Why Google Wanted Their Self Driving Car to Not Look Like a Car


Google’s self-driving car prototype doesn’t look like a car. It very deliberately looks unlike a car.

Early development of Google’s driverless vehicle technology was done by retrofitting normal cars from established brands. Google’s in-house prototype— the first “Google Car”—doesn’t make even the slightest effort to fit in. Though the form factor is similar to city-friendly smartcars, even among them Google’s cartoonish creation would stand out, partly due to the sensor array it has on top and partly because, head-on, it looks like its smiling.

Clearly this is not an accident. It is not as if Google’s designers were trying to make a car that looks like other cars and did a poor job. It’s also not that self-driving technology dictated this particular appearance: that development continues to be done on normal-looking vehicles demonstrates this.

So Google is consciously making the statement that this thing of theirs is not like a car. It isn’t a continuous evolution from cars that came before it, like electric and hybrid vehicles that are visually indistinguishable from all-gasoline cars.

Indeed, this is meant to be a fully autonomous car, the kind that doesn’t even need a steering wheel. That’s because during their testing, Google discovered that, in an emergency, relying on humans to take over didn’t work. After awhile, humans get lazy in these cars, making them ill-prepared to take over.

Borrowing from NASA’s design approach, Google’s new car copes with such eventualities, naturally, with more computers.

“It doesn’t have a fallback to human—it has redundant systems,” Nathaniel Fairfield, a technical lead on the project, said recently. “It has two steering motors, and we have various ways we can bring it to a stop.”

A weird design is probably the smarter move for Google. There wasn’t much hope of the self-driving car debuting as something that looks just like any other car. Does this matter? I think it is a strong indication of how transformative Google thinks its driverless vehicles are. It signals a willingness to treat this project as something that will break conventions as a matter of course.

Obviously, having robots take the wheel is a break with convention. Rather than minimizing that break, Google has made a deliberate choice to embrace it, hoping we’ll bask in the weirdness of their future vision.

Dan Luxemburg, CTO at Bandwagon