In the spring of 1961, President Kennedy said the United States would put a man on the moon before the decade was through. Improbably, eight years later, there they were: Neil and Buzz standing on the surface of a powdery grey rock that once seemed so far away. Right on schedule.
Present day, the giants of Silicon Valley and the big players in the auto industry are all aiming for a moonshot of their own. Only this time, it’s not quite going according to schedule.
Earlier this year, Elon Musk—the eccentric founder of Tesla Motors—told investors that his company could have an autonomous car ready in three years. Musk is vocal about his stance on driverless cars. At another conference, he went so far as to say human drivers could be banned from the roads in 20 years.
In 2012, Google co-founder Sergey Brin said his company would have self-driving cars for the public in five years. Three years in, Google pushed back the estimate, saying it would be, “closer to five years from now.” Carlos Ghosn, CEO of Nissan and Renault, also pushed back his companies’ goals (which were 2020), saying entirely autonomous cars now won’t be available until 2025.
To date, the State of California has issued self-driving car permits to eight companies at last count, including Volkswagen, Mercedes, Google, Tesla and Nissan.
One tentative step for mankind.
But, how real are “robocars” and, more importantly, what’s stalling their arrival?
Since 2009, Google’s self-driving cars have clocked 2.7 million kilometres on public roads. According to the company, its robot vehicles have successfully avoided hitting a child riding on a toy car in the street, as well as a woman in a wheelchair who chased a duck in front of one of the vehicles.
The biggest problem, so far, is other drivers. Google claims that none of the 14 crashes its machines have been involved in were the robot’s fault. (One happened when a human driver was in control, and the most recent incident involved a distracted driver who rear-ended the Google vehicle.)
Google is essentially asserting that robot cars are good; it’s the human drivers around them who aren’t.
While the thought is unnerving, many of us have already given up some control of our vehicle to robots without realizing it. Mercedes has a 3D stereo-camera-based system that automatically “sees” the road ahead and can slam on the brakes if it senses an imminent crash. Radar-based cruise control keeps you a certain distance from the car ahead by speeding up or slowing as needed, while lane departure control keeps you from veering into another’s path. These are features that are available now, off the showroom floor.
But to build a truly autonomous vehicle would require more than just sensors and radar and supercomputing power. It demands, if not artificial intelligence, a machine that can learn and adapt; something that can think compassionately and quickly, like a human.
For instance, what should a robocar do if faced with a choice between two life-threatening courses of action? (This is the Trolley Problem: a thought experiment that uses the example of a trolley hurtling towards five people who are tied to its track. The subject is standing next to a lever that can divert the trolley onto another track where it will only kill one person.) Could a robotic car recognize all mitigating factors a human would see? It’s hard to say if we’re even ready to let machines have such a great responsibility.
And then there’s the problem of legislation. Who is liable if your robocar hits a telephone pole, or worse, kills a cyclist? Is it the fault of the company that built the vehicle, the company that programmed its software, the city for having bad signage, or you, for sending the car to park itself?
The fact is that 1.2 million people are killed on the roads every year, according to the World Health Organization. Human error is a factor in 90 per cent of those incidents. The ultimate goal of robocars, Google’s Sergey Brin told the Associated Press, is to eliminate human error.
Robocars could save many of those 1.2 million lives, even if the technology is not infallible. There is the potential to reduce congestion, provide mobility to elderly and disabled people, and facilitate mass-scale car sharing.
So, when will you get that cyborg chauffer you’ve always wanted? The real answer is that nobody knows. There is no President JFK for robocars, nobody brave enough—or with the resources or authority— to set a hard deadline. Go ahead and renew your current lease, but watch this space.
It all makes landing on the moon look easy.