The staff here at Automotive.com loves cars. We’d have to, seeing as how we cover the autosphere twenty-four seven. But we have moments, usually stuck in traffic, when even we wish someone else was in the driver’s seat. Unfortunately, such public transportation is limited. Thus we were more than a little curious to the recent coverage on how Google’s so-called “self-driving” vehicles actually worked.
The secret, reveals the IEEE Spectrum, is in the details. More specifically, the road and terrain detailed gathered by the Google system. According to Google engineer Chris Urmson, the self-driving Toyota Prii create extremely detailed three-dimensional maps of their environment—roads, terrain, buildings, signs, etc.—using a roof-mounted laser. The robotic Prius then compares its maps to those stored in its database and notes the similarities and differences to allow it to drive safely and within the law. Google engineers will sometimes first drive unknown routes to gather details which they then download into the autonomous hybrid.
Other sensors, like cameras, radar, GPS, etc., help the self-driving car “know” where it is at all times and the location of other objects like vehicles and even people.
Stanford University professor Sebastian Thrun, who currently guides the self-driving cars, sees the technology one day reducing traffic congestion, improving fuel economy, and reducing vehicle accidents.
Automotive.com’s take: Would you allow a car to drive you to your destination? Personally, we’re looking forward in “riding” along in one of these self-driving Prius. As always, let us know in the comments below.
Source: IEEE Spectrum