Google's Self-Driving Car Had An Accident - And Google Takes Responsibility

Google's self driving Lexus RX SUV had a minor accident when it crashed into the side of a bus while pulling out onto the road on Feb 14 near Google's Mountain View headquarters in California. Google has now acknowledged that it was a fault by the car.

Google has been trialing the vehicles over the recent few months.

Google published a report admitting the car had been at fault on Feb 29. The report said that the car had detected the coming bus but assumed it would slow down and let the car go ahead of it.

The entire report discusses this one incident. This might be because Google has always made a case for its self-driving cars using the point of its lack of caused collisions.

The human driver in the Google car who is required legally to be there in case there is a malfunction did not take hold of the wheel or try to override the car's self-driving computer because he said that he believed that the bus would slow down too.

Google's self driving cars, which were revealed back in January to completed nearly 425,000 miles of testing on the roads, have been involved in 13 minor 'near-miss' accidents before. However, this is the first time that these cars were at fault.

This, according to Tech Week Europe, is proof that the cars still have quite some way to go in their testing process until their technology is perfect. 

In previous accidents the engineers riding the cars had to take control of the manual system of the car several times.

The car had been traveling at 2mph (3km/h) and the bus it collided with at 15mph (24km/h).

The driver who was in the car during the incident reported that it was possible that what confused the car was an unexpected obstacle - sandbags kept to mark a lane closure on the road.

Google, who has refined their self-driving algorithm, did assume responsibility - adding that they would meet with the California Department of Motor Vehicles (DMV) to discuss the accident and try and find out where the fault lies. 

Real Time Analytics