The Practical Path to Driverless Cars

As Silicon Valley races toward the future of driving, Tampa is keeping a more manageable pace.

Image
Shutterstock.com

TAMPA, Fla.—I am not sure about this, but it's a pretty safe guess that Jason Bittner is mad at me. He drove to work because there were plenty of good reasons at the time for him to use the car to run afternoon errands: egg cartons to recycle at the supermarket, library books to return, and a consignment shop drop-off for the clothes his kids have outgrown. But then I showed up and made myself at home in his office, asking questions and not paying nearly enough respect to his indoor Green Bay Packers wind chime. It quickly became clear to him that none of his chores are going to get done. This is why Midwesterners think New Yorkers are rude, I guess.

Bittner is into "trip-chaining," he says, in his personal life as well as his professional one. His job is director at the University of South Florida's Center for Urban Transportation Research (CUTR, pronounced cutter, not cuter) — home, Bittner says, to the largest collection of full-time faculty studying public transportation in the country. That's 43 faculty researchers and about as many student-researchers, working on 180 projects, as well as a new division devoted entirely to driverless cars called the Autonomous Vehicle Institute, which Bittner was integral in developing. This is how far ahead CUTR plans: there are 23 saplings planted on the front lawn.

By the time those saplings mature, driverless cars may be here, and, like horseless carriages a century and a half earlier, they will be a quantum leap removed from their predecessors. We are just now taking the first baby steps towards that leap. Only four states (California, Florida, Michigan, and Nevada) allow driverless cars. And the federal government has only authorized a handful of public-road test beds — Tampa has one — for so-called "connected vehicles," which are not necessarily driverless.

The National Highway Transportation Safety Administration has defined five levels of vehicle automation. At Level Zero, vehicles can detect their environment, as with a collision alert, but rely on the driver to act upon that information. Level One gives the car control over minor functions, including automatic braking and parking assistance. Level Two allows for features including adaptive cruise control, with the driver still expected to be the master controller at a moment's notice. Level Three cedes total control for limited scenarios for limited durations, with a buffer for the driver to get used to the driving situation before manual override. Level Four is the idealized fantasy we mean when we discuss driverless cars: fully functioning vehicles that need no input from a driver.

With the Google Car already roaming its streets, California is hoping to fast-forward to a driverless future with an end-stage, Level Four gizmo — to change car commutes from active driving to passive riding almost overnight. But technology tends to advance in increments, with nuanced evolutions from candlestick phone to rotary to touchtone to cordless to brick cells to fliptop mobiles to smartphones. Silicon Valley is building the iPhone of cars. CUTR — and more broadly Florida, where the driverless car law does not allow for a car without a person in it, ready for manual override in emergencies — is embracing a far more practical, level-by-level approach. Yes, experts say driverless cars may arrive by 2040, but those same experts say fully manual cars may not be off the road until 2070.

•       •       •       •       •

Bittner often says we're in the middle of a "Kodak moment" for transportation. "Kodak owned film, owned digital film, owned it all," he says. "But it didn't see the future in it, even when that future started becoming obvious." There was a Detroit before Motown, a New York before Wall Street or Broadway, a Los Angeles before Hollywood, and a San Francisco before Silicon Valley. Heck, Orlando was more than a century old when Disney moved in. There's nothing really stopping Tampa from becoming the driverless era's Motown.

Jason Bittner (left), with Denis Eirikis of CUTR, at the Automated Vehicle Summit in Tampa in November 2013. (Katy Hennig | University of South Florida)

True, Florida is not Silicon Valley or Cambridge or even Austin. It is, in fact, a place so stereotyped for its local yokels that there is an Internet meme — Florida Man — set up just to pool all the half-witted antics that Floridians do (to say nothing of the Floridians approaching senility). In that sense, it's easy to see Florida as a driverless beta-testing paradise: a fertile proving ground to play out the adage that foolproof systems tend to underestimate how foolish people can be. Is a "Stand Your Lane" law for human drivers' rights so far-fetched of an idea for Florida's legislature in 2025? Or a driverless equivalent of a "hanging chad" debate?

The state is also primed for robot assistants. It's currently home to 26 metropolitan areas, 14.3 million people, and 100 million annual tourists. In 2012, Florida suffered 2,424 of the nation's 33,561 highway deaths (only California and Texas had more), according to the NHTSA. By 2030, a quarter of Floridians are expected to be 65 or older, their more alert driving days behind them. Tampa, in particular, has many of the social ingredients experts think of when they envision the driverless era: retirement communities, college campuses, military bases, hotels and theme parks — contained settings well-suited for autonomous shuttle services.

