Imagining the Driverless City

Robotic vehicles, drones, and other cutting-edge technological advances could soon reshape urban land use as radically as the automobile once did. Here are some leaders’ thoughts on how the future might look.

Building facades are part of the University of Michigan’s demonstration site, Mcity, where robotic vehicles can be tested. (University of Michigan)

Building facades are part of the University of Michigan’s demonstration site, Mcity, where robotic vehicles can be tested. (University of Michigan)

Design consultants Marcus Martinez and Amna Ansari of Alloybuild imagine “Shuffle City,” a matrix of compact urban cells, on-demand shared workspaces, and a shared mobility network of autonomous vehicles. (Alloybuild)

Design consultants Marcus Martinez and Amna Ansari of Alloybuild imagine “Shuffle City,” a matrix of compact urban cells, on-demand shared workspaces, and a shared mobility network of autonomous vehicles. (Alloybuild)

On a 32-acre (13 ha) site in Ann Arbor, Michigan, a radically new type of metropolis is taking shape. Mcity, as it is called, is a place where transportation consists solely of robotic vehicles that steer themselves without a human at the wheel. The driverless cars must navigate several miles of an urban street grid—complete with traffic lights, intersections, bike lanes, pedestrian crosswalks, and tunnels—and rely upon software, wireless communications, and sensor technology to get to and from their destinations while avoiding collisions with pedestrians and other robotic vehicles.

Granted, Mcity is just a simulation; the buildings are just facades, and the inhabitants who unexpectedly step off curbs are just mechanized mannequins. “It’s pretty minimalistic,” admits Jonathan Levine, a University of Michigan professor of architecture and urban planning. “You wouldn’t mistake it for a movie set, let alone a real city.” But even so, researchers at the university’s Mobility Transformation Center have designed the facility to offer a realistic test of how driverless cars might function amid the daunting complexity of a dense urban environment. It is preparation for a future that is fast approaching.

According to a study released in April by the Boston Consulting Group (BCG), major automakers such as General Motors are already rushing to add gadgetry that would give cars some ability to pilot themselves in low-speed, stop-and-start traffic-jam conditions or in single lanes on highways, or even to find parking spaces and pull into them without human help. Meanwhile, they and others—including technology giant Google—have been working to develop “level four” automation, in which cars are capable of taking over all the critical functions from human drivers. BCG predicts that the first fully autonomous passenger vehicles will hit the market by 2025, enabling humans to sit back and relax while a robotic chauffeur drives them to their destination—and returns later to pick them up at the curb.

That could be the starting date of a revolution that many observers think could alter the urban landscape as profoundly as the electric streetcar and the Interstate Highway System once did. “I think it’s going to be the most important transformational change in 100 years,” says Randall K. Rowe, chairman of Green Courte Partners, an Illinois-based private equity real estate investment firm. “It’s going to change the way we get around, the way we transport goods, and how we look at land use.”

Experts in a variety of fields—from technology to architecture and real estate—believe that driverless cars will be just the start of the revolution. They will combine with other advances in automated transportation, ranging from self-driving buses and trucks, to robotic motorcycles and small drone aircraft that someday may deliver packages, and, in combination, altering practically everything, including commuting patterns, the price of land, the mix of land uses, and even the design and construction of buildings.

A Boost to Walkable Urbanism

Rings of roadways for driverless vehicles could connect urban cells in the “Shuffle City” demonstration. (Alloybuild)

Rings of roadways for driverless vehicles could connect urban cells in the “Shuffle City” demonstration. (Alloybuild)

Not everyone, however, is eagerly anticipating a driverless future. Some worry, for instance, that robotic cars will not be safe. But autonomous vehicle advocates point to Google’s experience, in which its experimental vehicles have traveled 1.7 million miles (2.7 million km) on regular roads, with just 11 minor accidents—all of which were caused by human drivers in other cars, according to an article by program director Chris Urmson. A 2013 report by the Eno Center for Transportation, a Washington, D.C.–based think tank, predicted that driverless vehicles would make roads dramatically safer by eliminating the human factors that cause 93 percent of crashes.

“A human has two eyes,” says former San Jose, California, director of transportation Hans Larsen, who has studied the technology and its applications. “A vehicle will have 20 sensors. It can detect pedestrians and puppy dogs. It can see in the dark. It doesn’t get tired or drive drunk.”

