Robots can get driver's license in Nevada

[Pythagorean]

http://singularityhub.com/2012/02/2...-plate-in-nevada-means-the-driver-is-a-robot/

"An extended campaign in Nevada by Google has led to a new host of provisions which will allow automated cars to legally drive in the state. Starting March 1st, 2012 innovators like Google can officially apply for a new kind of robot driver’s license that will give them permission to openly test their cars on the road"

 

[Ivan Seeking]

http://singularityhub.com/2012/02/2...-plate-in-nevada-means-the-driver-is-a-robot/

"An extended campaign in Nevada by Google has led to a new host of provisions which will allow automated cars to legally drive in the state. Starting March 1st, 2012 innovators like Google can officially apply for a new kind of robot driver’s license that will give them permission to openly test their cars on the road"

When I talk with elderly people, one of the most common complaints is the loss of independence when they lose the ability to drive. Hopefully technology will progress faster than my age.

 

[HowardVAgnew]

This sounds a bit scary on some levels, but if closed track tests are successful test, then yeah, its time to take it to the next level by testing on the open road, with careful monitoring and control, of course.

 

[Office_Shredder]

This sounds a bit scary on some levels, but if closed track tests are successful test, then yeah, its time to take it to the next level by testing on the open road, with careful monitoring and control, of course.

Google's driven 200,000 miles on open roads in California (about 140,000 before anybody even knew what they were doing)

 

[Ryan_m_b]

I can't remember exactly who said this (a top guy in GM perhaps?) but I read recently that in the last decade there has been a substantial development in the technology behind self-driving cars.

Personally I think this is fantastic. There was always going to come a point where testing needed to go out into the world (beyond the odd experiment by one group). Hopefully widespread testing in a safe and proper manner will bring about the widespread adoption of this technology sooner rather than later.

I'm very interested to see the full ramifications of self driving cars. How will it affect the haulage/freight industries? How will it affect personal car ownership? How will it affect public transport? How will it affect our opinion of machine skill vs people skill? How will it affect town/infrastructure planing? How will the automotive industry change?

 

[ThomasT]

http://singularityhub.com/2012/02/2...-plate-in-nevada-means-the-driver-is-a-robot/

"An extended campaign in Nevada by Google has led to a new host of provisions which will allow automated cars to legally drive in the state. Starting March 1st, 2012 innovators like Google can officially apply for a new kind of robot driver’s license that will give them permission to openly test their cars on the road"

I think this might turn out to be a very bad idea.

 

[dipole]

I think this might turn out to be a very bad idea.

Thanks for the opinion, care to justify it?

 

[Ivan Seeking]

Thanks for the opinion, care to justify it?

Next comes the right to vote?

 

[rootX]

I can see robots outperforming humans in driving. I hope things don't go bad like they happened in case of Toyota recently. Even small malfunctions can get massive bad PR and undo all the progress.Instead of getting full driving license, I guess I will just wait for my robot car :!) I have been waiting for that for quite a few years now.

 

[dipole]

I don't understand your point Ivan, but maybe you're joking. We have machines that operate autonomously all the time, trains are capable of driving themselves, airplanes can fly on auto-pilot, most of the technology and goods you own were probably built or assembled using robots at some stage of their manufacture...

If machines can offer a real improvement to human transportation by driving on roads autonomously, why is this a bad thing? Is a robot really less trustworthy than some teenager who's texting/blaring music/chatting with friends/and being reckless all at the same time while driving? We allow them to drive, and in fact the number one cause of deaths among young people is car accidents.

Naysayers really need to justify their opposition to this sort of technology.

 

[Ivan Seeking]

I don't understand your point Ivan, but maybe you're joking.

Yes.

 

[Office_Shredder]

Any time my dad and I talk about welfare politics he sticks with the argument that 'they should just become truck drivers, that pays a decent wage'. I wonder if he's at risk of being out of an argument

 

[TheStatutoryApe]

I can't remember exactly who said this (a top guy in GM perhaps?) but I read recently that in the last decade there has been a substantial development in the technology behind self-driving cars.

Personally I think this is fantastic. There was always going to come a point where testing needed to go out into the world (beyond the odd experiment by one group). Hopefully widespread testing in a safe and proper manner will bring about the widespread adoption of this technology sooner rather than later.

I'm very interested to see the full ramifications of self driving cars. How will it affect the haulage/freight industries? How will it affect personal car ownership? How will it affect public transport? How will it affect our opinion of machine skill vs people skill? How will it affect town/infrastructure planing? How will the automotive industry change?

