Thursday, June 2nd, 2011
[ I ran out of Drew Austin pieces to repost from Where, but lucky for us he wrote this original that I’m sure you’ll enjoy. Speaking of which, Brendan Crain has restarted the Where Blog, which is really awesome news. This one is an absolute must to subscribe to in my opinion. So check it out if you aren’t already familiar with his work. ]
Architecture has borrowed plenty from biology over the centuries, but the reverse is less common. After observing the central dome of St. Mark’s Cathedral in Venice, the paleontologist Steven Jay Gould coined the term spandrel, describing a side effect of adaptation that turns out to be useful in itself. Like the architectural spandrel–a triangular space where two arches meet–the biological spandrel may seem perfectly designed for a certain function though it’s actually more of a lucky accident.
Today, for anyone interested in understanding cities or improving them, data on urban phenomena represent a different kind of spandrel. Much of the data offering useful clues for city planners and researchers is out there accumulating whether anyone wants it or not. The ability to track the origin and destination of every taxi trip in New York or Boston, for example, was a recent byproduct of the self-swipe credit card technology that those cities have required all cabs to install. Last year, an article in Wired described how New York City harnessed the knowledge available from 50,000 daily calls to 311, using that information to map the distribution of problems like noise complaints throughout the five boroughs. Maximizing our understanding of the city means discovering what’s already out there–the spandrels–as much as it means actively collecting data when necessary.
The taxi perfectly illustrates the opportunities these incidental data sources create as well as the limitations of what they can tell us. Transportation is one of the trickiest and most critical problems in any city, and one of the areas where good data can help the most. By recording their own activity, taxis become sensors that roam the city painting a detailed picture of traffic conditions, travel demand, and even the locations where passengers give the best tips. We can learn a lot about the city from taxis, but we can learn even more about taxis themselves and their role in the urban environment.
It’s easy to forget, but the taxi has always been a critical form of public transportation. In cities without good transit, the taxi is often the only public transportation available. More importantly, mass transit cannot efficiently serve every type of travel that passengers demand, and the taxi is better suited to do so in many cases (think of the bus that never has more than a handful of passengers on board). Low-income city dwellers as well as the affluent rely on taxis where buses and trains don’t suffice. In the United States, where everything is seemingly built for the private car, modes of transportation that improve mobility for the carless are allies, not competitors.
Because taxis are privately operated and can’t be planned like mass transit, the opportunity they represent receives less attention than it should. Now, taxis are still private, but the rich data they generate means they are no longer the blind spot for transportation planners that they once were. We may not know exactly how to improve taxis, but we can start by deciding how they might ideally serve a city and then observing how they currently measure up to that ideal. Beginning with the principle that taxis are a form of public transportation, they should complement mass transit by filling in the gaps where transit service is less accessible, and the aforementioned taxi trip data makes it possible to see whether this happens naturally. As the maps below indicate, more taxi pickups happen where transit access is also quite good. Of course, demand is also higher near Boston’s center, and it’s difficult to say where unmet taxi demand exists (although it’s possible to infer this). As a “spandrel,” taxi data alone won’t tell us everything we need to know to answer a question like this–that is, the data collection wasn’t designed with this particular question in mind–but the more we grapple with the data, the more we learn what it can and can’t tell us, and the more useful it becomes as a means of enhancing taxicabs or countless other aspects of city life.