According to a study published by the Virginia Tech Transportation Institute, self-driving cars on average were involved in fewer accidents compared to vehicles driven by humans.
Johnny Luu, a spokesperson for Google, said that the company commissioned Virginia Tech to carry out the study to develop a methodology that will be able to meaningfully compare the safety of self-driving cars and human-driven vehicles.
According to Luu, the study will be very helpful in the future development of self-driving cars as an apples-to-apples comparison can now be made with human-driven cars from the developed methodology.
The study analyzed more than 50 self-driving vehicles of Google, which have travelled about 1.3 million miles in total across California and Texas without being driven by humans.
Google has previously reported incidents of minor accidents that involved its self-driving vehicles that are being tested. The company's fleet has been involved in just 17 crashes over the past six years, with none of the incidents being the fault of the self-driving vehicles, according to Google.
The Alphabet unit's self-driving car report for November 2015 discussed an incident when one vehicle was pulled over by a cop for moving too slow and another case when Google's autonomous Lexus car, the Google AV, was rear-ended as it approached an intersection.
Adjustments were made on how to classify the severity of the crashes and to account for accidents that were not reported to the authorities.
Last year, the National Highway Traffic Safety Administration discovered that about 60 percent were only property damaged vehicular accidents and 24 percent of accidents that resulted in injuries were not reported to the authorities. It is a law in California, meanwhile, for all accidents that involve self-driving vehicles to be reported to the authorities.
After the adjustments were made, the Virginia Tech study estimates that human-driven vehicles find themselves in 4.2 crashes per million miles, as opposed to self-driving cars that find themselves in 3.2 crashes per million miles.
It also found that the crash rates for traditional cars, at all the indicated levels of severity, were higher compared to the crash rates of self-driving cars.
The researchers have discovered statistically significant data that self-driving vehicles were less likely to be involved in serious accidents compared to human-driven vehicles. Although, they stressed that additional studies are needed to say for sure that self-driving cars are indeed safer than their counterparts, with human drivers behind the wheel.
One of the major issues with the Google-commissioned study is the sheer number of variables to control in the examination of the safety of self-driving cars. There are some cases when it would be very hard to determine what exactly caused an accident wherein a human driver was involved, on the other hand, self-driving cars are still not widespread enough to check if the safety technologies included in these vehicles actually hold up against the myriad of real-life situations that can be experienced.
A similar study was published (PDF) by the University of Michigan Transportation Research Institute in October 2015, which compared the crash rates among the self-driving cars of Google, Delphi and Audi. The UMTRI's study found results opposite to the Virginia Tech's study – with the self-driving cars having a higher rate of crashes compared to human-driven vehicles.
However, the UMTRI study also noted the low volume of miles driven by autonomous cars, with 1.2 million miles compared to the 3 trillion miles that were travelled per year by human-driven vehicles.
With such conflicting results and no defined national safety standards for self-driving cars, further studies are certainly needed to determine just how safe autonomous vehicles are compared to traditional cars with human drivers. However, pending these additional studies, researchers such as the team from Google will continue to progress their development of self-driving car technology to build upon the massive strides that have been made over the past few years.