I think it was about a decade ago that I read an article about self driving cars developed in Japan. I think that they had only done very limited road testing, mostly for demonstration purposes.

 

[CaptFirePanda]

It has the potential of reducing traffic fatalities by one for every car involved

 

[Office_Shredder]

It has the potential of reducing traffic fatalities by one for every car involved

This seem unlikely, unless everybody else has a chauffeur and I'm just a chump

 

[HowardVAgnew]

Naysayers really need to justify their opposition to this sort of technology.

Admittedly, I have not followed the topic religiously, but as I confess to being a 'naysayer' -- not a denier, but a "sayer" with some "nays," I would like to communicate my concerns. Humans do stupid things, and certainly the roadway is no exception, but until I see some convincing proof, so far as I am aware, the human brain is far more capable of handling dynamic situations than an artificial computer.

How well can a robot driver identify what is ahead? Can it analyze the erratic driving of another car caused by another driver and determine the likelihood that it is being operated by someone too intoxicated or exhausted to drive predictably, and adjust a safety margin around that vehicle to compensate? Can it see and analyze a puddle, determine the risk of hydroplaning causing loss of control, and adjust driving for safety around it? Can it determine when an obstruction on a roadway or other conditions necessitate temporarily passing the obstruction into oncoming lanes when it is safe to do so? Can it read temporary 'detour' signs diverting traffic from, say, one side of a freeway to the other that cause modern GPS navigation systems to determine you are driving the 'wrong way' and must 'turn around?' Can it react to material falling off a truck or rock at freeway speeds? Can it realize when getting to a destination is impossible because of severe weather conditions, completely stopped traffic or other hurdles and it is time to give up and go home instead?

As I said, I have not followed autonomous drivers, so I do not know the answers to these. Is the answer to each absolutely, yes, the robot driver can handle all of these situations at least as well as a human driver?

 

[CaptFirePanda]

This seem unlikely, unless everybody else has a chauffeur and I'm just a chump

'Twas just a bit of a joke.

 

[Office_Shredder]

Actually I'm a bit intrigued of what percent of the time the driver isn't required. For example every time you need to drive your kid someplace, you wouldn't need to do that anymore. Anytime you need to drop a package off someplace, you can just send the car to do that for you. But would people trust their vehicle to do these things? Probably you wouldn't let your car drive your kids anywhere without you seeing how people won't even let their kids play in the yard unsupervised anymore

 

[wuliheron]

They require a bond of 1-3 million dollars to drive your robotic car, but that certainly might be worth it for trucks. A driver that only has to stop for gas and repairs? It might give a whole new meaning to "truck stop".

In another ten years IBM should have their second neuromorphic chip and then we'll see the real terminators start to roll of the assembly line. Robotic cars that learn from experience going to simulated driving school. Cars marketed for the congenial personality of the robotic chauffeur. I just hope the economy models don't demand too much attention.

 

[Gokul43201]

Actually I'm a bit intrigued of what percent of the time the driver isn't required. For example every time you need to drive your kid someplace, you wouldn't need to do that anymore. Anytime you need to drop a package off someplace, you can just send the car to do that for you. But would people trust their vehicle to do these things?

How awesome would it be to send your car to grab your Chinese take-out from the restaurant down the street (when you could have walked instead)?

 

[Office_Shredder]

1) There's no way I trust the place to get the order right without me there to verify it :P

2) If everyone starts doing it, you have a single location where unsupervised vehicles pull up unlocked waiting for somebody to put something in it? Just hop in and drive away man.

 

[MarcoD]

Admittedly, I have not followed the topic religiously, but as I confess to being a 'naysayer' -- not a denier, but a "sayer" with some "nays," I would like to communicate my concerns. Humans do stupid things, and certainly the roadway is no exception, but until I see some convincing proof, so far as I am aware, the human brain is far more capable of handling dynamic situations than an artificial computer.

How well can a robot driver identify what is ahead? Can it analyze the erratic driving of another car caused by another driver and determine the likelihood that it is being operated by someone too intoxicated or exhausted to drive predictably, and adjust a safety margin around that vehicle to compensate? Can it see and analyze a puddle, determine the risk of hydroplaning causing loss of control, and adjust driving for safety around it? Can it determine when an obstruction on a roadway or other conditions necessitate temporarily passing the obstruction into oncoming lanes when it is safe to do so? Can it read temporary 'detour' signs diverting traffic from, say, one side of a freeway to the other that cause modern GPS navigation systems to determine you are driving the 'wrong way' and must 'turn around?' Can it react to material falling off a truck or rock at freeway speeds? Can it realize when getting to a destination is impossible because of severe weather conditions, completely stopped traffic or other hurdles and it is time to give up and go home instead?

