Google Is Building An AI Driver, Not A Self-Driving Car


Google is one of the largest tech giants working on developing self-driving car technologies and has been doing so for more than half a decade. It's safe to assume that half a decade would have been enough for a company to develop and manufacture a non-smart vehicle for the roads, but in reality, Google's top priority lies in developing self-driving software that's smart enough to safely push autonomous vehicles on public roads. This is an enormous challenge in itself, and as Google's head of self-driving technology told the media in a recent interview, the company's endeavors in the automotive market aren't as much about building a vehicle as they are about creating a superhuman driver.

In a way, creating software and an array of sensors to allow vehicles to drive themselves in a controlled environment isn't that much of a challenge, but building a system that can function in the real world and on public roads requires a lot more work. In other words, it's not enough for the sensors to detect objects and obstacles; the system also needs to act intuitively and on a case-by-case basis in accordance with changes in its environment. Furthermore, the company's vision is to create a self-driving system that can technically be used anywhere and regardless of the car model. This is the reason why Google's head of self-driving technology, Dmitri Dolgov, claims that Google is not developing a car, but rather a driver. "We're building a driver. We've been on Prius, Lexus; we have our own prototype, and we're now working with Fiat Chryslers on a new platform". Mr. Dolgov also added that as far as the software is concerned "it's like you getting into another car. You get a rental – maybe it's a little bit bigger, and it doesn't quite handle the same way as your own car – it takes you time to get used to, but the core tasks transfer".

As for the "magic" behind Google's self-driving car system, these "superhuman drivers" are designed to rely on a series of long-range lasers and radars that allow the car to "see" to a distance of at least 100 meters around it. The system can detect and understand hand signals (a symbolic language we humans take for granted) and react accordingly, so for instance, if a self-driving vehicle encounters a bicyclist who signals using his or her hand, the self-driving car (or superhuman-driver) should technically react much like a human would. Furthermore, the system is connected to a cloud and self-driving vehicles can communicate with, and learn from each other. On the other hand, this type of system can technically increase the risk of hacking, and for this reason, Google has taken extra precautions by separating the communication interface used by self-driving cars to share information with one another from the interface required to control the car. In closing, you can take a look at the image below to get a general idea of what a Google self-driving vehicle "sees" when roaming the streets alongside human drivers.



Share this page

Copyright ©2016 Android Headlines. All Rights Reserved.

This post may contain affiliate links. See our privacy policy for more information.
Senior Staff Writer

Mihai has written for Androidheadlines since 2016 and is a Senior Writer for the site. Mihai has a background in arts and owned a couple of small businesses in the late 2000s, namely an interior design firm and a clothing manufacturing line. He dabbled with real-estate for a short while and worked as a tech news writer for several publications since 2011. He always had an appreciation for silicon-based technology and hopes it will contribute to a better humanity. Contact him at [email protected]

View Comments