Google Tech to Help Autonomous Cars Communicate With Pedestrians

Google's newly patented technology is designed to give pedestrians a way to understand a self-driving car's "intention" at crosswalks and intersections.

Google autonomous car

Google has been granted a patent for a technology that would let an autonomous car notify pedestrians of the "intent" of the vehicle at crosswalks, stop signs and other such areas.

For example, a self-driving car equipped with the technology might use an electronic sign, lights, a speaker, or robotic hands and eyes to inform pedestrians of its intention to slow down, continue without change, or stop or yield at an intersection or traffic light.

An autonomous car has no way to communicate with pedestrians in the manner that human drivers do in certain situations, Google said in a patent published Nov. 24 and first reported on in Time. A typical driver, for instance, might be able to inform a pedestrian of his or her intention to wait for the pedestrian to cross a street by communicating using eye contact, hand gestures and other means.

Autonomous vehicles, on the other hand, have little besides the usual signaling devices such as turn signals, headlights, brake lights, high beams, horns and reverse lights to signal intention. "Autonomous vehicles lack the capability to directly communicate the vehicle's future behavior," the Google patent noted. As a result, pedestrians will likely be unsure about whether it is safe to cross a road in front of an autonomous vehicle. Merely stopping a vehicle at an intersection is unlikely to reassure pedestrians that it is safe to cross, Google said in its patent.

The patent describes various scenarios where the technology would be useful—for example, when a pedestrian might want to cross a road or has already begun doing so, or is standing at the edge of the crosswalk waiting for cues about the vehicle's intentions. Google's new technology will use data from the various sensors embedded in the vehicle to make decisions on what to communicate to pedestrians. Autonomous vehicles typically come equipped with lasers, radar, camera, sonar and other devices to gather data about the driving environment.

"Sensor data from one or more of these devices may be used to detect objects and their respective characteristics," with relation to position, shape, heading and speed, the Google patent said.

Google's autonomous car effort, launched as a moonshot project a few years ago, has been gathering momentum in recent months. The company has been testing a fleet of modified SUVs and its own, home-built collection of fully autonomous cars for several years now.

Most of the tests have been conducted in California, but recently the company began testing on public roads in Texas.

Google's autonomous fleet has already logged more than 1.7 million miles in test drives over the past six years. In that time, the company said its cars have been involved in just 11 mostly minor collisions, all when a human operator was in control of the vehicles.

Earlier this year, Google hired auto-industry veteran John Krafcik to head the autonomous car venture, which is still a part of Google's X-Labs operation.

Editor's Note: Google's autonomous fleet has already logged more than 1.7 million miles in test drives over the past six years, not 1.7 miles as originally written.

Jaikumar Vijayan

Jaikumar Vijayan

Vijayan is an award-winning independent journalist and tech content creation specialist covering data security and privacy, business intelligence, big data and data analytics.