Self-Driving Cars Are on the Road to Nowhere
Technology companies have been selling a vision of greener cities and safer roads. It's nothing more than hype.
In the popular imagination, the rise of self-driving cars will mean greener cities, safer roads, and happier workers. With technology at the helm, speeding, braking, and reaccelerating—which are responsible for much of the gas burned on roads today and the resulting air pollution—would be a memory. Automated ride-sharing technology would also allow city dwellers, attracted by cheaper housing, to move away from urban areas while also avoiding the drudgery of gridlocked commutes. And, most important of all, roads would be safer, because driverless cars wouldn’t get drunk, drowsy, or distracted—the main root causes of the more than 1.2 million global traffic-related fatalities every year.
This vision has understandably prompted heavy public investment in the technology. Singapore, for example, spent millions of dollars building a dedicated facility to test driverless cars. Munich followed suit with similar efforts. And then there’s Beijing. The city stands poised to open a $2 billion technology development zone—one aimed at establishing it as an epicenter for self-driving technology.
But if this driverless future is attractive, it’s also almost certainly wrong. That’s true for at least two reasons.
First, truly automated technology—at least the kind that requires no human involvement—is a rarity in every sector of the economy, and there’s little reason to think driving will be any different. Industries including energy, agriculture, and transportation may heavily rely on machines to turn a profit. But they still demand human muscle in one form or another, because machines can (and do) break down. This means a perpetual need for a human backup in case of an emergency. In the case of driving, that means human drivers will still bear ultimate responsibility for the cars they’re allegedly merely riding in.
A related challenge—particularly for driverless cars—is the complexity of the systems involved. For example, some cars today require over 100 million lines of software code to keep them running. More code means more functions—like rear-facing cameras, anti-lock brakes, and live traffic updates.
But more functions also mean more opportunity for software errors. Just ask Mercedes. In 2005, the German automaker—which has a longstanding penchant for perfection—recalled over 1.3 million of its cars. The reason: reports that onboard systems, which worked fine in isolation, could not do the same when paired together. Some drivers also discovered that playing with the car’s navigation system would sometimes cause their seats to move.
Mercedes isn’t alone. Some cars manufactured by Japan-based Toyota have also been prone to system failures—the type that can potentially lead to unwanted acceleration of the vehicle. The cause was so-called spaghetti code: long lines of software so complex and intertwined that it is difficult—even for the engineers who created it—to fully understand.
Driverless cars pose a challenge in this regard. That’s because these cars are far more complex than their predecessors, requiring hundreds of millions of lines of code (the space shuttle required just 400,000). More code means more opportunities for errors—errors that can lead to accidents.
Yet the greatest hurdle facing driverless cars isn’t technological, but psychological. Surveys consistently reveal global skepticism about the technology. Americans, for example, remain more worried than enthused by driverless cars. Furthermore, over half of the American public says it would not ride one given the opportunity. This reveals a deep-seated lack of trust in the technology. Similar sentiments have been expressed worldwide.
Public opinion matters. The United States, for example, boasts one of the highest motorization rates (the number of cars per 1,000 people) in the world. It also has higher than average disposable income per capita. In theory, this means not only could U.S. cities benefit from driverless cars, but that residents would have the cash to afford them.
Instead, consumers stand poised to shy away from self-driving technology due to safety concerns, which, in turn, would affect sales. Safety is also being used by politicians—ever sensitive to public opinion—to pass laws making it harder for self-driving cars to hit the road. All of this makes a driverless future far from certain.
What’s more likely is that human drivers will continue wielding some control over their cars, regardless of how automated those cars are. Put another way, driverless does not—and will not—mean humanless. Some may quarrel with this idea. But automakers have long been preparing for it. Just ask Nissan, Waymo, and Zoox, all of which are exploring ways for human drivers to retake control of a car should automation fail. One industry executive called this setup “the ultimate backup system.”
That it may be—but it is also one fraught with peril. Keeping humans at the wheel challenges the narrative that driving will, thanks to technology, soon be stress-free. It also raises questions about whether the environmental and safety benefits tied to driverless cars can ever be realized.
These benefits are particularly important for cities like Beijing. China’s capital is notorious for its air pollution. Government officials there have been forced to ban some cars from local roads. The result: momentary blue skies. Driverless cars would, of course, make this effect permanent by encouraging eco-driving and ride-sharing. But that’s unlikely if drivers must still keep an eye on the road.
Under these circumstances, the safety benefits of driverless cars also become suspect. The paring-down of road fatalities demands removing flawed human decision-making from the driving process. That’s hardly possible if drivers must occasionally do otherwise.
This reality should worry cities banking on driverless cars. The likes of Singapore, Munich, and Beijing may, in the short run at least, benefit from getting national governments to spend locally. But the long-term payoff is poised to be slim.