As I said, I have not followed autonomous drivers, so I do not know the answers to these. Is the answer to each absolutely, yes, the robot driver can handle all of these situations at least as well as a human driver?

Can it handle a child running after a ball it kicked to the other side of the street? I don't think so.

 

[Pythagorean]

1) There's no way I trust the place to get the order right without me there to verify it :P

2) If everyone starts doing it, you have a single location where unsupervised vehicles pull up unlocked waiting for somebody to put something in it? Just hop in and drive away man.

Naw, it will have a cargo compartment that you will give a temporay acces # to the restraunt for. No unlocked doors

 

[Pythagorean]

(that was meant as a prediction, not a statement of fact)

 

[Office_Shredder]

Can it handle a child running after a ball it kicked to the other side of the street? I don't think so.

Yes of course it can.

Pythagorean, yeah that's pretty obvious I don't know why I thought they would have to sit it in the front seat or something

 

[MarcoD]

Yes of course it can.

Pythagorean, yeah that's pretty obvious I don't know why I thought they would have to sit it in the front seat or something

Why would it? The ball is on the other side of the street, the kid is running in front of the car. Unless it knows that kids play and run after balls, it shouldn't be able to know stuff like that.

 

[CaptFirePanda]

It shouldn't matter what the child is running after, sensors would alert the "driver" that something is about to intersect its path and the car would stop or otherwise avoid the child.

 

[Office_Shredder]

Why would it? The ball is on the other side of the street, the kid is running in front of the car. Unless it knows that kids play and run after balls, it shouldn't be able to know stuff like that.

http://jalopnik.com/5851324/how-googles-self+driving-car-works

It tracks its surroundings omnidirectionally. So it will be able to detect the kid running towards the street and stop to allow him to cross

 

[MarcoD]

Possible, I didn't read the article. Guess as long as it doesn't drive too fast it's okay.

 

[CaptFirePanda]

I'm sure Google is very glad to have received your blessings, Marco.

 

[AlephZero]

Considering that most of Nevada has a population density of less than 10 people per square mile (according to wikipedia), close encounters with kids playing with balls is probably not much of an issue.

In that environment, I expect a human could pass the driving test without ever meeting most hazard situations for real.

 

[MarcoD]

Like anyone cares? It's just something I've witnessed once. It ain't nice.

 

[Greg Bernhardt]

From what I've seen human drivers do, I would welcome driving along side a robot!

 

[CaptFirePanda]

Like anyone cares? It's just something I've witnessed once. It ain't nice.

Was a robot driving the car?

 

[MarcoD]

Kid getting overrun by a car since it was only watching the ball it was running after.

Oh, no, a robot wasn't driving the car. Guess there would be an argument for it since response time might be faster. Then again, the kid was darting out from between two cars anyway. No idea.

 

[Office_Shredder]

That's a situation where the robot driver could be superior to a human, because the robot driver has constant 360 degree vision and literally can't fail to notice the kid (barring sensor errors or something) whereas a human driver might not be as attentive

 

[MarcoD]

Say you're next to a baseball field for kids. A human driver could, possibly, slow down because he or she knows the situation. A baseball wouldn't track on the sensors, neither would a party moving behind other cars.

It's a trade-off. But I wouldn't let a robotic car drive very fast in suburbs.

 

[CaptFirePanda]

That's a situation where the robot driver could be superior to a human

[emphasis mine]

I most definitely agree, but mistakes will happen. Whether it's a manufacturing defect or an installation defect, there is always room for error.

What bothers me most about this is the fact that if I cut off a robot driver and it honks at me, will it be hurt when I flip it the bird in my rearview mirror?

Say you're next to a baseball field for kids. A human driver could, possibly, slow down because he or she knows the situation. A baseball wouldn't track on the sensors, neither would a party moving behind other cars.

The data would likely allow for the robot driver to "know" it was driving past a baseball field and it could slow down accordingly. It would also be limited by posted speed limits and wouldn;t have the urge to drive just a little faster than legally allowed.

 

[AlephZero]

It's a trade-off. But I wouldn't let a robotic car drive very fast in suburbs.

I expect it would be programmed to obey speed limits at all times - unlike most human drivers.

 

[SW VandeCarr]

Considering that most of Nevada has a population density of less than 10 people per square mile (according to wikipedia), close encounters with kids playing with balls is probably not much of an issue.

In that environment, I expect a human could pass the driving test without ever meeting most hazard situations for real.

