It’s been a futurist fever dream since at least 1939, when visitors to the General Motors pavilion at the New York World’s Fair marveled at a glimpse of the automotive future: a miniaturized model of a city prowled by 50,000 robot cars, controlled not by human drivers but centralized radio waves.
That future may have been delayed a bit (the GM pavilion boldly predicted that humans would be out from behind the steering wheel by 1960) but it’s no longer a utopian daydream or pulpish science fiction: Several car companies say they’ll begin selling vehicles that can drive themselves — at least part of the time — by the end of this decade. And Florida will be right there at Ground Zero of the automotive revolution.
“We want to be the state that’s ready for innovation,” said Ananth Prasad, Florida’s transportation secretary. “We want to be the state where entrepreneurs and great minds can come to try things, rather than the state that just says no.”
One of just four states that permit experimental driverless vehicles to be driven on public roads, Florida has been the site of tests to see how their crash-averting sensors react to sudden and vicious thunderstorms. A section of Interstate 4 between Orlando and Tampa has already been equipped with transponders that feed the cars information about traffic and road conditions.
And when auto manufacturers and government officials from around the country held a summit last year to hammer out some of the legal and technological issues surrounding the vehicles, they chose Tampa, Fla., to host it.
During a car show accompanying the summit, Prasad — such an enthusiastic booster of the concept that he sits on the board of the national industry group Intelligent Transportation Society of America — took one of the prototypes out for a spin. Or, maybe, it took him.
“It felt a little odd, sitting there with my hands away from the wheel, on an interstate with people going by me at 70 miles an hour,” he recalls. “But it wasn’t scary or intimidating. Of course, obviously I was in a car full of Google engineers if anything went wrong.”
Google, the Internet company that has been testing luxuriant Priuses and Lexuses festooned with radar, lasers, and video cameras that allow the vehicles to be driven by its software rather than humans, says it expects its “autonomous driving system” to be on the road by 2018.
Several car manufacturers are only slightly less optimistic. Nissan, Mercedes and Renault have all announced plans to market cars in 2020 that can drive themselves at least part of the time.
“A lot of people question whether we’re going to see this in our lifetime,” Prasad said. “I think they’re going to be very, very surprised. … Autonomous vehicles are much, much closer than we think.”
In fact, they’re already here — kind of. Several of the key technologies in creating a full-blown driverless car have already been deployed. Some Mercedes-Benz models sold in Europe can be put on auto-pilot in low-speed traffic jams; taking cues from the vehicles ahead of them, they stop or go, speed up or slow down as traffic permits, while keeping themselves inside the lane.
Volvo equips some cars with a pedestrian protection system that hits the brakes hard if the car’s radar spots a person in the street ahead. Even the lowly Ford Focus has a parallel-parking-assistance option that does all the maneuvering while the driver controls the accelerator and brakes and looks out for fire hydrants or no-parking signs.
“Automated driving, in a certain degree, already exists today,” said Joachim Taiber, director of the International Center for Automotive Research at Clemson University. “You can already buy cars with collision avoidance systems, emergency braking system, lane-departure warnings.
“That’s how we’re going to arrive at what people most likely mean when they say ‘driverless car’ — a robot vehicle with nobody behind the wheel that functions on all roads under all conditions. That’s not within reach in 10 years or probably even 20. Instead, we’re going to move toward that in phases, and actually, we’ve already started.”
The prospect of a truly and totally driverless world like the one to which Taiber refers can excite normally sober engineers into flights of both transcendental fancy and apocalyptic morbidity.
Surely many of the 31,000 Americans who die in traffic accidents each year would be saved if human drivers and their unfortunate tendencies to drink beer, text, apply makeup and race trains were taken out from behind the wheel — not to mention the elimination of things like fog-bound pileups and collisions due to rearview-mirror blind spots.
“That death toll — it’s like having an airliner crash every day of the year,” said Paul Feenstra, senior vice president of the Intelligent Transportation Society of America. “If that were to happen, we’d have congressional hearings every day to get to the bottom of it. But it’s become an acceptable norm on the nation’s highways, and it doesn’t have to be.”
