The Algorithmic City

Rush hour traffic zipping in front of the Denver skyline.
Rush hour traffic zipping in front of the Denver skyline.

As a young professional living in Paris, I commute to work on a daily basis. If I'm lucky, my train will arrive less than 10 minutes late, at which point it will be so packed that I will be standing for 30 minutes, squeezed between people who obviously did not shower and felt like they absolutely needed to reach for the top handle bar above me. This happens in every city around the world, so somehow it must be a universal truth about urban living, right? Wrong.

Back in the old days, less than 10 percent of the world population lived in urban areas. That's what cities were designed to handle. Today, we reached 50 percent. By 2050, it will be 70 percent. Taking into account population growth, this means we will have 7 billion people living in cities in 2050, twice as much as today. Just imagine what morning traffic will feel like, or how unbearable public transportation will become. And we are not even talking about housing, energy, waste or food.

What we are experiencing here is an incapacity of existing cities to handle the growing number of people living in them. In other words, cities do not scale. The problem is that in virtually every case, inability to scale has led to failure. It happened to technology startups who could not handle their success. It happened to companies who could not ship enough products. And it will happen to cities.

Unfortunately, we cannot just take down cities and rebuild them, add more capacity in public transports, or create more physical space. The only thing we can do is become better at how we manage them. And the way we do it is simply by predicting what will happen!

This though, requires data. A lot of it. The more we measure what happens in our cities, the more we will be able to understand how they function, how they are impacted by various events, and how people interact with them. Once we understand this, we can build a model to predict future behavior, thus anticipating what will be needed and dramatically increasing the efficiency of existing infrastructures and services.

Let's take an example: predicting crime. How would we build such a model? One way would be to look at the history of crime in an area, and extrapolate from it. This approach is limited though, as crime patterns change, and cannot be observed until new crimes have been committed. This means the police will always be one step behind criminals. Fortunately, we can circumvent this by looking at what caused crimes in the first place, and using it in our models instead of just looking at past criminal records. This would allow police departments to intelligently dispatch their troops in places where crimes are most likely to happen, thereby preventing them from happening in the first place. This is called predictive policing and is already being tested in cities such as Los Angeles, with highly encouraging results.

What else can we predict? In theory, any event that is not random, provided we have enough data to model the context. Examples include passenger load in public transports, availability of parking spots, traffic jams, waste production, energy consumption and revenues of a shop in a specific street. These all share a common underlying principle: use context rather than history to predict behavior.

In themselves, each of these predictions could lead to amazing new products and services. The real power though comes from integrating everything together and modeling an entire city and its interactions with people. For instance, if you can predict where people will need to go tomorrow, then you can create optimal bus routes, minimizing time to destination and walking distance, taking into account predicted traffic, weather and garbage collection schedules. In this ideal system, all services would be optimal and available to citizens at anytime. We call this new way of designing cities "Algorithmic Urbanism".

Eventually, this will completely reverse the city-people interaction. People will no longer be constrained by cities, but rather, cities will adapt to people and to their changing habits. And this could happen today! We have the technology to do it, we have the data to model it, we have the talent to build it. All we need is for people to convince city operators, utility companies and governments to open their data so that more predictive models can be built, and cities efficiently managed.

If we can do that, if we can build this user-centered, algorithmic city, then and only then will we be able to sustain our growth and dramatically increase our quality of life.

This post is part of a series produced by The Huffington Post and The New Cities Foundation, to mark the New Cities Summit in São Paolo, June 4-6, 2013. The summit highlights what works to solve the great urban challenges facing all cities. For more information on the New Cities Summit, click here.