Tucked away in the basement of an iconic office tower shaped like four engine cylinders, engineer Werner Huber is telling me about the joy of driving. We’re here at BMW headquarters, in Munich, Germany—capital of Bavaria, and arguably of driving itself. But Huber oversees strategic planning for advanced driver assistance systems, so in a way, his job is to put an end to driving—at least as we know it.
“I think that in 10 to 15 years, it could be another world,” Huber says. He’s not willing to predict exactly what driving will look like then, but he’s certain humans will be doing a lot less of it.
For many people, automated cars call to mind those high-tech vehicles with a rotating periscope on top that Google has been driving around California. But Huber and executives at other European automakers say the automated driving revolution is already here: new safety and convenience technologies are beginning to act as “copilots,” automating tedious or difficult driving tasks such as parallel parking.
“Driverless” technology will initially require a driver. And it will creep into everyday use much as airbags did: first as an expensive option in luxury cars, but eventually as a safety feature required by governments. “The evolutionary approach is from comfort systems to safety systems to automatic driving,” says Jürgen Leohold, executive director for research at Volkswagen Group in Wolfsburg, Germany.
Both BMW and Volkswagen are among the companies already demonstrating cars that drive themselves. In 2010, Volkswagen sent a driverless Audi TTS up Pike’s Peak at close to race speeds. Like similar vehicles from Google, these automated vehicles use some combination of GPS, radar, lasers, ultrasonic sensors, and optical cameras to create a constantly updated, 360-degree model of the surrounding environment, which an in-car computer can use to navigate.
But European automakers say their strategy is to move toward greater levels of autonomy incrementally, depending on what does well in showrooms.
Buyers of European luxury cars are already choosing from a menu of advanced options. For example, for $1,350, people who purchase BMW’s 535i xDrive sedan in the United States can opt for a “driver assistance package” that includes radar to detect vehicles in the car’s blind spot. For another $2,600, BMW will install “night vision with pedestrian detection,” which uses a forward-facing infrared camera to spot people in the road.
Lasers, cameras, and other sensors are the most expensive part of autonomous driving systems. Some experimental self-driving cars are estimated to carry more than $200,000 worth of cameras and other gear. Those costs are also leading automakers toward a gradual approach that starts with sensor technologies and then extends capabilities to control driving tasks as well. In the high-end Mercedes-Benz CL, for instance, cameras not only tell a driver when he or she is leaving the lane but actually help the vehicle steer itself back. Several automakers already sell cars with so-called adaptive cruise control that automatically applies the brakes during highway driving if traffic slows. Next, BMW plans to extend that idea in its upcoming i3 series of electric cars, whose traffic-jam feature will let the car accelerate, decelerate, and steer by itself at speeds of up to 25 miles per hour—as long as the driver leaves a hand on the wheel.
The market for “advanced driver assistance” technologies was $10 billion last year, according to New York’s ABI Research, but it projects that the figure will reach $130 billion by 2016.
Some experts, however, think the evolutionary approach could reach its limit in a few years. Stefan Solyom, an engineer at Volvo, based in Gothenburg, Sweden (but recently purchased by Chinese auto group Geely), warns that incremental thinking can take the industry only so far. “At some point,” he says, “we need to make a dramatic shift when the vehicle is able to take over fully automatic tasks.”
The reason, Solyom says, is that the biggest benefits of automation can’t be realized until a driver’s hands are off the wheel. Volvo this year demonstrated self-driving cars that can travel together in a highway convoy, something Solyom says could cut fuel consumption by 15 percent (the cars would brake less often and be able to draft one another). Automation could also eliminate many traffic accidents caused by human error, estimated to account for 80 percent of all crashes. Finally, not having to pay attention would free up countless hours for “drivers” to focus on other tasks, like reading a paper or responding to e-mail.
Automakers say crossing the technological inflection point to full automation will require significant legal changes. In many countries, laws require a human driver to be in control of motorized vehicles or be able to take over instantly. Another obstacle, say manufacturers, is uncertainty over whether they will be held liable for accidents involving driverless cars.
Sven Beiker, a BMW veteran who is now executive director of the Center for Automotive Research at Stanford University, says automated vehicles are likely to face public-relations problems. “We currently have, sadly enough, around 33,000 people killed in road accidents in the United States every year,” says Beiker. “Imagine we can bring it down to 10,000 with autonomous vehicles, which would be a huge success. Would the headlines read something like ‘Autonomous Vehicles Save 23,000 Lives’ or would it be ‘Robot Cars Kill 10,000 People’?”
Consumer themselves are likely to be the final obstacle. For years, BMW’s advertising has promised control over “the ultimate driving machine.” Now, as the machine takes charge, companies need to persuade consumers to go along. One example is the BMW “parking assistant” feature that uses ultrasonic sensors to take control of the steering during parallel parking. It is sold with the tag line “Just because you can do something doesn’t mean you always have to do it yourself.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.