A 2014 RAND Corporation report pointed to another potential problem cited by critics of driverless cars: Because they will free people to use their commuting time to work, watch movies, or even exercise in the back of a driverless vehicle, RAND predicted that they might encourage people to live farther out from urban cores and commute longer distances, thus greatly increasing both sprawl and congestion. That view is shared by skeptics such as former Vancouver city planner Brent Toderian. “It’s entirely possible that driverless cars will do more harm than good,” he says. “If they’re privately owned and you can summon them to pick you up and drop you off and then go park someplace, that actually will result in more trips on the streets, and more congestion—unlike things such as transit, walking, and bicycling.”

But proponents say that cities will benefit if driverless technology leads a transition away from ownership to a shared model. They envision something akin to a robotic version of Uber, which would allow city dwellers to use smartphones to summon cars from a driverless fleet that continually circulates throughout the city, or which waits on the periphery of neighborhoods.

Carlo Ratti, an architect and engineer who is director of SENSEable City Lab at the Massachusetts Institute of Technology (MIT), says studies indicate that sharing of driverless vehicles could result in “a city in which everyone can travel on demand with just one-fifth of the number of cars in use today.” That, in turn, would mean shorter travel times, less congestion, and a smaller environmental impact.

(University of Michigan)

(University of Michigan)

Some people also envision urban cores in which only driverless vehicles would be allowed. Larsen says getting rid of distracted, impatient human drivers in favor of robots would further walkable urbanism. “You’ll be able to move around with more confidence on foot or on a bike,” he says. “And if you have to go out of the neighborhood, you can just call up a car.”

Costa Samaras, an assistant professor of civil and environmental engineering at Carnegie Mellon University in Pittsburgh, envisions autonomous vehicle zones in dense urban cores that allow drop-offs between 7 a.m. and 10 a.m., and then convert to pedestrian-only thoroughfares until evening. Such zones would not necessarily need bollards to block traffic, since the rules could be transmitted to cars electronically, he says.

Others say that driverless transportation could also make the flow of goods and services into and within urban neighborhoods faster and more efficient. Samaras envisions the development of motorcycle-sized robotic pods, made of lightweight plastic or carbon fiber, that could deliver packages, reducing the need for bulky, cumbersome United Parcel Service and FedEx trucks. Others envision repurposing shared-car fleets as delivery vehicles at night, when they are in less demand, or having shops outfitted to handle automated pickups of products, so that people could send driverless vehicles to pick up their pharmacy prescriptions or dry cleaning.

Marcus Martinez, an architect with the Houston-based design firm Alloybuild, envisions aerial drones, similar to the experimental robotic copter for which Amazon is seeking a patent, making deliveries to windows on the upper floors of high rises. “If people are worried that the packages might be dropped, areas vertically associated with the drop-off could be landscaped, so that nobody could walk under them,” he says.

Larsen cautions, however, that driverless deliveries could exacerbate congestion. “We don’t want this to turn into a system that somebody uses to order a cupcake from a store five blocks away,” he says. Some regulation will be needed, he says.

Repurposed Parking Space

The University of Michigan’s Mcity Outdoor Lab, as diagrammed on the opposite page. (University of Michigan)

The University of Michigan’s Mcity Outdoor Lab,

Perhaps the most profound effect of driverless vehicles would be to drastically reduce the need for parking structures and surface lots, which today take up a third of land inside cities, according to Rowe. Some of the garages and underground structures could be converted into storage spaces for urban dwellers who live in micro units, while the unneeded surface lots would be available for commercial or residential development or green space.

“You could solve the affordable housing problem,” Rowe says. “Right now, the value of land is so high that it’s too costly to create [affordable housing]. But suddenly, you’d have all this land that you could take back and repurpose.”

If city governments can be persuaded to ease parking requirements for developers as a result of shifting to driverless vehicles, the cost of erecting buildings could be reduced by more than 20 percent, says Stephen Conschafter, a Washington, D.C.–based urban planner and master-plan designer. The result could be an urban construction boom—and more room for a wider mix of uses.

“Repurposing parking is likely to increase the quantity as well as the variety of people and businesses in the urban core,” predicts Issi Romem, an economist for BuildZoom, a San Francisco construction database startup. “Greater scale and variety could, in turn, make the place more attractive to a large group of people who crave a walkable urban lifestyle, boosting residential demand.”

