MOUNTAIN VIEW, California — Google's fleet of self-driving vehicles, which include special Lexus RX 450h SUVs and pod-like prototypes, are learning to respond "smoothly to the unexpected," said Google in its June update.
"We use the same basic principles to guide us in all situations," the report said. "If we're not sure what something is or how it will behave, we'll slow down to give ourselves time to gather more information about the situation.
"If it's a moving object, we'll categorize it as a vehicle, cyclist or pedestrian so we can make reasonable predictions about its behavior — and be prepared for odd behavior. When in doubt, our cars will stop. Over time, we're getting better at modeling and responding smoothly to the unexpected, from cyclists riding the wrong way in a bike lane, to the new fad of lane-splitting electric skateboards."
Google said it reported two minor accidents involving its self-driving vehicles to the California Department of Motor Vehicles in June. Both occurred in Mountain View and involved Google Lexus autonomous vehicles.
The trend points to driver error and inattention on the part of occupants of the non self-driving vehicles, Google notes.
In both instances, the Google vehicles were hit from behind by another vehicle in low-speed crashes. No injuries were reported.
Google said its prototype vehicles were deployed on public roads in Mountain View last month and it is continuing to test its Lexus self-driving vehicles.
"We'll gradually introduce more prototypes to the streets over the coming months," it said.
The Internet giant is currently averaging 10,000 autonomous miles per week on public streets.
The Google fleet includes 25 prototypes, two of which are self-driving on public streets, and 23 Lexus RX 450h SUVs, also driving on public streets. Google said the tests are "mainly" in Mountain View.
The fleet has logged more than 1 million autonomous-mode miles, which means the software is driving the vehicle and safety drivers onboard are not touching the manual controls.
Edmunds says: Google's self-driving cars are learning to handle the expected and the unusual, as autonomous cars creep closer to reality.