No matter how far autonomous vehicles are driven, they can't be driven far enough to demonstrate their safety. Scientists have discovered that autonomous vehicles would have to be driven hundreds of millions of miles in order to create enough data to clearly demonstrate their safety.

Motor vehicle accidents are the leading cause of premature death in the United States. In fact, they're responsible for more than $80 billion annually in medical care and lost productivity due to injuries. Autonomous vehicles also hold enormous potential for managing this crisis. This is largely why scientists are so interested in creating these vehicles.

With that said, autonomous vehicles may not eliminate all crashes. In addition, the safety of human drivers is critical when it comes to comparing the safety of all autonomous vehicles.

While the number of crashes and fatalities from human drivers is high, the rate of these occurrences is low when comparing them to the number of miles that people drive. As an example, Americans drive nearly 3 trillion miles every year. In 2013, though, there were 2.3 million injuries reported. That's a rate of 77 injuries per 100 million miles driven. It also corresponds with a rate of about one fatality per every 100 million miles driven. This means that autonomous driving has to log a lot more miles to compare to regular driving.

"Our results show that developers of this technology and third-party testers cannot drive their way to safety," said Nidhi Kalra, co-author of this latest study and senior scientist at RAND. "It's going to be nearly impossible for autonomous vehicles to log enough test-driving miles on the road to statistically demonstrate their safety, when compared to the rate at which injuries and fatalities occur in human-controlled cars and trucks."

The findings show that it may be difficult to show the safety of these vehicles. The most autonomous miles that any developer has logged are about 1.3 million, and that took several years. Needless to say, this pales in comparison to the miles driven on a regular basis. While it's important data, it shows that developers will have to do a lot more driving when it comes to proving safety and reliability.

The findings are published online on the RAND website.