ANN ARBOR, Michigan -- A total of 122,000 miles of roads link Michigan's fertile central farmlands to its northern wilderness and its southeastern industrial core. But this month, 4.2 miles of new pavement will link that road system to the future, too.
Those streets are part of Mcity, a 32-acre test facility at the University of Michigan that's funded in part by automotive and tech companies. There, researchers from academia, government and private industry will explore two profoundly transformative automotive technologies: cars that drive themselves and cars that communicate wirelessly with each other.
The changes coming to cars will make recent decades' worth of automotive developments look like minor tweaks. Sure, electronic fuel injection improved fuel economy, airbags made cars safer and cupholders are convenient for coffee. But with computers in charge, we'll be able to take a nap on the way to work, avoid accidents before we can see them and send the car to pick up grandma. Cars will know when others are coming around blind corners and.
At least, they will if the industry can develop technology that's smart and reliable enough.
Mcity plays a role in this future. It lets sponsoring carmakers Ford, Nissan, Honda, General Motors and Toyota cooperate with each other and with technology companies like Verizon and Qualcomm to test self-driving vehicles in an urban environment. Mcity, which officially opened earlier this month, is designed to challenge self-driving car systems with everything from simulated pedestrians to stop signs that are faded to pink and covered with graffiti. Other difficulties include a roundabout (emblazoned with the university's blocky M logo), train tracks, traffic signals and a road section made of the same type of see-through metal grille that gives drivers the willies when crossing Michigan's Mackinac Bridge high above the link between Lake Michigan and Lake Huron.
Mcity will also influence national debates such as whether car communications should share radio airwaves with smartphones and other devices or have separate, protected airwaves. And where Google and other Silicon Valley companies University of Michigan's Transportation Research Institute and the Mcity project.in the early days of computerized cars, Mcity could help Detroit steal some of that initiative back, according to , director of the
"The future of mobility -- particularly as it's influenced by connected and automated technologies -- is playing out here," Sweatman said in an interview here in his offices at the university. "The facilities and capabilities we're providing are helping the industry in this part of the world move forward quickly. A little bit of healthy competition is a good thing."
Google Chief Executive Larry Page graduated from the University of Michigan -- but then headed to Stanford to co-found a company that now champions self-driving cars. Maybe the next bright student with mobility on the mind will stay put.
Keeping high-value jobs in Michigan is important to the state's politicians -- and the decline of the city of Detroit itself provides a vivid cautionary tale of economic failure. About 243,000 of the city's 683,000 residents fell below the poverty level in 2013, according to the US census. But the auto industry is broader than just Detroit, and the state's members of Congress are committed to keeping the economic engine running.
"Southeast Michigan is home to more engineers per capita than any other region of the country, and projects like Mcity will help grow Michigan's role in the auto industry of the future," Michigan Sen. Gary Peters said in a statement. Peters, along with Rep. Debbie Dingell and Sen. Debbie Stabenow attended the Mcity opening ceremonies on July 20.
When computers drive cars, fiddling with your smartphone while you're on the road can become fun and productive instead of dangerous and illegal. Computers don't get drowsy, have subsecond response times, can track multiple vehicles and pedestrians simultaneously and, with sensors looking in all directions, have no blind spots.
Sweatman predicts self-driving cars will cut accidents by a factor of 10. "There's no doubt the automated future, when we get to it, will be much much safer," he said. And when people see that, they'll demand self-driving cars, he predicted.
Google and others have begun, where permitted, but a dedicated facility has its advantages. For one thing, circumstances can be reproduced so engineers can judge precisely whether particular changes are improvements. For another, ordinary pedestrians and drivers aren't guinea pigs in self-driving car experiments. General Motors is eager to work with competitors on cooperative projects that would be impossible at a particular company's closed test facilities.
Vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, in which cars talk to each other and to road systems like traffic lights and highways, will help, too.
Kirk Steudle, director of Michigan's Department of Transportation, holds in his mind a vivid example of what can go wrong: a 193-vehicle pileup on Michigan's Interstate 94 in a January blizzard this year. With slippery conditions and poor visibility, cars and freight trucks slid off the road and into each other, killing a truck driver.
