It seems rather strange to be asking this question with the 21st century well underway, and with the threat of serious climate change on our horizon, but nonetheless it needs to be asked. When are we going to start thinking of cities as systems?
To be clear, when I talk about cities “as systems” I mean to think of, and engineer, cities in an integrated and holistic manner, in the way that we engineer and think about aircraft or space craft. Of course, cities do not have a single designer, nor a single design organization, and they need to deal with a host of political and social issues that do not arise, or arise in only a limited manner, in the design of even the most sophisticated space craft. Nonetheless, the systems perspective has much to teach us. It is borne of a unified concern for the entity as a whole, and is based on the understanding that the entity is composed of myriad interacting parts, whose interaction with one another shapes the overall behaviour or dynamics of that entity.
At the beginning, a systems perspective would say that we need to understand the overall dynamics of a city. This means to characterize the city in terms of a set of state variables and to understand how those variables relate to one another over time. Candidate variables might include things like population, population density, age structure, population distribution, temperature, concentration of organic biomass in the sewage and water systems, and the vehicle flow rate (vehicles/time) in a link of the traffic system. Clearly the number of such variables is very large, and the system interactions are complex. Nonetheless, I believe that there is merit in thinking about the system as a whole, and trying to determine appropriate levels of abstraction for urban analysis and design, secure in the knowledge that this can do no more than provide broad insights and general sign posts for how things might develop.
Having a systems perspective at the outset would help to eliminate the current information silos that exist in almost every urban environment. We would quickly see that our jurisdictional boundaries exist more or less to manage one another, but that they should not define the boundaries of information systems. If we took a systems perspective, information systems would be conceived and designed to integrate with one another. It would not be a matter of needing some special project to create an enterprise or pan-enterprise system.
Consider the following example. Suppose you take the train to the airport. The train travels through underground tunnels beneath the city, or on rails above the city, traverses the airport proper and arrives in the station. Electricity is supplied to the train from the electrical distribution system. This means that several different organizational units are involved in your journey, including the city administration, local electrical utility, local mechanical contractor (because you likely took an elevator to the train station), the supplier of electricity to the train, the transit company that designed, built and operates the train, and the airport authority where you arrive. How many of these organizations routinely share information with one another? Probably not very many. In how many places in the world are their respective information systems integrated with one another for operations, ongoing maintenance, and design of the facilities in question. I suspect that the answer is almost none. Taking a systems perspective would make this unthinkable.