Uber has suspended tests of its autonomous cars on public roads in Arizona in the wake of a deadly crash involving one of the vehicles in Tempe, but it’s not yet clear if other automakers will follow suit or be forced to do so by the state government.
The incident unfolded late Sunday when 49-year-old Elaine Herzberg was struck by a 2017 Volvo XC90 SUV equipped with Uber’s autonomous technology as she was walking her bicycle across a four-lane street midblock, outside of any crosswalk.
Police say the vehicle was traveling approximately 40 mph in autonomous mode when it hit Herzberg, and video from in the car showed no signs that it was slowing down, although the investigation into the matter is preliminary. Herzberg later was pronounced dead at a hospital.
An Uber employee, Raphael Vasquez, was behind the wheel observing the vehicle’s operation at the time of the incident, but police could not yet say if he took any action to avoid Herzberg. The car is equipped with a suite of sensors that work in conjunction with highly accurate maps to navigate streets and attempt to identify and avoid obstacles.
Arizona regulations on autonomous car testing that were put in place by executive order earlier this month require that the vehicles be monitored by a licensed driver who is either in the car, or doing so remotely. The vehicles can operate with an empty driver’s seat, as was recently depicted in a video released by Uber competitor Waymo. General Motors and Intel both have significant testing operations in the state as well.
A spokesman said the governor’s office was in communication with law enforcement, but did not say if it would call for any new or temporary restrictions on testing, and the Arizona DOT has neither put out a press release nor responded to emailed questions from Fox News on the topic.
Intel and General Motors, two of the other companies testing autonomous cars in the state, have not made any public announcements and would not comment on their immediate plans to Fox News, while Waymo has not yet responded to a similar request.