn't be a very interesting story if the answer were "yes." Whether Google or Toyota manufactures a self-driving car with manual override doesn't matter much from the point of view of a typical citizen or someone who's interested in the architectural implications of this technology.
As long as there's a manual override and a responsibility to use it, the car requires a capable driver paying attention all the time. It's an interesting beginning and it might improve safety but it doesn't fundamentally change things.
But without manual override, the car must be very reliably capable on its own. I suppose there might be a network connection and the ability to dispatch an Uber to pick you up whenever the car decides it is unable to proceed. But nobody will choose to use these cars if that scenario is remotely common. So announcing that this thing could ship in 2 years indicates to me that either Google's tech is doing way better than has been suggested in other recent press, or Google has started talking up vaporware to an unprecedented degree.
From a UI perspective if they rarely need manual control you can use a less ergonomic option. Such as a joystick which would be a pain for regular use but would take up less space and still provide plenty of control for situations like parking in a field for a concert and or a quick emergency override in case something goes wrong.
Yes, that's is a very interesting point. I would think that, long term, the only manual control a self driving car really needs is a "kill switch." How long will it be after self-driving tech goes mainstream until people simply don't know who to drive?
I can see my kids now, "I remember when I was a kid, my mom/dad actually had to control the automobile. Archaic!"
Not too long ago my friend's son asked me why my phone was attached to the wall. He had only ever known cordless phones. I image cars will go a similar route.
There's a funny mention of this problem in "I, Robot". When will smith takes the wheel he is scolded for taking such a completely irresponsible risk. Given how those robots drive versus how he drives, we likely will have to agree with this at some point.
I wonder if real life will play out the same way : will controlling a car the human way become an unacceptable risk, and how fast will this happen ? Because we're likely to see both effects play here : first, humans will become worse at controlling vehicles, and vehicles will become better at it, likely having superior senses and predictive ability. So there should come a point where humans become a risk.
So there will come a point where we get to choose between new policies and letting humans drive. If we had self-driving cars and we could eliminate speed limits on interstates entirely, in trade for not allowing human drivers at all, would we do it ?
I would not be at all surprised to learn that computers are better at driving now. One interesting (albeit expensive) way to test this out would be to put some computer controlled cars into demolition derby races and see how they fare in artificially hostile driving environments where the human participants are extremely risk-tolerant.
I'm very glad that California has legislation in place to allow this already, and think it's going to pay off in spades to be a leader in this area. My bet is that insurance will be baked into the price of the first generation of cars (ie subsidized by the manufacturers), liability rates will be extremely low, and there will be an avalanche of consumer switching over a fairly short period of 10-20 years.
They're citing >>300,000 miles under automatic controls with zero accidents. That's probably better than the human average. (Small sample, I know, but still.)
It looks like Americans average about 200,000 miles between accidents. I am not sure what the variance here is, e.g., depending on driver experience.
Still, the numbers are completely incomparable. Google's cars have always had a driver at the wheel, and we have no idea how many times a manual intervention has prevented an accident. Google's cars have also been driving in very limited conditions: they have not tried to drive in very bad weather, and they do not handle highway merges or difficult surface streets.
Finally, it isn't clear that "zero accidents" is completely correct. Google has said that their cars have been in accidents, but only while under manual control or while stopped at stoplights. But what does this really mean? Perhaps there was an accident caused by the self-driving car, where the human driver took control at the last minute. Or perhaps the car saw a yellow light up ahead, slammed on the brakes to stop, and was rear-ended by another car expecting them to go through. (Technically, the other car would be at fault in this situation, but it is still unsafe driving on Google's part.) There is no public information that lets us know the safety record of Google's self-driving cars.
"We’re planning to build about a hundred prototype vehicles, and later this summer, our safety drivers will start
testing early versions of these vehicles that have manual
controls."
From the stories that go into detail about google's self-driving cars, it really seems like at this point the driver-take-over situation is really more like a breakpoint than an actual safety valve. It's there for when the computer has multiple reasonable options and the driver's judgement and awareness is needed to help resolve the ambiguity not just for that ride, but for future rides as well (as a sort of feedback).
When they drive entirely on Google property (like this parking lot), they don't need controls. If they want to drive on public roads, the law requires manual control.