"Google has a lot of data from its fleet," says Bryant Walker Smith, resident fellow at the Center for Internet and Society at Stanford Law School, who's written about the legality of autonomous cars. "But it's data that shows you how specially trained, highly educated, early-adopter types mostly in their 20s and 30s can navigate the roads of San Francisco, where it doesn't snow or flood or all these things, in cars that are much nicer and newer than the average American's car." Here again Florida — Tampa, especially — is primed as a practical test bed (minus the snow), with its sizable population of college students, English-as-a-second-language citizens and immigrants, senior citizens, and tourists.

Downtown Tampa. (Chad Palmer /Shutterstock.com)

Technological progress is littered with short-sighted decisions, or at best decisions that were good but short-lived. (Just consider the nation's derelict, expensive pay-phone system.) We are always slaves to our present-day preconceptions; Smith likes to show a clip from those greatest of all autonomous-car evangelists — The Jetsons — where, despite technology that has cars flying at the flick of a joystick and folding into briefcases, women are still bad drivers. "It is literally the cartoonish stereotype of the 1960s preserved in their forecast for the 21st century," he says. For that matter, there were plenty of crashes and traffic jams in The Jetsons, too.

Silicon Valley's finish-first approach has its limits, he says. "We are not talking about technology. We are talking about technologies, plural, working in concert in scenarios even the brightest engineers can't imagine." He notes the recent massive recall by General Motors of 1.6 million cars whose ignition switches were faulty because their design allowed for weighty keychains to jostle the keys out of place, shutting off the cars, including shutting off air bags.

If Apple's map service bombs, or Facebook's privacy settings trigger mass confusion, or even if someone loses their life savings in Bitcoin, it's all pretty inconsequential. A customer would take headaches over head trauma any day. But Silicon Valley is used to unveiling tech idols to worshipful early adopters who happily bumble their way through so-called "hiccups" or "bugs" in, say, Google Glass or Twitter. Nobody dies when Siri doesn't understand what you're saying. But what if Siri were your chauffeur?

"We won't be relying in the future on a solution we think up in 2015," says Smith. "Silicon Valley will have to rethink the way it rolls out technology — not just the approach, but also the audience."

•       •       •       •       •

Bittner knows Florida's own battle is uphill. On a wall of his office he keeps an infographic poster called "Death & Taxes," which illustrates where taxpayer money goes. He pointed to a tiny circle, the U.S. Department of Transportation, and noted its research wing supplies a great deal of CUTR's $14.5 million budget. I asked him which of the pie slices on the tiny circle is the research budget. "It's, uh, not big enough to be on this," he says.

The CUTR campus, home of the Automated Vehicle Institute, complete with saplings. (Katy Hennig | University of South Florida)

Bittner's is the universal battle cry of the academic: we need more research. In that way, he is not only charged with being the leader of the institute, but defender of its existence as well. (Or, at least, was so charged; shortly after our meeting, Bittner left CUTR to settle in Madison, Wisconsin, where his wife took a job.) I was on campus in late February to hear him brainstorm for the first time with local academic, commercial, and political leaders on a multi-pronged approach to CUTR's future and how to position Tampa as the go-to hub of autonomous-vehicle innovation. But first, the day beforehand, he was tinkering with a PowerPoint presentation he would give to a class of 21 college seniors in a course called "Transportation and Society."

He came out of the gate swinging, reminding the students of a recent fiery wrong-way car crash that killed four USF fraternity brothers. What unfolded was a lecture so filled with panic and Reefer Madness-style suspicion of fully autonomous cars that one wondered if Bittner's Midwestern ancestors were not horse traders bankrupted by the arrival of the horseless carriage. A sampling:

"Will you ride it first? Will you? Will you send your kids to soccer practice in an autonomous car?"

"Would you trust Google Translate at 100 miles per hour?"

"For every job autonomous cars create, it takes away seven jobs, but mostly those jobs are running around in circles at the airport."

"If Mexico becomes a hotbed of automation, will we let driverless Mexican cars into the United States?"

He noted his car, a manual-transmission 2000 Volkswagen Golf, has seat warmers, intermittent wipers, and "a cruise control that works some of the time." When he raised the specter of hacking autonomous cars by saying, "What if I don't want the Broncos to beat my Packers in the Super Bowl, so I hack their bus and drive it off the side of the road?" some of the students outright cringed and muttered "Too far" and "Why?"

