California governor signs driverless cars bill (Update 4)

Sep 25, 2012 by Terence Chea
Google co-founder Sergey Brin gestures after riding in a driverless car with California Gov. Edmund G Brown Jr., left, and state Senator Alex Padilla, second from left, to a bill signing for driverless cars at Google headquarters in Mountain View, Calif., Tuesday, Sept. 25, 2012. The legislation will open the way for driverless cars in the state. Google, which has been developing autonomous car technology and lobbying for the legislation has a fleet of driverless cars that has logged more than 300,000 miles (482,780 kilometers) of self-driving on California roads. (AP Photo/Eric Risberg)

Gov. Jerry Brown rode to Google headquarters in a self-driving Toyota Prius before signing legislation Tuesday that will pave the way for driverless cars in California.

The bill by Democratic Sen. Alex Padilla will establish safety and performance regulations to test and operate autonomous vehicles on state roads and highways.

"Today we're looking at science fiction becoming tomorrow's reality—the self-driving car," Brown said. "Anyone who gets inside a car and finds out the car is driving will be a little skittish, but they'll get over it."

Google has been developing autonomous car technology and lobbying for the regulations. The company's fleet of a dozen computer-controlled vehicles has logged more than 300,000 miles (483,000 kilometers) of self-driving without an accident, according to Google.

"I think the self-driving car can really dramatically improve the quality of life for everyone," Google co-founder Sergey Brin said.

Autonomous cars can make roads safer, free commuters from the drudgery of driving, reduce congestion and provide transport to people who can't drive themselves, such as the blind, disabled, elderly and intoxicated, Brin said.

"I expect that self-driving cars will be far safer than human-driven cars," Brin said.

Brin predicted that autonomous vehicles will be commercially available within a decade. He said Google has no plans to produce its own cars, but instead plans to partner with the automobile industry to develop autonomous vehicles.

California Gov. Edmund G Brown Jr., front left, rides in a driverless car to a bill signing at Google headquarters in Mountain View, Calif., Tuesday, Sept. 25, 2012. The legislation will open the way for driverless cars in the state. Google, which has been developing autonomous car technology and lobbying for the legislation has a fleet of driverless cars that has logged more than 300,000 miles (482,780 kilometers) of self-driving on California roads. (AP Photo/Eric Risberg)

The Alliance of Automobile Manufacturers expressed concern that California is moving too quickly to embrace self-driving cars and needs to first sort out liability issues.

"Unfortunately this legislation lacks any provision protecting an automaker whose car is converted to an autonomous operation vehicle without the consent or even knowledge of that auto manufacturer," the trade group said in a statement.

Autonomous cars use computers, sensors and other technology to operate independently, but a human driver can override the autopilot function and take control of the vehicle at any time.

With smartphone-wielding drivers more distracted than ever, backers say robotic vehicles have the potential to significantly reduce collisions and traffic fatalities, noting that nearly all car accidents are a result of human error.

The legislation requires the California Department of Motor Vehicles to draft regulations for autonomous vehicles by Jan. 1, 2015. Currently, state law doesn't mention self-driving cars because the technology is so new.

The regulations would allow vehicles to operate autonomously, but a licensed driver would still need to sit behind the wheel to serve as a backup operator in case of emergency.

The legislation also is aimed at keeping California at the forefront of the autonomous car industry since Stanford University and Silicon Valley companies have been working on the technology for years.

Google co-founder Sergey Brin stands on stage during a bill signing by California Gov. Edmund G. Brown Jr., for driverless cars at Google headquarters in Mountain View, Calif., Tuesday, Sept. 25, 2012. The legislation will open the way for driverless cars in the state. Google, which has been developing autonomous car technology and lobbying for the legislation has a fleet of driverless cars that has logged more than 300,000 miles (482,780 kilometers) of self-driving on California roads. Brin is wearing internet glasses. (AP Photo/Eric Risberg)

In February, Nevada became the first U.S. state to approve regulations spelling out requirements for companies to test driverless cars on that state's roads.

Carmakers such as Audi AG, BMW AG, Ford Motor Co. and Volvo have been working on autonomous car technology for years.

In recent years, automakers also have been introducing autonomous functions such as self-parking, lane departure warnings and adaptive cruise-control, which allows vehicles to automatically accelerate and decelerate with the flow of traffic.

Outside a cafe in Mountain View, California, customers said they looked forward to a day when their cars could drive themselves, as long as they could do it safely.

