Self-Driving Cars

Would You Buy A Self-Driving Car?

  • Yes

    Votes: 51 14.2%
  • Maybe

    Votes: 44 12.2%
  • No

    Votes: 265 73.6%

  • Total voters
    360
"The accident occurred when the driver of a second vehicle "failed to yield" to the Uber vehicle while making a turn, said Josie Montenegro, a spokeswoman for the Tempe Police Department."

Sensationalism by association?
I was aware of that but I left it out there as bait. Obviously the self-driving car was incapable of analyzing and accommodating a fairly common real world situation.
 
Yes, things already fail when it is easy. Now, what if the traffic situation gets complicated.

Of course, it is easy to judge from behind a computer, because I can not judge if the self driven car did have any chance to avoid a collision.

A good driver however, probably foresees the behavior of others making mistakes, and simply avoids the no-escape-possible situation.
 
The idea of investing massively in the technology and adoption of self-driving cars rather than efficient, comfortable, and reliable public transportation is so ludicrous that I don't know whether to laugh or cry...
 
I take the Amtrak to a town about 2 hours away every 2 or 3 months. When I do two things always amaze me. How much nicer it is than driving (or being in a self-driving car on a freeway) and how ridiculous it is that it costs more than driving my car, which is not great on gas mileage.
 
Ever consider that this could have been a situation that you couldnt avoid 99.999% of the time? The above quote is in reference to #174 which is general/could mean many things. "Failed to yield", like many things in life, plots out as a bell curve.
First I considered the obvious - that the "guilty" car turned in front of the "smart" car and the "smart" car missed several routine visual cues and failed to take evasive action. One of the best ways to avoid accidents is to pay attention and avoid the accidents that the "other guy" causes.
 
Obviously the self-driving car was incapable of analyzing and accommodating a fairly common real world situation.

First I considered the obvious - that the "guilty" car turned in front of the "smart" car and the "smart" car missed several routine visual cues and failed to take evasive action.
Seems to me you like to play pretty fast and loose with the term 'obvious' here. Based upon the article and presumably the police report, there's nothing to suggest that a human driver would/could have avoided this collision. Your assumption and misdirected blame here is unfounded malarkey.

I was aware of that but I left it out there as bait.
:rolleyes: You clever devil you. Do continue to have fun on your fishing expedition. Bon voyage....:bye:

 
Last edited:
Seems to me you like to play pretty fast and loose with the term 'obvious' here. Based upon the article and presumably the police report, there's nothing to suggest that a human driver would/could have avoided this collision.
The suggestion came to me from decades of alert driving experience. YMPV - your mileage probably varied.
 
It needs to be a dead child to trigger further investigation, I guess.
Still, it proves also simple tasks are beyond control.

Of course, my comment only applies if I am not lured by some april 1 joke.
 
Last edited:
In an attempt to see what was going on, I googled "what is going on with self driving cars?" and got this recent news.

http://www.npr.org/sections/thetwo-...olls-out-new-guidelines-for-self-driving-cars
Interesting. Seems to be about the safety and regulation of the development cycle of this technology. Probably cuz the technology is so unusually shaky to begin with and involves putting public safety at risk. So the politics don't advance or validate the technology - they just attempt to deal with it.
 
As long as the car companies dont follow the "no way in hell is your lid gonna be compatible with my tupperware" model, we should be fine...:)
But even in that scenario the lid was fully functional as a lid, and the incompatible container was also fully functional as a container. Neither of the two key components of autonomous driving - detection and analysis - are anywhere near being fully functional. Nor will they be until some serious breakthroughs are made in machine "cognizance" if that's the right term.
 
But even in that scenario the lid was fully functional as a lid, and the incompatible container was also fully functional as a container. Neither of the two key components of autonomous driving - detection and analysis - are anywhere near being fully functional. Nor will they be until some serious breakthroughs are made in machine "cognizance" if that's the right term.

Interesting point. I wonder when they'll get to the point where they recognize that the cars would be better off communicating with each other, and also with some specific 'autonomous vehicle sensors' built into the roads or signage. It seems as though everybody is focused on building systems that result in fully 'aware' singular car, where I think they'd be much better off building a more cooperative system.

I've got a colleague who just started in an r&d position at one of the autonomous vehicle groups out here. His guestimate is that all the easy stuff has been done, but it will take another 5-10 years to cover all the edge cases. There's still no question in my mind that it's only a matter of time, but in the mean time, I wonder why they are letting these things loose on the road in their current state.

bs
 
One edge case they'll probably never solve is the flooded pothole. Discriminating between a shallow puddle and a nasty pothole filled to the brim with water is something people just develop a "sixth sense" for. Another one is "reading" a deer's mind who just seems to be hanging out beyond the shoulder. The GM Proving Ground in Milford MI had (and still probably has) a wide variety of real world road types and conditions. They need a place like that where the cars can deal with all the conditions (and each other).
 
dupe post. but....

reaction times will be less than the average human.

better awareness of surroundings and thus the ability to drive much more defensively

commute times will be lessened

do humans typically get injured or die when they hit a deer?
You're WAY overestimating the importance of hardware (we do that well) and WAY underestimating the importance of algorithm (we're actually quite stupid). So you put stupid algorithms "behind the wheel" of sophisticated hardware and what happens? Yes - sophisticated hardware gets into stupid accidents.
 
Back
Top Bottom