Google’s competitor in the race to map the world in preparation for self-driving cars is already HERE. That was not a typo. HERE is the maps division of Nokia, the company best known as a maker of cellular handsets (although it unceremoniously hung up on its phone business just over a year ago). Google has focused its engineers’ tremendous technical acumen on mapping every detail of 3200 kilometers (2000 miles) of roads surrounding its Mountain View, Calif., headquarters. But HERE has been taking a different route, so to speak. Over the past year or so, Nokia unit has used its fleet of robocars to map 2 million kilometers (1.2 million miles) in 30 countries on every continent but Antarctica.
In an bid to publicize its ongoing mapping plan for self-driving cars, two of the people leading its self-driving car effort gave reporters at Wired a first-hand look at HERE’s mission-control-like headquarters, a ride in one of the autonomous vehicles in its fleet, and even dropped by Wired’s offices to provide further details. Lucky them. Here’s what we can all take away from what the HERE engineers revealed.
To build its digital maps, HERE starts with satellite and aerial imagery and incorporates so-called “probe data” collected by trucking companies and other fleet operators whose vehicles contain GPS devices. This rich data source—which comprises about 100 billion data points per month—allows engineers at HERE’s Berkeley, Calif., control center to populate the maps upon which its fleet relies with up-to-the-second traffic information. But the probe data is still not as important to the operation as the details provided by HERE’s own fleet—nearly 200 cars equipped with GPS, cameras, and lidar.
Its fleet uses a lidar system that “spins around, shooting out 32 laser beams and analyzing the light that bounces back,” John Ristevski, HERE’s head of reality capture, told Wired. Talk about detailed? Ristevski told Wired that:
[The lidar] collects 700,000 points per second... An inertial sensor tracks the pitch, roll, and yaw of the car so that the lidar data can be corrected for the position of the car and used to create a 3-D model of the roads it has traveled. The lidar instrument’s range tops out about 10-15 stories above the street. At street level, its resolution is just a few centimeters.
Another issue for autonomous vehicles that HERE is addressing is how to, in near real time, deliver updates about things such as accidents and the congestion that occurs when hundreds of cars spill out of a stadium parking lot at the end of a sporting event. Just as important is relaying information about, say, a pothole or a downed tree that would allow a self-driving car to take precautions in a way that wouldn’t scare the daylights out of the “driver.” Peter Skillman, Nokia HERE’s lead designer, told Wired that,
“Sensors on future autonomous cars could feed information over cellular data networks to HERE’s map in the cloud, but…it could take several seconds for a car in San Francisco to beam its data to a data center in, say, North Carolina, and get a response. Getting response times down to tens of milliseconds—fast enough for a car to switch lanes to avoid some debris in the road spotted by another car ahead of it—will require applications that live inside the LTE networks and can be accessed locally.”
That, says Skillman, will be necessary to get humans to trust that a car suddenly swerving out of its lane hasn’t suddenly shut down or gone rogue. Well, that and an onboard digital map that gives them a preview of what the car is about to do and a sense of why the robocar makes certain choices.
Willie Jones is an associate editor at IEEE Spectrum. In addition to editing and planning daily coverage, he manages several of Spectrum's newsletters and contributes regularly to the monthly Big Picture section that appears in the print edition.