Having Google around is one of the things that makes working in technology fun. They always seem to be coming up with some new, slightly off-the-wall idea for new applications of technology. Now, many of these ideas don’t amount to much (or at least they haven’t, so far). but they are usually interesting, and a refreshing change from the routine crop of “innovative”. me-too products.
Recently, Google took the wraps off of another of its research projects: the development of an automated system to drive a car without human intervention. To that end, they have brought on board some key technical people who have been involved in the DARPA Grand Challenge, a contest sponsored by the Defense Research Projects Agency to promote the development of driverless vehicles.
To develop this technology, we gathered some of the very best engineers from the DARPA Challenges, a series of autonomous vehicle races organized by the U.S. Government. Chris Urmson was the technical team leader of the CMU team that won the 2007 Urban Challenge. Mike Montemerlo was the software lead for the Stanford team that won the 2005 Grand Challenge. Also on the team is Anthony Levandowski, who built the world’s first autonomous motorcycle that participated in a DARPA Grand Challenge, and who also built a modified Prius that delivered pizza without a person inside.
Google’s system uses video cameras, radar, and laser rangefinders to detect surrounding traffic, and also taps into Google’s enormous map data base, which now includes data on speed limits on specific roads.
Google has also done some significant testing of their system on ordinary California roads, as described in an article in the New York Times.
With someone behind the wheel to take control if something goes awry and a technician in the passenger seat to monitor the navigation system, seven test cars have driven 1,000 miles without human intervention and more than 140,000 miles with only occasional human control.
The tests are conducted using the system to drive a specially-equipped car, with a trained driver behind the wheel who can take over if something goes wrong. They have successfully navigated some reasonably challenging routes, including Lombard Street in San Francisco, and California Highway 1, the Pacific Coast Highway. So far, the results have been encouraging.
The idea of automatic control of a vehicle is, of course, hardly new. Aircraft have had automatic pilots for years, as well as mostly automated instrument landing systems. There are, of course, numerous obstacles — cultural, political, and legal — to be overcome before self-driving cars are common, even after the technical problems have been solved. But the idea is intriguing: an automated driver doesn’t sightsee, get sleepy or tired, or drive while intoxicated or fiddling with a cell phone; it can certainly have faster reaction times than a human driver. There is the potential to make travel by automobile safer, which is not a trivial benefit, given that, just in the US, more than 37,000 people are killed in car accidents each year.
This is still a project in the early research stages, and it is too early to say what Google might ultimately do with it. In an article in the Washington Post. Rob Enderle of the Enderle Group says that the project is an example of Google’s lack of concentration:
“The word ‘focus’ is a word Google has never learned,” Enderle said, pointing to projects involving electricity distribution, vehicle design and artificial intelligence. He said cars that can drive themselves would allow commuters more time to surf the web, something Google would encourage.
Actually, after saying that Google lacks focus, Mr. Enderle apparently proceeded to specify how the project fits Google’s focus. As I’ve said here many times, Google is in the business of selling Internet advertising; its strategy has been to pursue projects that will promote use of the Internet, because more use means more people looking at the ads Google sells, and thus more revenue for the company.
Personally, I’m hoping this project will be a big success. Considering the way many people I encounter every day drive, it is hard to believe that an automated system could possibly be any worse.