San Francisco saw its first traffic accident involving an autonomous car earlier this month.
It occurred on the afternoon of Jan. 8 when a Nissan Leaf autonomous vehicle operated by Cruise Automation hit a parked Toyota Prius.
The car was in manual control mode at the time, but an account of the incident says it started when the automation computer steered the car incorrectly.
"The vehicle began moving in its lane to the left, then began correcting to the right at which point the drive decided the take over manual control," the company said in a report to the California Department of Motor Vehicles.
"After taking over manual control, the operator did not change the path of the vehicle and it collided with an unoccupied Toyota Prius."
The report doesn't say how much time the driver had to correct the car's course before the collision but Kyle Vogt, CEO of Cruise Automation, told IDG News Service the driver was alerted several seconds before the crash.
"There was enough time," he said. "He took over manual control but unfortunately made a mistake."
There were no injuries, but both cars suffered minor damage in the accident, which happened on 7th Street just ahead of the intersection with Bryant Street in San Francisco's South of Market district.
Cruise Automation is developing an after-market highway autopilot system called the RP-1. The $10,000 device works with 2012 and newer Audi A4 and S4 models. The company says it has already presold its initial production run of 250 of the units.
Units were supposed to be delivered in late 2015, but it's unclear if this happened. The company's website still refers to them as yet to be delivered and Vogt declined to comment on whether they had been shipped.
The California DMV requires reports on any accidents on public roads that involve cars in tests of autonomous driving. A handful have been reported since October 2014, the majority involving Google in Mountain View. That's not a surprise, as Google has the largest autonomous fleet and the most aggressive testing schedule.