The quixotic dreams of urban planners for vast tracts of parkland and bike paths and playgrounds might actually become affordable if cities could get rid of all their parking lots. And, maybe, they could.
“You probably don’t need parking garages any more once you have driverless cars,” said Chandra Bhat, director of the Center for Transportation Research at the University of Texas. “You’d just call an autonomous taxi to pick you up in the morning and return you in the afternoon. Somewhere from 20 to 80 percent of urban land area could be repurposed.”
Some theorists believe the entire foundation of private car ownership will be shattered. “There’s a lot of talk about people joining fleets or clubs, like Netflix for cars, where you’d just order one up when you needed it,” said Loren Smith, an analyst who follows government automotive policy for the research firm Capital Alpha. “You wouldn’t have a car sitting the driveway for 23 hours a day, doing nothing.”
The productivity, and possibly the space-alien-zapping skills, of American workers would zoom through the roof as commute times — traffic jams would be a thing of the past as computers mass driverless cars into perfectly spaced convoys traveling at uniform speeds — and workers can spend the time working on their laptops instead of trying to cut in front of the guy in the next lane.
“Especially in the information technology industry, a lot of people think of driving as a big waste of time,” said Clemson’s Taiber. “If you work on emails or reports, or play video games, or even just sleep, a lot of people would welcome that.”
The lame will walk and the blind see. Well, almost, anyway. Google thrilled the industry a couple of years ago with a video of a blind man in the driver’s seat of one of its automated cars as it carried him to first a fast-food restaurant and then to pick up his dry cleaning. Driverless cars promise new independence to the disabled and — not inconsequential in a country where the massive baby boom generation is passing into retirement — senior citizens whose driving skills have waned.
But driverless cars might also ruin the environment, worsen global warming and lay waste to our cities. “Once I can relax in my car on the way to work, read a book or watch a video, travel time won’t be a burden to me anymore,” UT’s Bhat said. “I might move further away, travel longer distances to work. That could expand the geographical footprint of our cities and increase our use of fossil fuels and massively increase pollution. It’s practically invited urban sprawl.”
We might even hear a giant, strangled glug-glug-glug as the American housing industry goes even deeper underwater. “If you don’t have to drive your car, you’ll want to have more fun in it, be able to move around and put bigger seats — maybe even a sofa — in it,” Bhat added. “Vehicles will get bigger and bigger. In an extreme case, would we even have homes anymore? Or just huge vehicles that become our giant mobile homes?”
All those scenarios, however, whether utopian or dystopian, probably lie well over the horizon. Not because the technology for truly driverless cars doesn’t exist — most of it is ready to go or soon will be — but because the road to that horizon is filled with legal, political and social potholes, some of them pretty deep.
One of the biggest: liability law. “No matter how good the technology is, there are going to be accidents, especially if you have a mix of human drivers and robots, which will be the situation for decades as the new cars penetrate the market. It can be something as simple as a tire blowout, or as complicated as a power surge that knocks out some of the technology while the human driver is asleep and can’t take over.
“So is the human driver — the car owner — responsible? The car manufacturer? The technology supplier? How do you apportion that out? Lawyers and insurance companies are going to be watching very carefully, and even though we have some models to work with — aircraft accidents that occur while the plane is on auto-pilot is a good one — it will be long and time-consuming.”
Technology failures might not be just legally complicated, but terrifying: What if a virus or a hacker invaded the software of vehicles and sent them reeling out of control? What if terrorists start using driverless cars to deliver suicide bombs that don’t require suicides?
Infrastructure is another problem that may take years to overcome. Nobody has even ventured a guess as to how much it will cost to embed America’s 8 million lane-miles of roadways with the network of transponders and sensors necessary to relay road and weather conditions to vehicles.
And the other critical communications link in driverless technology — vehicle-to-vehicle, which lets cars tell each other what they’re doing so they won’t collide — may be headed for political difficulties. It requires a big chunk of broadband, which the Federal Communications Commission set aside for carmakers in 1991.
But the proliferation of smartphones and video devices that stream movies and other video has gobbled up much of the broadband spectrum, and now the big cable companies have banded together to lobby Congress to let them share in the part reserved for autos. If the devices used by cable companies cause interference with the cars — nobody really knows the answer to that yet — that could lead to disaster.
