One of the worst crashes yet involving a Google self-driving car happened on Sept. 23, when another vehicle ran a red light and barreled into the passenger side of one the company’s Lexus SUVs.
When the incident was first reported last month, Google emphasized that the light “was green for at least six seconds before our car entered the intersection.” But on Oct. 6 in its monthly self-driving car project report the company released a more startling detail: the crash was severe enough that its test driver “voluntarily went to a local hospital.”
Here’s what happened, according to Google’s latest report:
As the Google [autonomous vehicle] proceeded through a green light at the El Camino Real intersection, its autonomous technology detected another vehicle traveling westbound on El Camino Real approaching the intersection at 30 mph and began to apply the Google AV’s brakes in anticipation that the other vehicle would run through the red light. The Google AV test driver then disengaged the autonomous technology and took manual control of the Google AV. Immediately thereafter, the other vehicle ran through the red light and collided with the right side of the Google AV at 30 mph. At the time of collision, the Google AV was traveling at 22 mph. The Google AV sustained substantial damage to its front and rear passenger doors. The other vehicle sustained significant damage to its front end.
The driver turned out to be fine, “evaluated by medical staff and released,” according to Google’s report. Even so, the anecdote is a bit unsettling when tech companies and automakers alike are rushing to get autonomous technology ready for the road, and Uber has begun putting real passengers into its driverless vehicles in Pittsburgh.
Earlier this week, Quartz reported on two minor scrapes that Uber’s self-driving cars have already gotten into in the Steel City. Higher profile incidents involving driverless technologies this year include a self-driving Google car hitting a public bus in Silicon Valley, and a crash that killed the driver of a Tesla Model S while autopilot was enabled.
For now, tests of truly driverless cars like Google’s and Uber’s always involve an engineer or other form of safety driver behind the wheel. This can add an even greater degree of uncertainty, as human operators are inevitably forced to make split-second decisions about whether they should intervene, or trust the vehicle to negotiate a tricky spot. Google’s report does not comment on whether the test driver responded appropriately by turning off the car’s autonomy, but the phrasing of its report implies that the driver may have interfered with the car’s ability to avoid a collision (“its autonomous technology detected another vehicle … and began to apply the Google AV’s brakes in anticipation that the other vehicle would run through the red light”).
At the bottom of September’s self-driving car report, Google notes that future reports will not mention “collisions that occurred while our car was being manually driven.” The company did not elaborate on why. It will “continue to include manual collisions that occur when the test driver takes control at the last moment,” as seems to have been the case in the Sept. 23 accident.
Google’s self-driving car project began in 2009 and on Oct. 5 the company announced that its vehicles had driven more than 2 million miles on public roads. Google, now a division of Alphabet, said it had reached the 1 million mile threshold just 16 months earlier. Its self-driving fleet in the US consists of 24 Lexus SUVs and 34 prototype vehicles, operating on the roadways of four different states.