Google: Self-driving cars are mastering city streets

(CNN) — Long a veteran of the highways of rural California, Google’s self-driving car is working on becoming safer in the city.

Over the past year or so, Google has been fine-tuning how the software running its fleet of automated vehicles handles the complexities of stop-and-go driving in heavily populated areas.

“A mile of city driving is much more complex than a mile of freeway driving, with hundreds of different objects moving according to different rules of the road in a small area,” Chris Urmson, the head of Google’s self-driving-car project, said Monday in a blog post.

Urmson said engineers have improved the cars’ software to recognize situations like pedestrian traffic, buses, stop signs held by crossing guards and hand signals made by cyclists.

And, he says, self-driving cars have the potential to handle all of that even better than we do.

“A self-driving vehicle can pay attention to all of these things in a way that a human physically can’t — and it never gets tired or distracted,” Urmson wrote. “As it turns out, what looks chaotic and random on a city street to the human eye is actually fairly predictable to a computer.”

Since 2011, when self-driving vehicles became street-legal in Nevada, Google has logged nearly 700,000 miles with the cars, mostly on highways. The only reported accidents have happened when one of the cars was being driven by a person, or they were the fault of another driver.

Autonomous cars are also now legal in California, Florida and Michigan, although all states still require a human driver behind the wheel.

Google has been testing the cars around its Silicon Valley headquarters in suburban Mountain View, California.

There’s more to learn before testing them in another city, Urmson wrote, “but thousands of situations on city streets that would have stumped us two years ago can now be navigated autonomously.”

The cars’ technology includes a laser radar system and a laser-based range finder that lets software create detailed 3-D maps of the surroundings.

In a YouTube video also posted Monday, one of the cars is shown recognizing and changing lanes in a construction zone, negotiating a railroad crossing and making a right turn at an intersection crowded with cars, cyclists and pedestrians.

“With every passing mile we’re growing more optimistic that we’re heading toward an achievable goal — a vehicle that operates fully without human intervention,” Urmson wrote.