He showed a video by the Florida DOT that called autonomous cars part of "a century-long vision for independent mobility for all." It didn't mention CUTR. But, in rotary-to-touchtone fashion, it went through the five levels of autonomous vehicles, ending the final fully-autonomous description with the caveat that "this level of automation is still years — if not decades — from reality." Sorta-maybe-kinda up there with flying DeLoreans.

That's fine with Anthony Townsend, research director at New York University's Rudin Center for Transportation Policy and Management, as well as research affiliate at the Silicon Valley-based Institute for the Future. "I like this Tampa approach," he says. "This is the darker, harder challenge. They're starting with user needs, not imposing solutions. For any Google Car to work everywhere, it will have to adapt to regional and demographic needs anyway, so it's smart to start from that point. We're not good at imposed solutions: look at how slow the rollout of hybrids has been. And they shut down Google PowerMeter because the engineers made all the decisions without realizing that customers are not algorithms. The last manually driven car will probably still be on the road until 2070. We need a system that accommodates that."

•       •       •       •       •

To rough-sketch that system, Bittner headed a meeting the following day of about two dozen local leaders — overwhelmingly white men well into middle-age — to break into four groups for back-of-the-napkin brainstorming on business opportunities, transit and transportation demand management, policy and planning regulations, and technology and test-bed strategy.

Joe Waggoner, executive director at the Tampa-Hillsborough Expressway Authority (THEA), set the tone before the breakout sessions. "More importantly," he said, "we want to make Tampa Bay a leader in the state and a leader in the country, if possible. If you don't think big, stay home. Maybe there's a potential to generate a new hi-tech industry here. Maybe we can attract manufacturers and software designers, programmers." He added: "We're in the transportation version of eHarmony here. We're creating our profile, looking for a new relationship."

It was an odd mix of hope and desperation, of rhetoric both inspired and generic, more Monorail Song than Moon Speech. Tampa is not a Rust Belt town hoping to revive its steel mill. I kept thinking this is what it must've been like at the fateful meeting in 1885 that would brand Tampa "the cigar capital of the world." Would they even want to become an autonomous-vehicle hub if they were still the global cigar capital? Maybe. Maybe not.

Chris Spencer, an aide from State Senator Jeff Brandes (who pushed Florida's autonomous vehicle law) joked: "Just ban the words 'sky' or 'net.' I guarantee we're going to get calls from constituents saying they don't want their car talking to other cars. They don't need people — especially government — knowing their business, their whereabouts."

"What's the engineering world's most low-hanging fruit?" Bob Frey, planning director at THEA, asked at a session on policy and planning. "What could we use to have an interim solution? If you want people to use it, keep it as close to what they know as you can."

•       •       •       •       •

The meeting had a blue-collar sensibility probably missing from Silicon Valley Skype sessions: if driverless cars rely on visual cues of where the lane markings are on the road, Tampa is going to have to re-prioritize the way it maintains roads. Maybe fresher paint is more important than patched potholes in an autonomous world.

At another meeting, this one on technology and test beds, Peter Brett, a traffic manager with the City of Tampa, noted that Tampa's traffic signal system is from the early 1980s, with the bandwidth of the time — "and even then it wasn't state-of-the-art, so we're probably talking about a 1970s or even 1960s infrastructure," he says. Is that a weakness? Or a strength? Google Fiber, after all, started in Kansas City, not Austin, Boston, New York, Portland, Seattle, or even San Francisco.

At yet another meeting, on business opportunities, Mark Sharpe, a local county commissioner, was griping about bureaucracy. He recalled a time when a company asked Hillsborough Area Regional Transit to test out a new lubricant, and the agency was so hesitant that the company went elsewhere. "We need a one-stop shop, an entrepreneurship center, a point of contact to say, 'We can work you through our local system,'" says Sharpe. He could've just as easily been talking about filmmaking tax credits, Obamacare, or college financial aid counseling.

As she packed up her things, Lisa Montelione, a Tampa City Council member who chairs the city's Livable Roadways Committee, joked, or half-joked: "Too many people in this town think about 'the future' and think only about their retirement — their own future, not the city's future."

Bittner, for the record, got most of his tasks done on the day we first met in his office despite my distracting him (though the consignment shop was closed). He said he forgave me: "You can't do it all at once all the time."

Top image: kosmos111/Shutterstock.com

This article is part of 'The Future of Transportation,' a CityLab series made possible with support from The Rockefeller Foundation.

About the Author

  • Richard Morgan is a writer in New York whose work has appeared in a number of publications, including the New York Times and the Wall Street Journal. His first car, acquired in 1995, was a brown stick-shift 1983 Honda Civic he bought for $600, spray-painted with all manner of graffiti and named "The Free Expression Car."