That's just an average. Ever been to Las Vegas, Reno? They're cities, just like cities everywhere. The cities are where the cars and darting kids are. Most visitors to Las Vegas arrive by air. When you're in the city, you're not even aware that the city is surrounded by endless miles of desert*. It's not relevant.

* Except when the wind blows sand into the city.

 

[Office_Shredder]

What bothers me most about this is the fact that if I cut off a robot driver and it honks at me, will it be hurt when I flip it the bird in my rearview mirror?

It doesn't come standard but should be an option in most mid-line sedans by 2020

 

[mheslep]

I can't remember exactly who said this (a top guy in GM perhaps?) but I read recently that in the last decade there has been a substantial development in the technology behind self-driving cars.
...

Indeed:

DARPA Grand Challenge—a raucous race for robotic, driverless vehicles sponsored by the Pentagon, which awards a $2 million purse to the winning team. Armed with artificial intelligence, laser-guided vision, GPS navigation, and 3-D mapping systems, the contenders are some of the world's most advanced robots. Yet even their formidable technology and mechanical prowess may not be enough to overcome the grueling 130-mile course through Nevada's desert terrain. From concept to construction to the final competition, "The Great Robot Race" delivers the absorbing inside story of clever engineers and their unyielding drive to create a champion, capturing the only aerial footage that exists of the Grand Challenge.

The engineering on the winning Stanford vehicle in particular was IMO exquisite.
http://www.pbs.org/wgbh/nova/darpa/about.html

 

[ThomasT]

@ dipole,
Actually, I was just joking. Maybe I watch too much scifi.

I think it's a great idea to develop the technology, however I'm not sure why we need (or why there would ever be viable market for) robot cars and/or robot drivers. Aren't there enough people out of work already?

To comment on a few of your points:

We have machines that operate autonomously all the time ...

It's one thing to put a plane in the open sky on autopilot for a while, but auto traffic is another thing. Lots of variables.

... trains are capable of driving themselves ...

They're on a track.

... most of the technology and goods you own were probably built or assembled using robots at some stage of their manufacture...

Doing very limited, repetitive tasks.

If machines can offer a real improvement to human transportation by driving on roads autonomously, why is this a bad thing?

A big if. But assuming that acceptable robot driving technology is someday developed, who's going to be able to afford to buy it? Would it put a certain segment of the population out of work?

Is a robot really less trustworthy than some teenager who's texting/blaring music/chatting with friends/and being reckless all at the same time while driving? We allow them to drive, and in fact the number one cause of deaths among young people is car accidents.

This assumes that robot cars/drivers are going to be implemented on a massive scale wrt individual vehicles.

A simpler solution to the problem of teen driving is to raise the driving age.

 

[Gokul43201]

Would it put a certain segment of the population out of work?

If we'd bowed to the seemingly unpleasant ramifications of this question every time someone promoted a technological advancement, we'd probably be a population with a 100% employment rate, but everyone would be either hunting or gathering.

 

[ThomasT]

If we'd bowed to the seemingly unpleasant ramifications of this question every time someone promoted a technological advancement, we'd probably be a population with a 100% employment rate, but everyone would be either hunting or gathering.

You might be right about that. Anyway, I think they're a long way from putting taxi, bus, and truck drivers out of work. Just too many variables for robots to safely handle I think. But I might be wrong about that.

 

[CaptFirePanda]

We'll likely see robotic public servants first.

 

[wuliheron]

We'll likely see robotic public servants first.

The next republican candidate for president maybe?

 

[SW VandeCarr]

We'll likely see robotic public servants first.

Not if the public service unions have anything to say about it, and they do!

 

[ThomasT]

We'll likely see robotic public servants first.

Or just servants. Yeah, I can see that. But still, a niche market.

 

[Ryan_m_b]

You might be right about that. Anyway, I think they're a long way from putting taxi, bus, and truck drivers out of work. Just too many variables for robots to safely handle I think. But I might be wrong about that.

There's also the question of whether or not it would create jobs. It's not a straight jump to self driving cars but the SARTRE project requires people to drive trucks all day up and down main highways. Also there may be the possibility for this to create jobs by making commuting far easier for people e.g if a fleet of public transport self-driving cars was put into operation in an area of http://www.poverty.org.uk/75/index.shtml. This will be aided by the increased speed that self-drive cars have the potential to travel at (IMO) thanks to all the cars driving as one block (no erratic driving slowing everyone down), cross roads replaced by smooth merging (cars turn into spaces made for them) and on major roads speed limits can be increased to whatever the cars can take without the risk of accidents.

We'll likely see robotic public servants first.

I would wager that this falls under a similar thing to AI: once we can do it we'll no longer call it a robot.