America loves driving. It’s a fact. On average, if you gathered ten Americans together, eight would own a car.
We use our cars as much as we possibly can. Eighty-six percent of us drive to work every day, rather than take public transport, or burn calories on foot. If it’s convenience versus mass transit or a light sweat, we choose the air-conditioned, music-filled, wonderfully private, four-wheeled metal box every time.
That said, driving is increasingly becoming untenable, unlivable — unbearable, even. If you live in one of the country’s ten most traffic-dense cities — and, unsurprisingly, here in D.C. we do — you’ll spend on average two days stuck in traffic over the course of a year. Two days. Think of the Netflix binges you could have crammed into those wasted hours.
If most of our driving is routine, to work, to the grocery store, to the gym when it’s cold and you just can’t be bothered, wouldn’t it be easier if the car could drive for us? Wouldn’t autonomy go a long way to making even the most tedious of commutes a little more bearable? That’s a question being posed by almost every major automaker.
Indeed, in just five years, General Motors, Mercedes-Benz, Nissan, Renault, Audi, BMW, Tesla and Google all expect to offer automobiles with some degree of autonomy, whether maintaining a lane on the highway or winding through a city. Like it or not, the future is autonomous, and the ramifications for our roads will be huge. PricewaterhouseCoopers believes traffic accidents will be significantly reduced, by a factor of 10. Morgan Stanley is throwing some big figures around for autonomous cars, forecasting annual savings of over one trillion dollars due to lower fuel consumption, reduced repair costs and a boost in productivity from drivers free to work in their cars on the commute to and from the office.
The idea that a car can adequately navigate the urban jungle, or a bustling highway, or throw itself round country bends, may seem like the stuff of science fiction, but it’s surprisingly closer to reality than you might think. Astro Teller, the head of Google’s semi-secret research facility, Google X, spoke this week at SXSW, offering insights into the search giant’s self-driving car program, revealing that the company’s employees have been beta testing autonomous cars on their commutes. They retain full control if needed, but essentially, Google was bringing its employees to work with no input necessary once the car was on the freeway.
However, as it transpires, humans aren’t exactly the best monitors of an autonomous car. If a system requires that the driver remains aware, it’s ultimately doomed, according to Teller, calling us fleshy meatbags a poor companion for a computer. “The assumption that humans could be a reliable back-up for the system was a total fallacy!” His proof? Google asked that employees testing the cars remain vigilant, monitoring the driving and checking for bugs in software. That didn’t happen. Once the system was activated, drivers would switch off. Teller didn’t go into great detail, only saying that “people do really stupid stuff when they’re driving.”
The solution? Remove the need for the driver completely. Most of Google’s test vehicles are built from standard road cars, such as Toyota’s Prius, but the company has its own homegrown autonomous car — one that lacks any driver controls at all. Instead, Google gives total control to the car’s computer brain. “Drivers” get in, input their destination, and sit back while the car whizzes them to their destination and parks itself. As a concept, it certainly has its appeal.
And, as it transpires, Google isn’t alone in condemning humans as an inefficient method of car control. Tesla co-founder and CEO Elon Musk this week sat down with NVIDIA’s CEO Jen-Hsun Huang to discuss his vision of self-driving cars. His answer? “You can’t have a person driving a two-ton death machine.”
Ouch. Right in our humanity. Musk, whose company makes the all-electric Model S, talked with Huang at length about autonomy. Musk imagined a world where humans actually driving cars would be outlawed as we were simply too dangerous when behind the wheel. For Huang, future vehicles would look dramatically different — if no one ever crashed, safety features such as airbags and heavy chassis would be rendered obsolete, leading to a rethinking of basic car design. Still, it’s not a future we need to worry about any time soon. Even if autonomy were refined to perfection tomorrow, Musk opined that it would take two decades before every vehicle on the road became autonomous, due to how many cars are owned, not just in America but around the world. Still, when it happens, Musk is confident about Tesla’s place in an autonomous world. “We know what to do, and we’ll be there in a few years.”
Even now, autonomous capabilities have come a long way in the past few years. Cars now do a lot of thinking for themselves. Cruise control systems can maintain a set speed, a specific distance, keep you from straying across lanes and even brake when necessary. Parking sensors are now combined with surround cameras to let cars park themselves, in both bays and parallel parking — some will even drive themselves back out of the space. Cars can detect potential crashes and either brake before the driver does to prevent it occurring, or activate security features — such as rolling up windows, tightening seatbelts, and moving pedals out of the way — to prepare for the inevitable crash.
Right now, in the right model, drivers can maintain a bare minimum of control, as their car keeps a steady pace and interior temperature, activates the wipers and switches the lights on and off as necessary. Is autonomy such a drastic next step? Perhaps, if all control is removed. I know that I, for one, would hate to lose the ability to throw my car round a few corners every now and then. That said, if I can check email rather than crawl through traffic, I say bring on the autonomous revolution.