As autonomous technology becomes more commonplace, driverless cars will soon become a reality in our everyday lives, with completely autonomous cars and taxis expected to roll out next year. But how safe are driverless cars?
First, let’s look at what a driverless car is and what that means. The term ‘driverless’ or ‘autonomous’ in this context refers to a vehicle that can guide itself without human interaction. The autonomous car is operated by a machine that has sensory data inputted into it through cameras, sensors, and in some cases, radar. This data creates a 360-degree picture, allowing the car a kind of awareness of its surroundings: other cars, buildings, and pedestrians. This will enable the car to understand where to drive in relation to what’s around it.
Due to the superior processing power of a machine over a human, driverless cars should in theory mean a decrease in the 1.3 million road fatalities caused by dangerous driving each year. An additional 20-50 million more people are injured or disabled in car accidents every year, so the case for driverless cars is a compelling one in the face of such tragic, and avoidable, injury. There’s much to be said for a car operated by something that will never be drunk, tired, angry, or distracted. As 94% of car crashes can be blamed on human error, the evidence is hard to ignore.
The downside, of course, is that driverless cars are machines, and all machines are designed, built, and instructed to follow strict rules. Deviation from those rules could be deemed a bug, malfunction, or design flaw, and deviation from the rules of the road are often what cause accidents in the first place. But what happens when a driverless car follows those rules, perfectly, and there’s still a crash? Who is at fault?
The answer is unclear, with nearly all autonomous cars operating today requiring a human presence behind the wheel to take control at any moment. In theory, the person present should be alert, ready to switch from passenger to driver at a moment’s notice. There are arguments that the car’s safety features and advanced tracking systems should be able to detect potentials threats and correct accordingly. Others point out that the driver, whether active or not, should always be held responsible as the person operating the car. However, when cars become fully autonomous, we won't actually be driving them at all.
Once we transcend the need for a human driver and can rely solely on cars to operate themselves, the answer becomes at once more clear and more complicated. With the human element removed, the strength of the car’s code, technology, and construction will be the difference between viable modes of autonomous transport, or dangerous fads that will be swiftly taken off the roads.
The key is extensive testing, smarter engineering, and checks at every level to ensure nothing but the best. In the UK, the Gateway (Greenwich Automated Transport Environment) project has just finished a pioneering experiment to bring autonomous cars to the streets of London. Gateway involved a fleet of seven electrically powered self-driving cars which rode around Greenwich.
Working in autonomous engineering is an exciting challenge, one that is expected to generate billions of revenue and change the world we live in.