Will Driverless Cars Dominate Our Future?

I’ve long been fascinated by the study of traffic. Those hours we all spend in the soft glow of brake lights help researchers turn questions about human behavior, urban planning and risk assessment into some stunning, slow-moving statistics.

Here are a few. Every year the average commuter spends about 34 hours crawling in traffic. That translates to an extra $713 in lost productivity, wasted gas and higher costs for goods delayed by congestion. Thousands of scientists study traffic patterns and city growth, but few do it better than the aces at the Texas A&M Transportation Institute. Those numbers come from TTI’s annual Urban Mobility Report, which has other illuminating findings. One I particularly like: Friday rush hour in America is the worst of the week, racking up 200 million more people-hours than Mondays.

With a growing population, more cars being built and limited road space, it’s a tough problem to solve. Urban planners have divided themselves into a few different camps. Some believe the answer is to make highways bigger. Adding a lane to the busiest roads–like Los Angeles’ 405 Freeway, the nation’s capital of congestion–can alleviate the flow of cars. Although some studies show that when cities add new lanes, all they’re really doing is inviting new cars onto the road, which makes the problem stay stagnant at best.

[vimeo 37751380 w=250&h=250]

What peak traffic efficiency might look like.

Then there are folks who get behind congestion pricing. Just like we have toll roads today, they support making more roads accessible only to people who want to pay. It’s a free market approach to a societal problem that grants open roads to the highest bidders—and in theory, clears up the regular roads for the people who opt out. “The idea is to allow cars to use the road as intended, at appropriate speeds, predictably and efficiently,” writes The Economist’s Ryan Avent, author of the book The Gated City.

There is a lot that makes sense about it, including boosting cities’ thinning budgets. London is trying it with some success. But the idea has problems too, such as being sluggish at changing prices based on ever-changing demand. It also solves the problem mostly for affluent people. Until now, I’ve always thought of traffic as the great equalizer. Rich or poor, we all have to sit bumper to bumper sometimes.

Which brings us to driverless cars, a technology still in its infancy, but appears the most promising as a game changer. At its heart, traffic is rarely caused by narrow roads or major blockages. Traffic researchers have found that the chief cause of congestion is human decision-making. Thousands of cars on the same road mean thousands of individual drivers making snap judgments. How far should I stay from the car in front of me? How fast should I go when things start moving? Should I rubberneck at that accident? Road traffic can’t really flow at peak efficiency because drivers are acting out thousands of slightly different interests—often irrational interests, some economists would say.

The idea behind driverless cars is to outsource that irrationality to a hyper rational computer. Loaded with censors, cameras and a super fast internal computer, the car itself can make decisions about speed, keeping a safe distance, and just when it’s safe, make a left turn with oncoming traffic. Then it can interact with other cars’ computers to know, for example, at the upcoming intersection. Google is one of the top companies working on this technology, which already exists. GM, Audi and BMW are already at work designing prototypes. The challenge might be implementation, and convincing even the most stubborn among us that letting a computer take the wheel might actually be safer and faster.

We’ll be returning to traffic and driverless cars more on Change Reaction, as I see them as an undeniable part of our future. Sure, when it comes to safety—including what happens when some cars on the road become driverless and some still have people behind the wheel—there’s still a lot more research to do. But I don’t think it’ll be supremely long before we’re all being chauffeured around.