Self-driving cars have long been a contentious topic among drivers, especially with behemoths such as Google bringing them into the mainstream. One of the most hotly-debated topics concerning self-driving cars has been their relative safety on the road as they coexist with human drivers.
This issue received some attention last year when Google released accident data regarding their self-driving cars. In a thorough detailing of their various collision incidents, it seemed to be that, while the cars did experience a substantial amount of minor collisions with human drivers, none of them were the fault of the self-driving cars. While this data indicates that self-driving cars may perhaps follow the rules of the road better, there was nonetheless a threat of collisions with self-driving cars – and no measure of how these futuristic vehicles stacked up to human-driven cars in terms of collision rate.
Fortunately, a recent study published by the Virginia Tech Transportation Institute has attempted to fill in that blank.
Crunching the Numbers on Crash Rates
In the report, the Institute concludes that the collision rate of self-driving cars is considerably lower than the national crash rate of human-driven cars. Self-driving cars in the past year have had a crash rate of 3.2 collisions per million miles, while conventional vehicles have a rate of 4.2 crashes per million miles. While this may not seem like significant data, it’s the first time that researchers have had a significant amount of comprehensive data on conventional cars to compare with the meticulous data that is gathered electronically by self-driving cars.
Additionally, the data also reiterates the fact that all of the collisions experienced by self-driving vehicles were at the fault of the human party involved.
Drawing Conclusions From the Numbers
This may seem like a conclusive win for Google and the other self-driving vehicle manufacturers, but there may still be a few points to consider:
- While the data suggest that self-driving vehicles may experience less-severe collisions, the amount of data is not sufficient enough to make a decisive conclusion.
- The data for conventional vehicles is comprised of more than 3,300 vehicles driving more than 34 million miles, considerably more than that of self-driving cars.
- The study itself was commissioned by Google, although the Institute clarifies that the opinions and conclusions drawn are solely their own.
Ultimately, while all evidence seems to suggest that self-driving cars are indeed better drivers than humans, that may not necessarily mean they’re safer drivers. The seemingly imminent prevalence of self-driving cars may mean that we as humans be more alert and aware of the things happening on the road – promoting a potential need for more comprehensive driver safety for everyone.