Google self-driving cars near miss and survive 13 serious accidents in two months
Google itself reveals the figures that its self-driving cars have 13 near miss incidents in two months. Latest figures from the Google about the self-driving car test reveals that there were 272 failures recorded during 424,331 miles covered in September and November last year.
There were around 13 failures recorded as very serious out of those 272 in total. Those 13 incidents were very close miss accidents and survived because drivers took control of the vehicle but after a short time, the controls were taken back by the vehicle itself.
What kind of failures they were?
All the 272 issues recorded as technical failures and recorded as technical issues but in all 272 issues, drivers handed back control to the vehicle.
The majority of issues recorded as hardware or software failures.
What has to be done to make it a safer option?
Before the Google self-driving car become reality and part of the real-time traffic, it has left hundreds of questions unanswered about the safety and reliability of the vehicle or its passengers. It also has given many clues about what has to be done to make it safer on the road, because, if the software fails while driving even for a fraction of time, it can lead to some horrible incidents.
It can risk the lives of people on the road. It is simply like if driver dies while driving, what would happen to the vehicle? No one knows but a serious accident is sure.
What is the frequency of software disengagement?
Of the recorded issues, Google reported that 69 times drivers faced a software disengagement and it is a big concern for safety. Out of those 69 issues recorded, 56 issues were account for real life conditions.
Google has simulator to test the software, it drives more than three million virtual miles every day to detect the life on the road, reaction of the cyclists and pedestrians to determine if the driver intervention is necessary?