“A dropped call on a cell phone caused by interference is no big deal,” Taiber said. “But the loss of even a little data on a car’s collision-avoidance system could be fatal.”
Privacy issues associated with driverless cars probably won’t kill anybody, but they may prove nettlesome. The technology in a driverless car will record exactly where it’s been and when, information that the owner might not want to have available to a nosy government agency or a truculent divorce lawyer.
Then there are potential corporate commercial clashes with privacy, too. What if Google, which knows you have a weakness for Big Macs because of information it has collected from your web browsing, strikes a deal and instead of mapping your driverless car on the shortest route to the office sends it to one that takes it past a McDonald’s?
“The privacy people are going to go crazy when they figure out all the implications of this stuff,” Robert Poole, editor of the industry newsletter Surface Transportation Innovations, predicted flatly. “It’s massive data collection, the GPS big-brother-box-in-the-car.”
Poole, though a fervent supporter of driverless-car technology, thinks there’s “a lot of mindless optimism” among its backers. He believes its adoption will take much longer than anyone expects and fall considerably short of the complete overhaul of human society that they hope for.
“You just aren’t going to see blind people traveling alone in these cars on freeways, or traffic jams disappear, or cities completely reengineer themselves,” he said. “As strong a technology as it is, things break and stuff goes wrong, and there will always have to be a human driver sitting there, ready to take control.”
WHAT THE FUTURE HOLDS:
Three upcoming key moments in the development of driverless cars:
—Any day now: The U.S. Department of Transportation will announce the results of a $25 million field test that began in Ann Arbor, Mich., in 2012. About 3,000 vehicles were equipped with sensors and wireless communications devices to see if they could successfully pass warnings about potential collisions to one another. The preliminary indications are good: The state of Michigan has already decided to take over the pilot program and expand it to 30,000 vehicles.
—Late 2014: The National Telecommunications and Information Administration, the federal office that advises the White House on telecommunications policy, is expected to come up with important recommendations about the part of the broadband spectrum set aside for driverless cars. Cable companies and other big broadband consumers want the spectrum to be shared with smart phones and other commercial devices; car companies are worried the smart phones will interfere with the new collision-avoidance systems on their vehicles. The NTIA’s preliminary announcements have seemed favorable to the automakers.
—Late 2016: The Obama administration says it will issue proposed rules establishing a common standard for vehicle-to-vehicle communications and a timetable for their installation in new vehicles. Vehicle-to-vehicle communications are the cornerstone technology of driverless cars, allowing them to warn one another of impending collisions, and a single standard will ensure that all vehicles can talk to one another. Once the rules are proposed, there will be a period of at least three months for public comment and debate, which could be a real turning point for smart cars.
Potential roadblocks for driverless cars:
—Liability law: America has more than a century of laws and judicial precedents on the books for who’s responsible for damages in a traffic accident. That all goes out the window when cars are driving themselves — probably way out the window if a human driver and a robot car collide. The plaintiffs’ bar, the trial lawyers who represent accident victims, is legendarily combative over changes in liability law. And insurance companies will certainly have an opinion or two.
—Broadband spectrum: Cable companies, one of the wealthiest and most powerful lobbies in Washington, are hungrily eyeing the part of the broadband spectrum set aside for driverless cars. The cable companies say their customers’ smartphones can safely share that spectrum with the cars, and besides, driverless cars are years in the future and smartphones are here now. Car manufacturers say driver safety should trump the ability of Starbucks customers to stream movies on their smartphones. If this dispute can’t be resolved amicably, or at least definitively, it will be devastating for driverless cars.
—Infrastructure: Just for starters, there are about 8 million miles of traffic lanes in the United States and about 300,000 traffic lights. Then there are stop signs, bridges that ice up, low-lying roads that flood in heavy rain — the list goes on and on. For cars to be truly driverless, they’ve all got to be equipped with sensors and transponders (and the lanes fully mapped.) Even when driverless cars start to roll off the assembly lines, it will be decades — if ever — before they really fulfill their utopian promise.