2016 was as big a year for autonomous cars as it was for false promises. The revolution, as it was claimed, is imminent. Tesla Motors co-founder, Elon Musk, wants to build trains that travel at 760 miles an hour between California and Europe. The tech tycoon said that his “fully autonomous” car would be ready to launch by 2018. It’s all created an enormous amount of hype with media painting the picture of our future: a world where we sit in pods and read or listen to music as we’re driven to work, instructing our cars to park.
Yet even the most optimistic examination of the present state of autonomous car technology suggests that a takeover of the transport system’s status quo is about as likely as the long-awaited arrival of the jet pack as seen in comic books from the 1960s. We examine some of the hindrances to the adoption of autonomous cars and see if it really is all that it’s hyped up to be.
Autonomous cars don’t drink and drive, fall asleep at the wheel, text, talk on the phone, or put makeup on while driving. With all their artificially-intelligent bells and whistles, they’re able to navigate roads without any of these human failings that can result in accidents.
But there’s something that autonomous cars don’t deal with very well – the unexpected. The human brain is still light years ahead of any computer when it comes to making decisions in the face of sudden, unforeseen events on the road.
Computer algorithms can ensure that self-driving cars obey the rules of the road – allowing them to turn, stop and slow down at appropriate times. But this technology can’t control the driving behavior of others on the road. Autonomous cars will have to deal with drivers who speed, overtake on dangerous stretches of road and drive down one-way streets.
One solution is to equip the cars with transponders. In a similar way to how airplanes avoid each other in the air, this artificial intelligence will allow cars to “talk” to one another. Current speed, position, and direction can all be sent back and forth in a process known as vehicle-to-vehicle communication (V2V). The problem with V2V is that it’ll only be effective when there are large numbers of V2V-enabled cars on the road. And it’ll be some time before that happens.
If snow, fog, rain, and any kind of inclement weather make driving difficult for humans, how much more difficult will these conditions be for a computer? Autonomous cars stay in their lanes using cameras and sensors which track lines on the road and pavement. But they’ll struggle to do this if, for example, the road is covered in snow.
Autonomous car manufacturers have listed weather as one of the major causes of system failure, after which drivers have to assume control. But they’re confident that this technology can be improved and that this challenge can be overcome. Mercedes-Benz, for example, already offers a car with 23 different sensors that detect anything from guardrails to oncoming traffic to help keep the vehicle within its lanes.
Google’s bubble-shaped autonomous cars rely heavily on very heavily detailed maps – far more detailed than those in the conventional Google Maps. These communicate the location of intersections, traffic signs, signals, and other obstacles. The cars use these maps along with data from their sensors to navigate.
But very few roads have been mapped to this degree of detail and even if they were, infrastructure is constantly changing. The construction of new roads and highways leads to road closures and detours; an intersection that was a four-way stop may have traffic lights installed or become a roundabout. To try to resolve this, many companies are focusing on enhancing their mapping software but it’s constant game of catch up.
Consider this scenario: In the midst of fast-moving traffic, a ball bounces across the road, pursued by two small children. The autonomous car’s only options are to hit the children or veer off the road, potentially striking another obstacle and causing harm to the car’s occupants. What is the best decision to make? Should a computer be given the authority to prioritize pedestrians or passengers?
These are the sorts of questions that continue to hang over the autonomous vehicle industry. And engineers are constantly trying to find solutions to these challenges. In the event of an imminent accident, a human driver will make a series of split second decisions. But in a car controlled by artificial intelligence, it’s a predetermined choice controlled by a programmer. This is one of the greatest challenges to autonomous vehicle manufacturers and, for now, there’s no concrete solution on the cards.
Other issues, such as cost and user adoption, will be decided in the marketplace. There’s no doubt that autonomous driving technology will enable tremendous business and service-level innovation, but the question over when the technology will be good enough still remains. As with every major technological transition, there’ll be winners and there’ll be losers. Every decision that manufacturers make now will influence whether it is them pioneering the world’s first feasible autonomous vehicle, or their competitors. Read more on the future of this rollercoaster industry in our article, which highlights some of the changes that autonomous cars will bring to our everyday lives.