Short in the Saddle

Driverless Cars | Are The 'Cars of the Future' Just Too Safe?Ready or not, the age of driverless cars is dawning on America’s roads. The impending wave of self-driving vehicles is forcing states across the country to take another look at regulations that for more than a century have assumed that someone with arms and legs will be sitting behind the wheel.

And which states are leading the way in aligning state laws to permit a rapid influx of self-driving cars onto their highways? Hint: the answer may surprise you.

We know what you’re thinking: the leader of the driverless car transformation has got to be California, the state that invented the Freeway (and, an hour later, the traffic jam). After all, while the sun still rises in the East, the future—for better or for worse—always seems to arrive first on the Left Coast, right?

Think again. It turns out that a thick tangle of regulations, some of which are still in the proposal stage, actually is slowing down the introduction of driverless cars in the Golden State. Meanwhile, a state with wide-open spaces—and a libertarian mindset that seems to enforce speed limits on a voluntary basis—is giving a big howdy to the unfolding driverless revolution.

That’s right, we’re talking about Texas.

Google currently is conducting an ambitious test program of self-driving cars in Austin, TX, using four of its bubble-shaped driverless cars and four retrofitted Lexus models. The program began in August with the blessing of Austin Mayor Steve Adler who, according to an AP report, used talking points written by a Google lobbyist to hail the experiment.

State transportation officials in Texas reportedly aren’t sure driverless cars—which also don’t have steering wheels or pedals—are a good idea. But the Lone Star State may have to pass some new laws if it wants to stop them: Google maintains that current Texas law does not prohibit autonomous vehicles, and many legal experts agree.

When the AP asked the Texas Department of Motor Vehicles to comment on the legality of the Austin test program, the DMV referred the question to the state Department of Public Safety, which then punted it over to the Texas Department of Transportation. A TX DOT spokeswoman told AP the agency is waiting on guidance from the Texas Legislature—which doesn’t meet again until 2017. So it looks like Google’s driverless cars have a green light in TX.

Not so in California. AP reports that California’s Department of Motor Vehicles is nearly a year late in writing rules for public access to driverless cars, frustrating Google’s efforts to expand its original test program, which began in Silicon Valley in May. The CA-based tech giant responded by reaching out to Austin, which gave the program a Texas-sized welcome.

[Editor’s Note: several observers have suggested that Google deliberately chose the Texas capital to put the heat on California, since TX has been aggressively courting CA businesses for several years by telling them they can free themselves from stifling regulation by moving to the business-friendly Lone Star State.]

In hailing the arrival of self-driving cars in Austin, Mayor Adler predicted the autonomous vehicles will reduce accidents and traffic congestion as well as provide mobility for people who can’t get around easily. The Austin police have declined to comment publicly about the experimental program. To avoid cultural shock, Google is putting its technicians in the driver’s seat of its driverless cars in Austin and assuring city residents that these faux “drivers” can quickly take control of the autonomous vehicles if anything goes wrong.

And what, you ask, could go wrong with a driverless car? Well, for one thing, early tests in California suggest that robot cars have trouble interpreting the idiosyncrasies of human drivers. For example, the Google cars are programmed to wait for all vehicles approaching a four-way stop sign to come to a complete stop before they go forward; the robot cars become paralyzed when human drivers in other vehicles inch their way into the intersection like we all do when trying to figure out who goes first.

During a test run in Mountain View, CA this year, a driverless Google car approaching a crosswalk was prompted by its “safety driver” software to stop to permit a pedestrian to cross the street. The human-driven vehicle behind the Google car promptly slammed into the rear end of the driverless car.

“The real problem is that the [Google] car is too safe,” Donald Norman, director of the University of California’s Design Lab told The New York Times in a recent interview. “They have to learn to be aggressive in the right amount, and the right amount depends on the culture.”

If Google plans to expand its driverless car experiment into the nation’s largest city, that learning curve will be steep. A polite, autonomous little bubble car that doesn’t know how to cut in and out of traffic in midtown Manhattan? Fuhgeddaboudit. Memo to Google: if you want to make it in the Big Apple, you’ll need to install a robotic middle finger in the space where the steering wheel used to be.

Among the other kinks that need to be ironed out before driverless cars rule the road is what to do if and when they break down on the highway and, more ominously, whether the software that controls autonomous vehicles could be remotely hacked by someone intent on causing vehicular chaos on America’s roads.

We may not have much time to figure all this out: Google has been dropping hints that it would like to put its driverless cars on the market as early as next year.