Waymo, autonomous vehicle technology division at Google, has admitted it stopped developing self-driving technology requiring human intervention a few years years ago.
John Krafcik, Waymo's CEO, told Reuters and other outlets that five years ago, the division was developing assisted driving technology as a quick route to market.
Like the level three autonomous driving system that will soon be available in the Audi A8 in some markets, Google's system would drive on highways, and hand responsibility back to a human on other roads or in circumstances beyond its programming.
During a tour of Waymo's facilities, Krafcik admitted: "What we found was pretty scary. It’s hard to take over because they have lost contextual awareness.”
In a filmed experiment, Google's test drivers were seen playing with their phones and applying make up at speeds up to 90km/h. One was even spotted napping at the wheel.
Not long after this test, management decided to focus solely on developing level five autonomous vehicle technology, which would require no active driver intervention.
The autonomous Chrysler Pacifica plug-in hybrid people mover Waymo is trialling as a ride-hailing service in Phoenix, Arizona, has just two driver operated buttons: one to start a ride, and another to request the vehicle pull over at the first available opportunity.