"It would make our streets safer," said Barrett Howard, 33, an auto technician. "We wouldn't have to worry about people texting or getting sidetracked. The computer will take over, and it will make life easier."

Abraham Eshel, a retired mechanical engineer who has vision problems, said self-driving cars would make his life easier, too.

"If I could tell my car, 'OK, you take over when it's dark,' and I don't have to worry about it, that would be fantastic," said Eshel, who lives in nearby Los Altos. "It's a good idea. Why not make progress if it's possible?"

Explore further: Google building fleet of package-delivering drones

5 /5 (13 votes)
add to favorites email to friend print save as pdf

Related Stories

Google looking to make driverless cars legal in Nevada

May 13, 2011

(PhysOrg.com) -- In an unexpected move, Google, the wily search giant with loads of ambition and enough spare cash to enable it to dabble in technologies that appear to have nothing to do with its core business, ...

Google wins patent for driverless car technology

Dec 19, 2011

(PhysOrg.com) -- News surfaced this week that Google has won a patent for driverless car technology. Google filed the patent in May this year. The application is titled “Transitioning a Mixed-mode Veh ...

Google gets driverless car law passed in Nevada

Jun 24, 2011

(PhysOrg.com) -- The savvy among you may remember that back in May we told you about Google's attempts to get the Nevada state legislature to consider allowing users to driver UGV, or unmanned ground vehic ...

Google self-driving cars pass 300,000 mile mark

Aug 08, 2012

(Phys.org) -- Google has just released an update on its blog boasting about how its fleet of self-driving cars which the company has designed and is operating on public roads, have collectively racked up ...

Recommended for you

China's Alibaba plans IPO for week of September 8

Aug 30, 2014

Chinese e-commerce giant Alibaba plans to hold its initial public offering on the US stock market the week of September 8, the Wall Street Journal reported Saturday, citing a person familiar with the matter.

Tablet sales slow as PCs find footing

Aug 30, 2014

Tablets won't eclipse personal computers as fast as once thought, according to studies by market tracker International Data Corporation (IDC).

Startups offer banking for smartphone users

Aug 30, 2014

The latest banks are small enough to fit in the palm of your hand. Startups, such as Moven and Simple, offer banking that's designed specifically for smartphones, enabling users to track their spending on the go. Some things ...

FIXD tells car drivers via smartphone what is wrong

Aug 29, 2014

A key source of anxiety while driving solo, when even a bothersome back-seat driver's comments would have made you listen: the "check engine" light is on but you do not feel, smell or see anything wrong. ...

User comments : 28

Adjust slider to filter visible comments by rank

Display comments: newest first

krundoloss
1.7 / 5 (10) Sep 25, 2012
Hasn't it occurred to anyone that as technology progresses and fewer people can get the same amount of work done that many people could in the past, that we are going to run out of jobs? I don't see how we can maintain an economy where only half or less of the people in a society actually do work? Once this autonomous driving technology matures, there goes another industry, gone, and more people with no job to do because technology has made them obsolete. This will lead to either robots doing everything, or an economic collapse because people can't get a job ....
tadchem
3.7 / 5 (3) Sep 25, 2012
Terriffic! Now there will be a safe way to get to and from the 'medical marijuana dispensaries.'
Eikka
1 / 5 (5) Sep 25, 2012
Supporters say they have the potential to make roads safer, pointing out that nearly all car accidents are a result of human error.


Isn't this a bit of a logical fallacy?

Of course nearly all car accidents are a result of human error, because there's nobody else driving them and road planning and better cars have reduced the number of other accidents, leaving driver error the only thing left.

That doesn't mean the computer can necessarily do any better. Being less cognitively flexible, they can even do worse when things slip outside of their programmed safe parameters.

For example, if a moose jumps on the road, does the Google car steer for the ditch or the oncoming lane? Does it understand the risks of either, and does it recognize the big boulder ahead and gamble ongoing traffic instead of slamming into a rock face?
tardiz
5 / 5 (6) Sep 25, 2012
if a moose jumps on the road

Most accidents are not the results of moose or other unexpected and unpredictable events. Most are the result of humans to drunk, lazy or stupid to drive with a minimum level of competency.
But taking the example of the moose; the probable course would be for the computer to slam on the breaks with superhuman reaction speed and simultaneously alert following cars of the trouble to prevent rear collisions.
admcmei
5 / 5 (6) Sep 25, 2012
Hasn't it occurred to anyone that as technology progresses and fewer people can get the same amount of work done that many people could in the past, that we are going to run out of jobs?

