There will be over 10 million car accidents by the end of this year, and most of them will be due to human error. Of those 10 million car accidents, over 30,000 will be fatal.
Annual US traffic fatalities per billion vehicle miles traveled (red) and miles traveled (blue) from 1922 to 2012.
Safety technology in cars has come a long way, but the graph above suggests safety benefits are beginning to plateau. Simply put, advancements in car safety are not as impactful as they once were.
So how do we continue to make big improvements in car safety? Improve the driver, not the car.
And by “improve the driver” I mean remove human operators and replace them with near-flawless computer systems that operate the car. I'm talking about making self-driving cars.
94% of all car accidents in the United States result from user error. This is a horrific number which could be reduced to less than 10% in the near future.
After six years and almost 1 million miles, Google's self-driving cars have been in 11 minor accidents. In each of those 11 minor accidents, however, it was not the self-driving car's fault. It was the human-operated car that caused the accident.
Google's self-driving cars are primitive, yet they prove that these innovative vehicles have the potential to drastically improve road safety. A self-driving car senses everything around it. And by “everything around it” I literally mean its suit of sensors detects activity with 360° visibility. And by “suit of sensors” I'm not talking about the shit in your mother's Roomba. The sensors on self-driving cars will exceed the capabilities of our human eyes and brain.
Self-driving cars will collect more information about the surrounding area than you. Self-driving cars will process information faster than you. Self-driving cars will pay more attention to the road than you. Self-driving cars will get lost less often than you. Self-driving cars will make better decisions than you.
Self-driving cars will drive better than you, and they will reduce car accident fatalities.