By John Herrman
Posting in Design
It turns out the automated car of the future has a wild side.
Late last year, Google stunned the world with an announcement that it was working on a self-driving car. What was most shocking wasn't that the company was working on such a system, but that it had been testing the system with a fleet of prototypes, logging 140,000 on public roads.
Granted, these cars were manned. But Google maintains that its software was generally able to handle the routes it was presented with without much human intervention:
Our automated cars, manned by trained operators, just drove from our Mountain View campus to our Santa Monica office and on to Hollywood Boulevard. They’ve driven down Lombard Street, crossed the Golden Gate bridge, navigated the Pacific Coast Highway, and even made it all the way around Lake Tahoe.
Except for a few tame demo videos, Google hasn't really shown us much of what these cars are capable of on the road. Now, courtesy of Danny Sullivan at SearchEngineLand, we can see the system flexing its muscles, from both outside and inside a demo car. I've posted the inside view below; the second video is available at Danny's site.
This particular course was designed for aggressive driving, and Google's engineers instructed the car to drive accordingly. This may not seem like the best way to assuage peoples' immediate and visceral worries about automated cars; predictably, the video has already spawned countless jokes in comment sections around the internet. (Example: "Today I learned how Google maps directions want you to drive to get places on time," wrote one user on Reddit.)
Counterintuitively, this kind of demo should actually be quite heartening. Why? Because it provides a more accurate representation of potential uses for Google's technology than previous videos of a calmly piloted Google-powered cars. Consider this: It's hard to imagine a near-future scenario in which cars are allow to truly drive themselves; issues of safety and liability abound, and even constructing a practical--not to mention legal--framework for active driving assistance will not be a fast or easy process. The application of such technology in less conspicuous role is much easier to imagine. A role like, say, accident prevention.
Various computer-aided accident prevention features have been included in luxury cars for years now. Google's software includes sophisticated methods for avoiding collisions, many of which could conceivably be utilized in cars that are primarily human-controlled. Aggressive driving as seen in the above video could be employed in a defensive capacity, with reaction speeds and decision-making clarity that a human operator might lack. This could serve as the first step in the increasingly inevitable march forward for automated cars.
Another possible path? Google's software as an incremental outgrowth of the cruise control systems built into nearly every car on the road today. This is the vision that Google seems to have (rightly!) promoted and cultivated so far, and will probably serve them better in their long-term goal of actually helping bring a fully automated car to market.
Still, it's nice to see that the self-driving car of the future has some driving chops.
Mar 4, 2011
@12 the car doesn't use the GPS to drive silly, it only uses the GPS to plot the course and determine distance bewteen turns, the GPS sat cannot tell the car where people are or where other cars are, duh! that sucker is covered with cameras! They have a system in place (Korea I think) that has cameras on the rear bumper that prevent the car from backing up if another vehicle or person is coming towards you when you try to back up. It literally puts the vehicle in park until the sensors detect nothing is coming towards you.
I think there are plenty of practical applications for this. How about long-haul trucking, say across the desert or to Alaska? It wouldn't have to be busy city streets, but the long distance driving for shipping or even buses could really use something like this. And of course there's military application? I'm guessing this will take off in other countries before the U.S.-- like places where mass transit rules and this kind of system could actually improve safety quite easily.
This technology will eventually be a boon for seniors and other disabled persons. My mother has had to give up her driver's license and her car because of age and eyesight problems. Her life has been greatly constricted as a result. Any senior will tell you giving up their license is one of their most feared events as they age. If, say, 25 years from now this technology is mature, it will provide myself and others a much better quality of life in late retirement.
... which located me Tel Aviv, Israel in Kansas (http://s204809728.onlinehome.us/SouthSeasCave/wp- content/uploads/2011/03/Google_Tel_Aviv.jpg), it explains the wild side. :-)
What happen if the satellite is out of wack... I can see there are going to be collisions. Unless the system realize the sat is out of commission and should be switch over to manual.
I am surprised at the negative comments made to this post. You do not have to be a tech geek to be impressed with the engineering capabilities that were demonstrated. Anyone who is unable to see the value in the technology and get excited by the accomplishments must be direct descendants of the people who told Columbus the world was flat.
perhaps traffic jams will decrease with robotic controls as the robots will not care about all of the distractions humans do.
Accudet rates for these cars? If there are no accidents, how does that compare to average human drivers. Can these cars handle rush hour in Los Angeles? How about grid lock in NYC? snow in Colorado? Rain in Florida? If those things are done well, the robot car could be ready now. Too bad it takes the Courts around 20 years to catch up with technology. I know people who I would trust less than a robot on the road right now.
Have you seen how people drive on a Saturday night? I will agree that the technology that has been demonstrated by Google and DARPA's Grand Challenge has a ways to go, but in ten years, I would be much happier driving with 10,000 autonomous cars on the road than 10,000 drunk and / or fatigued drivers on a Saturday night. Therein lies the greatest positive this technology will bring. Most people will want to drive their cars, but I think a lot of people who are in a situation where they are "kinda" impaired, will opt to let the robot drive them home rather than risk hurting themselves an others. A lot of accidents are caused by people who are a "little" buzzed or a little tired, but not so bad off that they think it would be worth it to sleep it off or get a taxi. Also, how many fender benders and squished pedestrians will be prevented by people letting the car drive while they put on makeup, shave, text, etc.? I'll admit, as a control systems guy, I'm biased towards computer control, but I don't think the systems are as far from being ready for prime time as many seem to think. We're already starting to see some "accident avoidance" systems (auto-breaking, lane deviation warning) in cars. More sophisticated systems are on the way. Once those are considered "normal" and have a proven track record, it will be a lot easier for people to accept full autonomy.
Combined with GPS, I can see how it can be used for the boring highway driving in the near term. This is not much different than auto-pilot in airplane. You still need a human driver in the most complex driving conditions but it will do a better job than human in less complex conditions because it does not get fatigue or distracted. It would be extremely valuable it can handle slippery, icy, and foggy conditions, where most errors in human judgment occur.
The technology is being developed, that's why the driver has his hands close to the wheel. We can't say that it will not become more autonomous over time. The prototype is not supposed nor meant to be set loose all by itself by the millions right now, obviously.
Administrator is correct, the video show a car in a controlled environment. How about putting 10,000 of those things in a city on a Saturday night?
140K miles on public roads is impressive. I just wonder how the operator handled situations where everything looks fine on radar but he or she could see a potential problem.
I remember when cruise controls first became available, and some idiot set the cruise control on his RV, got up and went to fix himself a sandwitch, and was surprised when he rolled over in the ditch.
The problem for Google's fantasies is that money-to-inventions conversion rate is very low, even lower if a single general (or even worse, irrelevant) business company wishes to produce a product that requires a whole new class of radical inventions, alone. They might exploit a few patents, trademarks and that will be all. The above videos doesn't prove anything. This is a closed, controlled environment and still, the driver feels insecure enough to hold his palm 2 inches from the wheel during the whole ride. Real conditions can get 100 times harder than this -you always have to take the worst case into account. Not to mention, a bad or imperfect "accident prevention" system might actually cause an accident under some circumstances -except from accidents caused by too relaxed drivers...