Google's self-driving cars have proved they're capable of navigating highways under ideal circumstances—and so far there aren't any accidents to speak of. But what about in a city like New York where drivers are particularly unpredictable? In a blog post on Monday, Google said that its next phase of development is to master the chaos of city streets.
Director of Google's self-driving program, Chris Urmson, explains that the search giant has made several software upgrades in the past several months, imbuing the technology with some great new capabilities. For starters, Google said its software can now detect hundreds of distinct objects simultaneously, including pedestrians, buses, city landmarks and even a stop sign held up by a crossing guard. Urmson admits that going one mile on the highway is much different from going a mile in a city, and Google is confident its technology is capable of that.
"A self-driving vehicle can pay attention to all of these things in a way that a human physically can't—and it never gets tired or distracted," Urmson said.
Google's software updates are so advanced that what looks chaotic to the human eye, such as jaywalking pedestrians or a car suddenly pulling out of a driveway, are "fairly predictable to a computer." Each self-driving vehicle is equipped with that recognizable little bucket on the top, which contains 64 lasers that collects three-dimensional information in all directions by bouncing radar waves off objects within 500 feet.
Going further, there's also a camera that looks out toward the front of the vehicle and is capable of reading everything from traffic signals to street signs. Imagine setting a location on your GPS, and the car getting you to your destination without incident. Google shared a video that shows off what the world looks like from the software's perspective, and how it ultimately is able to drive passengers around as if a human was at the wheel.
As we've encountered thousands of different situations, we've built software models of what to expect, from the likely (a car stopping at a red light) to the unlikely (blowing through it). We still have lots of problems to solve, including teaching the car to drive more streets in Mountain View before we tackle another town, but thousands of situations on city streets that would have stumped us two years ago can now be navigated autonomously.
More than wearable computers and predictive technology in my phone, a self-driving future sounds like my kind of utopia. Down the road, Google hopes the next car you buy is equipped with the technology necessary to drive on the city or highway without human intervention. If it's in my lifetime, I will absolutely be one of the first to line up.