I’d rather take my chances in a computer-driven car in California any day!
Google’s self-driving cars don’t have a flawless safety record, but it’s clearer than ever that careless human drivers remain the greater threat. A commercial van running a red light struck one of Google’s autonomous Lexus SUVs as it crossed a Mountain View intersection, seriously damaging the side of the vehicle. Neither the Google observers nor the van driver were hurt, thankfully. However, this is likely the worst incident to date for Google’s driverless cars — previous crashes have typically involved humans rear-ending the robotic cars at low speed.
Google will likely shed full detail on the collision in its self-driving car report due at the end of the month. It makes no bones about what happened in a statement, though (see below). The Lexus didn’t enter the intersection until “at least” 6 seconds after its traffic light turned green. There was no question that the van driver was at fault, in other words. Google also stressed that red light violations are the “leading cause” of car crashes in US cities, and that 94 percent of those are due to human mistakes.
Although the outcome could have been much worse, the crash underscores a key problem with moving to autonomous cars: that piloted and robotic vehicles will likely have to share the roads for a long while. Google, Uberand others can design driverless systems that follow the law to a tee and adapt swiftly to unexpected road hazards, but it might be near-impossible to protect against human drivers who throw caution to the wind. Crashes like these likely won’t disappear unless self-driving tech becomes the rule.