After decades of movies where self-driving cars seemed like a fantasy, the last few years of development have shown that they could become a more realistic option. Now, we have started to look at autonomous vehicles as a way to increase safety on the roads.
While many accidents are the result of driver error, there may still be engineering and computer programming hurdles that could prevent self-driving cars from reaching the pinnacle of driving safety.
These are some of the latest realizations that have come about with research and development focused on creating a safe, self-driving vehicle.
People have multiple priorities
The human brain is capable of processing various high-priority processes at the same time and can alternate between what is most important as circumstances change. As you drive to work, you may be thinking about what route is fastest, how bad the traffic is, and what you will do when you get to work. Any of those thoughts could take an immediate backseat to another driver swerving through traffic.
Fortunately, autonomous vehicles do not get distracted by to-do lists or songs on the radio. The dilemma, however, is creating a program that changes priorities as seamlessly as a human brain when circumstances change.
Experts say, in order to create a car that drives more safely than human drivers, the programming involved would always have to “prioritize safety over speed and convenience.”
Prediction is still a problem
The difference between human drivers and autonomous vehicles is that humans can anticipate actions that machines are still unable to predict. Not long ago, there was an accident with an Uber test vehicle where the vehicle could not foresee that Elaine Herzberg would cross in front of the car.
While humans make many mistakes when driving, they tend to be able to predict what other humans will do, such as a child suddenly running out into the street after a ball. For now, at least, humans remain the safer drivers.