Such is the marvel of human reflexes. Yet cheeky Google wants to supplant said reflexes with transistors and LIDAR and microprocessors, to replace human drivers with motorist-o-matics. This could be the first wayward step towards the domination of artificial intelligence and the apocalypse, the end of humankind as we know it! Then again, it could be a step towards a greener, safer future.
The Internet of Automobiles
Google was once content to be everyone’s neighbourhood know-it-all, but in 2010 it announced its intent to build a marketable self-driving vehicle. Two years later, millions watched online as Steve Mahan drove a modified Toyota Prius – or rather, a Prius drove Mahan – about town for some errands.
Fast forward five years and 1.8 million miles. It is summer 2015, and Google has a driverless car built from the ground up. No more kitted Lexus SUVs or fleets of Toyota Prii or gaggles of golf carts. Instead, Google has an adorable two-seat autobot with a top speed of 25 mph and no steering wheel, pedals or manual controls whatsoever. Welcome to the technocracy.
Google’s entourage of self-driving cars suffered a total of 12 accidents over their first five years of life. In one accident, a human driver rear-ended the autonomous car. So did another. And another. Two other drivers broadsided the hapless prototypes. “Not once was the self-driving car the cause of the accident,” insisted program director Chris Urmson. With apologies to Mark Twain, the statistics don’t lie: The Google car drives much better than you.
The secret to Google’s success is LIDAR, a laser-guided mapping system somewhat similar to echolocation. The Velodyne 64-beam laser atop the car records 1.3 million readings every second. Stereo cameras in front create overlapping fields of vision to measure object size and distance. Four radar systems measure the speed of other cars. A dedicated camera views traffic lights. A GPS anticipates the road ahead. That flood of information, a whopping one gigabyte per second, cascades into the computer which creates a three-dimensional map of the car and its environment.
Clean, Connected Cars
Each Google car remembers information learned by the others. The cars share a communal brain, and thus, a Google driverless vehicle pulls over for an ambulance, waits for jaywalkers, and ignores raised middle fingers. As of now, the system cannot function properly in heavy rain, accumulated snow, or on roads without clear markings.
Yet none of these technological obstacles stop Google and its aficionados from dreaming. Soon, cars will park themselves. People will summon autonomous taxi cabs via text message. Lines of cars will platoon, each matching the acceleration of the other, saving precious time, gasoline, and road space. Cars will “see” one another using vehicle-to-vehicle Wi-Fi. Drunk driving will (hopefully) become a historical footnote.
Google is not the only company with skin in the game. Audi, Mercedes, and even Apple have engineered or expressed interest in self-driving cars. Many hope the cars will save lives. Others hope they will save the earth or at least improve upon the green technologies put in place with other hybrid and electric vehicles.
Every day, the United States consumes 375 million gallons of gasoline, according to Duke Energy. Such avarice begets sulfur dioxides, PM 2.5 pollution, carbon monoxide, and rusting mountains of metal waste destined for landfills. But a team of transportation scientists at the Organisation for Economic Cooperation and Development reimagined the streets of Lisbon, Portugal filled with autonomous “taxibots.” Nine out of 10 cars would disappear, they predicted. Oil, and America’s dependence on it, might disappear as well.
Or the cars will gather in gangs, learn to hack the World Wide Web, and take over the earth.