The internet giant says its fleet of autonomous vehicles - mainly sporting Toyota and Lexus badges - have now clocked-up thousands of kilometres while driving virtually unaided around the streets of Mountain View, California.
In a relatively short time, Google engineers have ‘taught’ the car to recognise temporary roadwork signage, unpredictable movements from cyclists, the complexities of railway crossings and more.
As seen in the video (top of page), Google’s autonomous car is now equipped with the knowhow to deal with an ‘unusual’ lane-change, caused by road works and indicated only by temporary signage.
The footage shows a cyclist playing ‘daredevil’ with the Google car, using hand signals and moving into the car’s lane to simulate a cyclist who may have failed to give way; demonstrating more of the car’s new autonomous features.
The car can also detect cyclists pedalling within cycle lanes, along with vehicles parked by the roadside while still occupying a small portion of the lane.
"A self-driving vehicle can pay attention to all of these things in a way that a human physically can't -- and it never gets tired or distracted," Google said in its official blog.
“As it turns out, what looks chaotic and random on a city street to the human eye is actually fairly predictable to a computer. As we’ve encountered thousands of different situations, we’ve built software models of what to expect, from the likely (a car stopping at a red light) to the unlikely (blowing through it).”
Google said the autonomous car is now equipped to handle situations on city streets that would have stumped the car’s software just two years ago, and that more than 1.1 million kilometres have now been logged by its autonomous cars.