What's Putting the Brakes on Google's Self-Driving Car

A Google's street view car is seen on a busy street in Jakarta Photograph by Mast Irham/EPA via Corbis

Google is weighing building its own line of self-driving cars independent of the automakers, according to new report by Amir Efrati on JessicaLessin.com. Efrati doesn’t name his sources, but he’s a veteran Google reporter formerly of the Wall Street Journal so I have little reason to doubt them. But it does raise an interesting question: Can a tech company—even one with the resources and innovation drive of Google—build an automobile from scratch?

First, the details of the report: Efrati’s sources said Google is making no headway with the entrenched automakers over partnerships to build self-driving vehicles. So it’s opted to go around them, talking to auto-components designers Continental and Magna International about having them build cars to Google’s design. (German paper Frankfurter Allgemeine also reported Continental has struck a deal with both Google and IBM.)

Efrati’s report added that Google might use these cars as part of a “robo-taxi” service that prowls cities picking up passengers on demand.

Google is no stranger to tackling big hardware projects. It became an ISP when it launched Google Fiber. It makes its own tablets and smartphones, it’s been designing its own data centers for year, and it’s even attempting to create the world’s first borderless network flying in the stratospheric winds with Project Loon.

But make a car? That requires considerable commitment of finances and resources, not to mention the need to recruit a different type of engineer—mechanical rather than software. Silicon Valley has created car companies before, but it’s not as if Elon Musk started Tesla Motors as a side business of PayPal. He helped found a whole new company devoted to making an electric vehicle, received investment from the automakers, and recruited some of the best and brightest of the auto industry to help run it.

While Google is probably one of the few companies in the world with the resources and wherewithal to start an auto-manufacturing operation in-house, you’ve got to wonder if all the trouble is worth it. When Google made its Nexus line of phones, it was mainly trying to create concept phones that demonstrated Android’s full potential. When it got serious about making phones it went out and bought a phone maker: Motorola. I seriously doubt Google is going to go out and buy a struggling automaker.

The Nexus, however, might be a good model if Google is sincere about proving its autonomous car concept. Google could be aiming to create the Nexus equivalent of a vehicle: The car would be built by someone else, while the software and technology that made it autonomous, intelligent, and connected would all be Google.

Still, it does seem that the auto industry and Google are heading toward the same goal even if they’re getting there in entirely different ways. While Google appears more focused on building a sensor-driven vehicle, automakers like Ford, Honda, and BMW have been trying to combine sensor technologies with networking technologies that allow cars to communicate their positions and intent with one another.

Those automakers could be moving far too slowly—focusing on “assisted driving” rather than taking the driver out of the equation—for Google’s tastes. So maybe Google believes a “Nexus car,” one demonstrating the full capability of its driverless technology, is the kick in the pants the auto industry needs.

Also from GigaOM:

What the Next Business Model for EVs Might Look Like (subscription required)

A Decade Later, Tesla Now Officially a Threat to the Auto Industry

On Second Thought, Ballmer’s Exit Doesn’t Look All That Smooth

Hey! New Microsoft CEO, Whoever You Are, Take Care of Azure

Five Things You Should Know About Google’s Cloud Platform

Before it's here, it's on the Bloomberg Terminal.