Yes, I have my eye on one of these little beauts, although it’s hard to judge the exact scale.

Presumably the strange looking pedal-like assembly underneath is some kind of advanced retractable landing gear.

David

Think about it: if your automatic car crashes into a bus full of catholic schoolgirls and nuns, killing them all in a horrible fireball, who is at fault? There is no way the car companies will want to wear that kind of publicity or liability.

Speaking of publicity… TimG points out that about 42 thousand Americans die each year in car crashes. If we assume that the first iteration of self-driving cars aren’t perfect and te death toll is reduced to about 20 thousand, which headline will sell more papers?

“Self-Driving Cars Cut Death Toll in Half”
or
“Phycho Robot Cars Kill 20 Thousand”

I work in image processing and my professional opinion is: IF we could outlaw humans driving (and really get it enforced), then twenty/twenty-five years we could get it developed and deployed (deployment actually takes much longer than you’d think, so twenty years is “right around the corner”). IF we can’t outlaw human driving then I’d put it at least at the forty year mark. The reason for this is the “New York cabbie problem”: everyone who drives in New York knows that New York cabbies do truly insane things, so upon seeing a yellow cab they hang back a little to deal with problems. But the cabbie’s know this, so they know there are maneuvers they can pull using this knowledge, and push things as much as they can without getting attention from the police. The cabbie’s actively use the way that people view them to increase their driving options.

In the context of self-driving cars, if there are no human drivers then you can assume that other cars will either behave sensibly or they’re malfunctioning and will be behaving “randomly”. This is a hard, but do-able problem. What they aren’t going to do is try and take advantage of the fact that a program, designed to avoid killings, injuries and other damages, is driving the car. But if there are also human drivings, some of them will decide “since that self-driving car should be able to respond quickly in order to avoid a crash, I’ll cut in front with minimal clearance”. Likewise the envelope of various other “self-centred” driving techniques will get pushed right to the edge. Since just a few deaths that appear to be caused by self-driving cars will cause a backlash, any companies that want to deploy them will need to have the controlling programs completely capable of dealing with an intelligent and actively “hostile” other human drivers whilst still giving a reasonable ride. (If the car emergency stops every five minutes to avoid a human gaming the system it’s not something people will pay for.) This is a much harder problem (indeed the 40 years incorporates a guess of increasing “artificial intelligence” productivity as the years go by, it’s a more than twice as difficult problem.)

I’d actually expect things like surgical robots and robots for helping the elderly to be deployed sooner than self-driving cars precisely because you can safely assume all the humans involved will be co-operating rather than gaming the system.

At last trying to put thinking and technologyinto some ‘practical’ use. But Why have the car wait for you at some central location.
If I could go down the pub, restaurant, club or airport, be dropped off by the car which either selects the nearest parking space (or returns home to my garage) and then comes and picks me up – Great! No need for the wife to go teetotal on a night out.

Mind you I don’t think Taxi Drivers or Limo Drivers would be too happy. Yet another redundant career?

Biggest problem with automated driving is emergent mass behavior, of course. When you have large numbers of autonomous agents following simple rules en masse, big things happen, and not always what you’d expect. It’s true enough of human traffic patterns (seen the paper on shockwave modelling of traffic jams? great stuff) that when you turn the agents into robots, I would expect some rather….interesting behavior to emerge.

lol actually I just had an image of The Game of Life being recreated in traffic patterns as seen from overhead, with repeating structures cascading across the grid, etc.

People wll get used to having it, use it liberally and we’d figure out oh, we are safer in general this way. Can you imagine the efficiency increase?

I’m a car enthusiest and I too love sitting behind the wheel but don’t tell me a majority of people would not be better off with automated cars.

Drivers drive cars because they’re crazy individualists. And of course we all know that most drivers consider themselves better than average.

They’re never gonna give up control.

Sure it could easily be implemented, but the market won’t be there for it.

At least not in the US as said by LabLemming. I can’t even imagine it here in Denmark.

Sad. But true.