“Right now, parking is a huge drag on developers’ creativity,” Levine explains. “It’s a big drag on mixed use because it separates people from land uses.” In addition, he says that a switch to shared driverless vehicles—in which transportation would become a variable, per-trip expense rather than a fixed investment—would change the economics in a way that would encourage people to stay closer to home and shop locally—and add dining and entertainment to their shopping experience.

Changes in Building

Drivers were allowed in the cars that toured the Mcity test site during its grand opening. (University of Michigan)

Drivers were allowed in the cars that toured the Mcity test site during its grand opening. (University of Michigan)

The transition to robotic vehicles also could profoundly influence architects, developers, builders, and urban planners, allowing them to reimagine everything from the structures they build to the flow and function of streets and spaces around them.

Since driverless cars would be pulling up to the curb to let off passengers, coming up with a way to speed up that process will be critical for preventing traffic jams in front of buildings. John Eddy, chairman of the North and South American infrastructure practice for the global design and engineering firm Arup, envisions ground floors becoming more porous, with numerous entrances on every side of the building. If lightweight, minimalist robotic pod-vehicles ever become commonplace, Eddy envisions an even more exotic solution—drive-in entrances equipped with car-sized elevators that haul passengers to their floor, and then ride back down to the street and drive off. “It would be ideal, since they don’t have to park the car in the building,” he explains.

At the same time, robotics might enable the construction of trickier, more flamboyant buildings. Carnegie Mellon University urban designer Mitchell Sipus envisions office or residential towers being fashioned from three-dimensional printed parts and assembled by aerial drones equipped with robotic arms. “There are things you can design now but can’t construct, because they’re too difficult or expensive,” he says. “But drones may make that possible.”

The streets around those structures also could be transformed by automated vehicles. Alloybuild’s Martinez, along with his colleague Amna Ansari, have created a demonstration project called Shuffle City, which reimagines car-centric Houston reshaped into a matrix of more compact “urban cells” connected by rings of roadways for driverless vehicles. Heavy, thick road surfaces may become unnecessary, he says, because vehicles would be made of lighter composite materials; crash avoidance technology would reduce or eliminate the need for a lot of steel reinforcement in the vehicles.

Traffic lights also could become a thing of the past, with the advent of smart intersections such as MIT’s experimental DriveWAVE, which would manage the flow of networked vehicles so that they slip past one another without coming to a stop.

Driverless vehicles’ ability to detect and avoid pedestrians could allow planners to redesign urban streets to be more suitable for walking and more aesthetically pleasing. “Instead of blacktop, you could have a more European style, with cobblestones,” Conschafter says.

Rowe suggests a more exotic possibility: Because robotic cars could be networked to move efficiently in smaller spaces, “they don’t really even have to be above ground at all,” he says. “You might be able to put the roads underground, and then take the street surface and turn it completely into walkable, human-scale urbanity. You could put in more green space, and create a closer connection to nature, with pocket parks, outdoor recreation space, and other things that would help make the densification of cities more pleasant.”

Experts describe such visions for a driverless future with an important caveat: They say it is critical for urban governments to work with developers and technologists to make sure that zoning and other regulations are revamped to allow for optimal benefits from innovation. “Parking requirements, for example, won’t change automatically,” Levine says. “When it gets to city councils, there may be a lot of political opposition, from constituencies who still favor the old lower-density development that promotes driving.”

Given the decades-long life cycle of buildings, Rowe hopes to see governments provide incentives for developers to start preparing now for the driverless future. One such idea, for instance, would be for regulators to ease parking requirements slightly if a developer designs parking structures or lots so that they can be more easily converted to other uses, rather than for maximal capacity.

“None of us really knows how fast this change is going to come about,” he explains. “But we need to be ready for it.”

Patrick J. Kiger is a Washington, D.C.–area journalist, blogger, and author.

Patrick J. Kiger is a Washington, D.C.–based journalist and author.
Members Sign In
Don’t have an account yet? Sign up for a ULI guest account.
Members Get More

With a ULI membership, you’ll stay informed on the most important topics shaping the world of real estate with unlimited access to the award-winning Urban Land magazine.

Learn more about the benefits of membership
Already have an account?