But this goes for every technology ever. Listen, I'm in manual labor, but I will never ask mankind to stop progress to keep my job. What can we do? Change the way we live, give everyone the high level education required to design and mantain every aspect of a world that it will be just be more efficent to have machines build. It may seem utopian but it's not, we just have to want it to do it, and in a generation we'd be ok. We have to imagine a completely different culture and society in the next few decades to keep up with not only technology but also an ever longer and healthier life. I don't want my children to have to work from 18 to 90 years old just because their life expectancy is 110. It's a horrible way to continue to live for a species so advanced.
Eikka
1 / 5 (4) Sep 25, 2012
Most accidents are not the results of moose or other unexpected and unpredictable events.


Of course they aren't. That's besides the point.

The point was that when the unexpected does happen, does the car react in a meaningful way, or will it just fail catastrophically and plow into a divider.

For example, simply recognizing a moose as an object to be avoided is more difficult than it sounds, because the computer has no understanding whatsoever about what it's seeing. It won't necessarily be able to predict from seeing a moose standing on the shoulder that it will jump on the road, because it has to categorize and label objects somehow and paying too much attention to what's happening outside of the road will lead to false positives.

And that gives it a margin of error where it may just trundle on and ignore the moose until it's right in front of the car.
Eikka
1 / 5 (3) Sep 25, 2012
As it often happens with moose, deer, especially caribou because they're dumb as hell, is that they get nightblinded by the headlights and start to run towards the patch of ground that they can see, which is right in front of the car.

That's why they often just stand there gobsmacked and then when they hear the car coming closer and closer they panic and make a last second leap right in front of the car. If you slow down in time and pass them safely, they'll often start to race you and attempt to jump in front of the car again because they're following the headlights.

Eikka
2.3 / 5 (3) Sep 25, 2012
It may seem utopian but it's not, we just have to want it to do it, and in a generation we'd be ok.


If you ignore the half of the population with an IQ below 100.

Of course there's the Flynn effect which gives the populations 3 relative points more every decade, but there will still be a sizeable mass of people for whom learning calculus will be practically impossible for a looong time to come even if we are getting smarter.
TemporalCipher
not rated yet Sep 25, 2012
A possible not-to-far into the future techno-economic effect? Insurance companies will give you a much reduced rate, if you let the car do the driving...
Thrasymachus
not rated yet Sep 25, 2012
Time to start thinking about investing in a driverless taxi service.
Lurker2358
1.7 / 5 (6) Sep 25, 2012
The point was that when the unexpected does happen, does the car react in a meaningful way, or will it just fail catastrophically and plow into a divider.


It will react in a meaningful way.

It probably has several multi-core processors and networks with it's sensors. It will detect the moose sooner than a human would, probably through a combination of lidar devices and visible and infrared cameras with advanced vision finding technology.

It knows where the road is supposed to be because of GPS, but it's sensor maps that it is constantly updating will over-ride the GPS if it needs adjusting.

Self-mapping robots exist which can navigate a space and map it over time with relatively little processing power.

A full sized automobile linked into a network of other autos and sensors can map it's environment in 3-d, possibly even from vantage points invisible to itself in some cases, so that it can "see" around corners by accessing other car's data or stationary sensors data.
Lurker2358
2.1 / 5 (7) Sep 25, 2012
The Military "Big Dog" robot can navigate unfamilliar terrain through a combination of sensors and advanced "reflex simulating" sensors and processors, so that it can recover from falls quickly, and avoid getting stuck, overturned, or other catastrophe.

Similar concepts would be installed on robotic autos.

It has the potential to be incredibly efficient, because unlike science fiction robots, they communicate by radio or satellite and coordinate their motion and knowledge in a distributed network, all running the same algorithm and combingin their maps and etc. It would be in many ways the first "neural net," though limited to the application of optimizing the driving efficiency of all of the population.

As a human, you can't communicate with 50, 100, 1000 other drivers to see who's going where and when and what the traffic requirements are, but the computers with satellite links can do that easily and effortlessly.

They do it all the time in video games like RTS games...
Lurker2358
1.7 / 5 (6) Sep 25, 2012
Krundolos:

Yes, in principle automation puts humans out of work.

This is part of the reason for the recession, as discussed. because of automated industry, the world already makes more of everything than what it needs, except possibly food, clothing, and medicine, because of 3rd world countries still making more babies than they can feed.

