Just as Rome was transformed by the aqueduct and New York by the elevator; cities around the world are now being shaped by digital technology. Smartphones, Transportation Network Companies (TNCs—like Uber and Lyft), and the flows of online retail have all made the city of 2017 remarkably different from that of 1997. However, we are yet to see the fundamental shifts that previous waves of technological advances wrought on urban life and form. We are still in the early stages of the digital revolution, and it is difficult to see just how it will be manifest.
I’d like to offer three ways—from the small to the profound—in which I expect cities to change over the next three decades. I have chosen these less because they are the most important, but rather because they best illustrate how technology might transform the urban—not just by introducing new information into our spaces, but by making new behavior patterns feasible and desirable.
The Self-Driving Vehicle
The most visible change to the city we can expect is the emergence of the self-driving vehicle. It’s still difficult to say how quickly this technology will be adopted, but it’s safe to predict that the most obvious difference between images of 2017 and 2067 will be the types of vehicles on the street—as is true today if we compare photos from 1917 and 1967.
Much has been written on how autonomous vehicles (AVs) might be used—and for that matter how their usage might change cities—and the general expectations are as follows:1
First, self-driving vehicles should be dramatically safer than human motorists. Driver distraction, the cause of roughly half of all US road collisions today, would be effectively eliminated. Further, since self-driving vehicles can be programmed to obey the law and drive with high levels of prevision, the public realm in the urban environment will be much safer. After all, the potential for a car to run a red light or fail to yield is much lower with a computer following directions than with a frustrated urban driver.
Second, self-driving vehicles should eliminate the need for on-site parking. When a car can be programmed to guide itself to a parking space two or twenty minutes away from a drop-off point, the need for nearby parking diminishes. Vast amounts of space currently devoted to curbside parking could be opened up. Imagine if all of Manhattan’s arteries could be de-clogged of automobile storage and their throughput could be significantly improved without any other changes.
Finally, self-driving vehicles could enable far more efficient driving patterns. Instead of accelerating out of one stoplight and slamming on the brakes at the next, AVs can be programmed to travel at speeds optimized for safety, ride comfort, and energy efficiency. They can even adapt to surrounding driving patterns to improve overall system performance before all vehicles are self-driving.
The key phrase here is that self-driving vehicles can be programmed to do these things. They are choices—not inherent qualities of the technology—so to achieve these results, standards must evolve. If riders have the option to choose an ‘aggressive’ mode for their ride, the safety benefits of the AV will be largely lost. On the other hand, if vehicles always prioritize pedestrian safety, traffic might come to a standstill with pedestrians no longer fearing for their lives, crossing streets whenever and wherever they please.
These problems raise what is perhaps the greatest question regarding self-driving technology: whether it will simply replace today’s personal vehicles without the hassles of driving, or if it will foster an entirely new set of uses that are, on balance, better for the city.
If people abandon auto ownership, rely on self-driving vans, and are willing to accept slight detours with pooled trips, it is easy to imagine AVs creating more pleasant city streets. The International Transport Forum’s simulations for Lisbon found that shared-ride AVs could replace all bus service in the city, offering far better overall travel service to both drivers and bus-riders. However, if people own their own vehicles, and send them home to save on downtown parking rates, the AV would become an agent of increased congestion and sprawl. Whatever the case, it is safe to say that as the cost of driving falls—in the form of a driver’s time—we will see a spike in the average miles traveled per vehicle.
The Managed Curb
While the autonomous vehicle may be a visible player in images of the future city, photos are unlikely to capture a second remarkable change: the shift to a dynamically managed curb. Today, the curb is one of the most poorly-managed spaces in the city: under-priced, under-regulated, and over-congested. Most of it is used for automobile storage, and at best, spaces are reserved for different activities on a first-come, first-served basis.2 This lackadaisical management fundamentally conflicts with the just-in-time nature of urban transportation, and the double parking that plagues our busiest streets today are just the most obvious symptom.
I’d argue that digital technology will force planners to consider curbs to the point where they become some of the best-managed spaces in the city. Today, a driver looking to park their car for an hour or more is willing to go a few blocks out of the way. But a passenger in a for-hire vehicle expects to be dropped off right at their destination, especially important during pick ups when drivers need to reliably find their passengers and vice versa. Similarly, when consumers shops at a retail store, deliveries can be made in bulk to a viable loading zone near the store’s freight entrance. However, when a UPS or FedEx driver needs to deliver a parcel, they have only minutes per stop and must park as close to their destination as possible: a curbside spot.
