With multiple road trials now under way it seems inevitable that driverless cars will soon become common on our highways. Soon we will be sharing the road with vehicles that are controlled by unseen computer programming, GPS and elaborate algorithms. These vehicles will be autonomous and left to make decisions on their own. But how safe is this technology? The trails that are now under way across the country hope to prove how safe this really is and work out any kinks that may be found along the way.
The manufacturers of driverless vehicles assure us that they will have almost superhuman abilities recognizing the world around them. They are loaded with sensors and gather tons of data about their environment. All this in the hopes that they can replicate the way human drivers handle road conditions under all circumstances.
Cameras, radars, lasers and ultrasonic sensors along with GPS play the parts of the eyes ears and directional sense of a person driving a car. All this data must then be fed into a brain that will be making the important decisions.
For most of the new vehicles the brain consists of an integrated computer system about the size of a license plate. This computer brain uses machine learning to teach itself about the driving environment and to recognize the ever-changing conditions that exist there. This machine learning is what allows driverless cars to amass a great deal of knowledge about all the possible threats that can arise and what to do about them.
But how safe does this really make the driverless car? According to the manufacturers it’s a matter of how many hours and miles of real driving under real world conditions the cars are exposed to that makes them ever safer.
Well there’s only one way to know for sure and that’s by trail and error. Let’s hope that driverless cars can learn to be safe quickly.