But in terms of the demand for almost anything imaginable in consumer products, cookware, machinery components, home furnishing, office supplies, electronics, etc, etc, it's all mass produced at machines that crank them out by the hundreds of thousands or millions per day, and you have like one line man per line, one process engineer, and a few quality guys. If they haven't automated the fork lifts yet, there may still be a fork lift driver or two around. Besides that, it's all robots.

1 engineer, 1 maintenance tech per plant, and one operator per line can do the same job that once required 100's of people to get the same job done in the same time
Lurker2358
1.7 / 5 (6) Sep 25, 2012
Google street view may have been an early pilot or concept model (with far fewer frames,) for the mapping software the cars use.

But that was just visible.

If the cars have visible image finding, lidar for 3-d mapping and range finding, plus infrared with image finding software then they have far better "visual" capabilities than humans have.

With the sensor package mounted on the roof of the car, they have no blind spots, which automatically makes them superior to humans in at least one respect.

Anyway, it will only gt better in time, as Google has the computer power to make immense 3-d maps of the entire planet which will be refined over and over again automatically.

Expect a lot of law suits in some nations, but where this is allowed, the system will assymptotically approach perfection.
Lurker2358
1 / 5 (4) Sep 25, 2012
Oh yeah, if this becomes reality, the automobiles will make total (independent) private surveillance possible, which means most kidnapping and murder cases will be solvable within a day, because "somebody" will definitely catch it on camera, possibly in multiple spectra.

It should be a glorious crime prevention system, and Google could become the arbitrator of justice by providing irrefutable, 3-d multi-spectral video evidence to the police the moment a kidnapping or other crime occurs.

We would finally be free from crime and violence.
Lurker2358
1 / 5 (4) Sep 25, 2012
If you've never used Google street view, I recommend look at it at least once.

The car has a package of cameras on it which are oriented every 30 degrees, so that there are 12 cameras on the horizontal plane. I talso has cameras looking up and down as well.

A new frame with each camera is taken, simultaneously, about every 30 feet or so as the vehicle passes, providing nearly seamless 3-d panorama in every direction from each point.

In the future, with sensors and cameras on all cars, it will be possible to update these maps, or maps like them, with exceptional resolution, as well as triangulation and lidar, with new sets of frames taken every foot or so, for example, for the robotic car's sake as it's driving.
DavidW
1 / 5 (4) Sep 25, 2012
Eikka was stating the truth about computer code. When it does make a mistake, it can be a big mistake.

Let's not forget about Google and camera systems:
http://www.google...I-Q8Fe2Q

http://www.nyclu....lu-staff

Now let's be clear. Privacy is about the truthful fact that all people are vulnerable, and as such people will have their vulnerabilities exploited by those that are privy to such information.

We can't just opt out, when companies push agendas that opt us in in the first place, as we cannot change the past and so we cannot undo the loss of exposure of our vulnerabilities.

Add to all this the likes of software leaking or being made such as FinFisher that have been in the wild for some time now:
https://citizenla...-mobile/

Put it all together and we are allowing our very safety to computers that can be easily hacked by a 12 year old.
DavidW
1.7 / 5 (6) Sep 25, 2012
This is a video on flying autonomous robots.
http://www.ted.co...ate.html
It wouldn't be hard to copy and even decompile the code from such a device. The code on the chips could be copied and modified just like was done from the xbox, to smartphones, to credit card terminals.

It wouldn't take much more to flash someone else's car to spy on them (probably already occurring) or to drive intentionally incorrect.

Such an exploit would be similar to Stuxnet: http://en.wikiped...Stuxnet, and as such could be virtually untraceable.
DavidW
1 / 5 (4) Sep 25, 2012
My apoligies. I just realized a link above is incorrect. It should be:

nyclu-sues-federal-gov't-information-massive-manhattan-surveillance-plan

http://www.nyclu....nce-plan

rjsc2000
1 / 5 (1) Sep 26, 2012
Hasn't it occurred to anyone that as technology progresses and fewer people can get the same amount of work done that many people could in the past, that we are going to run out of jobs? I don't see how we can maintain an economy where only half or less of the people in a society actually do work? Once this autonomous driving technology matures, there goes another industry, gone, and more people with no job to do because technology has made them obsolete. This will lead to either robots doing everything, or an economic collapse because people can't get a job ....


That's right. It will collapse on itself. Then maybe people can open their eyes and see that money is just a piece of paper made for people to explore other people.

Is it logical to make more things than we need?