Both for-hire rides and deliveries encourage double-parking because they aren’t there that long. One doesn’t double-park for an hour or two; it’s too obvious that you’ll get a ticket. But if you only need the space for ten, five, or even three minutes, then double-parking is fairly safe. Our parking enforcement system does not generally have the capability to enforce violations that are only three minutes long.
This creates the potential to make the curb highly efficient because the digital technology that enables TNCs and inexpensive home deliveries also makes it possible to manage their use of the curb. A Lyft driver knows exactly where they need a parking space, and with a five- to ten-minute lead time, when they need it: not “about 10:30am” but “from 10:32 to 10:35”. The same is true with a FedEx driver, who know where their next stop is and exactly how many deliveries need to be made there. Unlike drivers today—for whom arranging parking in advance would require a plus-or-minus window of fifteen minutes or more—these users can predict their own usage of the curb with great precision.
These same technologies can enable constant curb enforcement. Although GPS is too imprecise a technology for this use, the array of sensors and cameras that digital technology allows will make it possible to enforce and charge for curbside usage on a constant basis, making it a far better deal to arrange—and pay—for a three-minute parking space than to occupy someone else’s spot and be guaranteed a large fine.
Taken together, the demand and the potential both suggest that cities are likely to move to highly managed curbs. Citizens who rely on free on-street parking are likely to resist in the near term, but I expect these approaches to be first applied to areas where such parking is currently prohibited, like airports and downtown loading zones. Over time the crush of double-parking will become a bigger detriment to most neighborhoods than the lack of parking spaces—especially as self-driving vehicles come into widespread usage—and so the curb might become the most efficiently used space in the city.
The Inversion of Transit-Oriented Development
It is possible, however, that the most profound effect that digital technology will have on the shape of cities will be neither as visible as the self-driving vehicle nor as efficient as the dynamic curb. Rather, these advancements have the potential to invert the way physical transportation nodes have traditionally shaped urban form.
Cities have almost always been built on the unevenness of transportation access created by physical infrastructure.3 The most obvious example is the way that density naturally clusters around rail transit hubs; but it is true even at lower levels of ‘fixity’, around bus stops and highway interchanges. The value of land generally responds to proximity to transportation nodes, and for more than a generation, American urban planners have prioritized ‘transit-oriented development’ as a cornerstone for thoughtful urban growth.
Digital technology offers the potential to undermine the importance of these physical channels. If digital wayfinding becomes the standard, the ‘easy-on, easy-off’ location near the highway exit loses its value. In a world of flexibly dispatched buses, bus stops aren’t meaningful: instead of passengers congregating at fixed places to access a bus, buses will adjust their routes to reach passengers. And if the vision of a world of shared, self-driving taxibots becomes a reality, there is theoretically no location that is more or less accessible than any other, in part because driving time can be used for other purposes. Such uniform access to mobility suggests a return to von Thünen’s “undifferentiated plain”, where distance from a center is the only determinant of land use.4
But technologies that diminish the importance of physical channels altogether reveal the potential of these digital systems to generate channels of their own. It’s already been noted that Waze can create traffic problems where none previously existed; and that the invisible pricing systems of UberPool, LyftLine, and similar ride-pooling services can make some routes or points cheaper without anyone noticing. Rather than following any nefarious scheme, these would likely follow the logic of the transportation network–just as airlines charge higher fares at “fortress hubs” than at competitive airports—but can still lead to unintended inequities.
For example, if the logic of pooled-ride services become widely used, they should result in dramatically lower costs to travel on routes where additional riders are likely to be found. However, this could lead to unintended land-use differences on the basis of class or race, just as New York City yellow cabs still rarely leave Manhattan’s wealthy core because overwhelming patterns of demand rarely compel them to serve the outer boroughs.
This suggests that demand and pricing may one day become the key shaper of urban form. Routes with dependable levels of demand—or where certain premium customers are reliably present to generate profits for car services—will likely have lower costs and more frequent service. Routes with less demand—or, perhaps, with a high concentration of price-sensitive customers—may suffer. In this way, the urban transportation patterns of the future may be more akin to today’s airline map, where routes with reliable business travel (New York to London, for example) are much easier and cheaper than shorter trips with different demand profiles. This isn’t necessarily bad, but it isn’t necessarily good, either. It’s just different, and urban planners will need to consider both how traditional land-use decisions can shape these digital channels and how the invisible overlay of digital transportation management may be shaping their cities.
Imagining the future
All told, digital technology promises to transform cities, in ways that will range from the obvious and the desirable to the subtle and ambivalent. These are just three areas where one observer thinks change is likely. Few tasks are as fraught as envisioning the future, but the history of urban technology promises that the changes of the digital age are likely to be both profound and complex.
–