V2V and V2I send signals at the speed of light that there's a problem. In initial versions of the technology, cars will pass the warning on to drivers, but later, cars will act on their own.
"This technology may not have prevented the initial accident, but it would have prevented the follow-up pile-on," Steudle said. "Thirty-three thousand people died last year on America's roads. It should be zero."
Toyota is building V2V and V2I abilities into 2015 cars in Japan, and with its version of the technology, which it calls Super Cruise.
"Ultimately it will become mainstream," said Cem Saraydar, director of General Motors' electrical and controls systems research lab. "We decided to take a leadership role in this technology and start production."
Chicken or egg?
Saraydar acknowledges that V2V and V2I suffer a common computing industry plight: the chicken-or-egg problem. V2V is useful only when cars and infrastructure support it, but there's little incentive for carmakers and road builders to invest in the technology when it's not yet useful. Ultimately, GM believes the technology's day will come.
"Alexander Graham Bell had a similar problem when he was launching the telephone," Saraydar said.
It could come a lot faster if the US government requires it. In May, the National Highway Transportation Safety Administration advanced its schedule from 2016 to 2015 to release a proposed V2V requirement.
The industry and government have to embrace V2V and V2I technology soon, said Richard Wallace, an analyst at the Center for Automotive Research in Ann Arbor, or else it may not happen at all.
"If we don't do it very quickly, that window is going to pass," Wallace said. Google and other tech-savvy companies will sidestep the technology, relying instead on cars' built-in sensors and perhaps.
Another V2V challenge: interference. The auto industry has had a 5.9GHz band of radio frequencies available for car communications since 1999 but has made little use of it so far beyond research, and the Federal Communications Commission proposed opening the band up to use by anyone.
Mcity will let researchers put the technology to the test. Can V2V work, with a car successfully broadcasting news that it's suddenly braking even though nearby people are busy watching videos on their smartphones?
Mcity also will let Verizon test another idea that could lower the cost of the V2V and V2I communications technology. "Rather than embedding a DSRC [dedicated short-range communications] radio in the car, we are exploring how can we do it in smartphones," said Amit Jain, director of corporate strategy for Verizon's Internet of Things work.
Mcity also dovetails with a V2V and V2I study that's been under way in Ann Arbor since 2012 and that's now expanding to interstates and other roads around southeast Michigan.
"The goal is to have 125 miles around Ann Arbor and to have 20,000 vehicles equipped," including vehicles from manufacturers' fleets, private owners, and possibly the Michigan state government, said Steudle of Michigan's Department of Transportation.
The V2I infrastructure isn't cheap -- it's getting $20 million in federal funds to bring the radio communication infrastructure to part of I-94, for example -- but the broader V2V and V2I test will help MDOT assess the safety and economic benefits.
"That is the biggest question," Steudle asked. "How do you make a decision between fixing a bridge and deciding on roadside [radio] units that help traffic move?"
Challenging Silicon Valley
Google grabs self-driving car headlines, Tesla Motors' CEO Elon Musk thinks they're, and the extraordinarily profitable , too. It's no wonder that automakers see a competitive challenge coming from Silicon Valley.
But increasingly, there are bridges across the divide.
"We're seeing a lot of automotive manufacturers creating R&D labs in the Silicon Valley so they can be involved in the innovation happening in technology there, and we are also seeing a lot of large tech companies opening offices or hiring employees in Michigan to be closer to the auto hub," said Chris Borroni-Bird, vice president of strategic development at mobile chipmaker Qualcomm Technologies. Qualcomm, based in San Diego, is using Mcity to test vehicle-to-pedestrian communications and the Qualcomm Halo technology for charging electric vehicles wirelessly.
Silicon Valley remains three time zones away, but the computerization of cars means automakers are in practice getting closer and closer.
"The cultural transformation is taking place," Verizon's Jain said. "It used to be that mechanical engineers controlled the conversation at auto companies. Then a decade ago it transitioned to electrical engineers. Now the computer engineers are gaining the upper hand."