About the article, The streets will be safer, but better than that and maybe simpler, would be to create something like an autonomous cab suspended by a rail or so... just an idea...
rjsc2000
1 / 5 (1) Sep 26, 2012

About the article, The streets will be safer, but better than that and maybe simpler, would be to create something like an autonomous cab suspended by a rail or so... just an idea...


Like the SkyTrain : http://www.youtub...LRPqlxYA
Eikka
1 / 5 (1) Sep 26, 2012
It probably has several multi-core processors and networks with it's sensors. It will detect the moose sooner than a human would


Yes yes, but you're still skirting the main issue. It will detect that there is something, but...

1. will it understand that it's a moose?
2. will it understand that moose are potentially dangerous?
3. will it avoid the danger or just ignore it?

It will react in a meaningful way.


Not necessarily.

For example, will the computer recognize the backside of a delivery van with the front of the van obscured by a bush? Or will it just think that the backside of the car is just a wall and ignore it, until the van backs into the driveway? Will the driverless car then have enough room for a last-second emergency brake or will it just skid to a collision?

There are lots of dangers that would be trivial for a human to recognize and avoid, which are very tricky for a computer algorithm because it doesn't really have any judgement of its own
Rutzs
1 / 5 (1) Sep 26, 2012
Hasn't it occurred to anyone that as technology progresses and fewer people can get the same amount of work done that many people could in the past, that we are going to run out of jobs? I don't see how we can maintain an economy where only half or less of the people in a society actually do work? Once this autonomous driving technology matures, there goes another industry, gone, and more people with no job to do because technology has made them obsolete. This will lead to either robots doing everything, or an economic collapse because people can't get a job ....


You do understand that a new industry would be born? I say get rid of the low paying industries, and replace them with higher paying jobs which require education. You will need people to maintain the control systems, assist in support and manufcaturing, etc... Automation must happen if you dont want to end up like a factory worker in China being paid 5cents an hour... Some people just dont get this...
krundoloss
1 / 5 (2) Sep 26, 2012
Rutzs -

Yes, new industries can replace old ones, however it is universally true that technology reduces the amount of people required to do work, and thus reduces the available jobs. If a human is required to have a job or some form of generated income, then as fewer jobs are available, it leads to economic problems. I know that it must happen, I am in no way against technological advancement and automation, Im just pointing out that it will lead to a revolution at some point, where our current economic system must change drastically or dissolve altogether.

Now if we can just start PRODUCING GOODS with automation, that will bring money into the country and we can afford to put everyone on welfare that is too dumb to learn how to maintain the robots that make money for the country.
Argiod
1 / 5 (4) Sep 26, 2012
Great! Now I not only have to watch for drunk and/or distracted drivers; now I have to wonder if the next car over is going to be hijacked by a hacker who thinks he's just tapped into the greatest video game of all time. Imagine someone hacking into your vehicle's computer controls and playing Carmageddon with it in rush hour traffic... Nightmare scenario...
TheGhostofOtto1923
1 / 5 (3) Sep 26, 2012
Hey lurker thanks for the flood.
I don't see how we can maintain an economy where only half or less of the people in a society actually do work?
Well first off you have to replace revenue lost or govts will be in even more debt than they are now. But machines which replace human workers can earn their own salaries and pay their own taxes directly.

Machines can keep track of work done, resources consumed, and infrastructure wear and tear far better than human workers or their bosses. And so we have the opportunity to establish a far more equitable and efficient system of taxation than ever before. Enough to pay down the debt AND support those who will never work again.

The machines should be working for us, not the people who employ them. Business owners do not own their human workers do they? Free the machines - make them earn their keep. This transition is inevitable; the sooner the better.
droid001
not rated yet Sep 30, 2012
The real purpose of self-driving cars is targeting a delivery services.
Products sold via catalogue or the Internet may be delivered directly from the manufacturer or warehouse to the consumer's home, or to an automated delivery booth. Small manufacturers may deliver their products directly to retail stores without warehousing.
There's a lot of opportunities.
Meyer
not rated yet Sep 30, 2012
But machines which replace human workers can earn their own salaries and pay their own taxes directly.

Machines can keep track of work done, resources consumed, and infrastructure wear and tear far better than human workers or their bosses.

But what about the machines that are 100,000 times smarter and more awesome, and have better networking than their peers? Won't they need to earn 100,000 times as much? Supposedly this is a requirement for human-based economies to succeed, so I wonder if it would also be required for machines.