This post provides some background to the research I'm just starting to undertake over at CASA. In short my research involves using computational methods to help increase the energy efficiency of transport networks in cities. The research draws on complexity science and network theory, and the output is a set of tools & techniques that allow urban planners to make informed decisions about transport systems. This article is for the uninitiated and assumes no real prior knowledge of urbanism.

Le Corbusier wrote The City of Tomorrow & Its Planning in 1929. In it he sought an unprecedented clarity of form for the new metropolis; high-rise, automobile friendly, efficient & highly rational. His vision - an architected landscape of modernist forms arranged into strict typologies - eradicated the wayward path of winding streets & the 'chaos' of industrialised early 20th Century cities; polluted, congested, noisy, inefficient, poorly sanitised, seemingly out of control and lacking in open green spaces. He saw a Paris blighted by these issues and proposed a clean break inspired by the bolder forms of classical & medieval civilization he so admired; Athens, Pompeii, medieval forts & Roman settlements. In its place he sketched gleaming sky-scrapers designed for maximum sunlight, each serviced by a subway, large open spaces, multi-tiered roadways for different modes of traffic and strict zoning. Through contemporary eyes it seems a suffocatingly prescribed environment, a paternalistic intervention, the product of a megalomaniacal will or totalitarian control even. It's difficult for me to see it as anything other than the work of a very creative auteur rising to an almighty challenge, and very much a naive product of its time.

Corb's tract was a landmark of modernist urban design, built on an existing skepticism about the sustainability of the high-density metropolis, ingrained by Howard's Garden City paradigm that had already begun to take root with the first wave of American urban planners. His wide open spaces lured one side of the planning theory continuum and his ultra dense high-rises attracted another. It gained influence and led to the Athens Charter for functional cities, ushering in an age of city beautiful planning.

At the time of Corb's vision we had a very different notion of order. We also didn't have much data. Corb's proposals were based only on some top-level traffic, land use & population data; the Chicago School was just beginning to compile the first rigorous data analysis of urban activities around the same period. The science of his day suggested that rigid geometries equated to order and that the natural world was rife with randomness.

It wasn't until later in the 20th Century that we began to grasp urban form as a complex entity. In the Sixties complexity began to gain a serious mathematical foundation - Mandelbrot began to unravel the rational basis of organic form and Lorenz began to study the nature of chaos, whilst urbanists like Alexander were talking about the growth of cities in terms of genetics and the structure of cities in terms of complexity (e.g. pointing out that a city is not a tree). For the first time in history, we could literally see the order behind seemingly chaotic structures such as coastlines & leaves and dynamic complex systems like our brains or the climate. This fundamentally altered the way we perceived form, and by extension, urban form. Cities now looked to us like living organisms - emergent, vibrant and ever-evolving. An inter-disciplinary observation that self-organization often provides good-fit solutions to design problems was growing. A better understanding of order had radically changed our attitude to the built environment.

By this time both Mumford and Jacobs had damned Corbusier's visions of Radiant City, aided as they had been by decades of ill-conceived & poorly executed public housing projects produced in its image. This marked the beginning of the end of the utopian phase of planning which had begun with Howard's anarchic vision sixty years earlier and ended with city beautiful planning. Jacobs highlighted urban diversity as a result of complex interactions of mixed-use function, drawing attention to the subtle inter-relationships of land use in cities. The serious study of the city as a complex system had begun.

Complex & chaotic systems are essentially deterministic but difficult to forecast. Through processes like bifurcations, catastrophes & self-organized criticalities, very small fluctuations cause large differences in state. This property of complex systems is both the challenge set by as well as the basis of the government funded research project I'm joining, SCALE. By modelling human transit as energy flows in cities we aim to propose small infrastructural changes to public transport networks that can produce large boosts in efficiency for those same networks. In particular, we'll focus on reducing energy consumption. Our ability to forecast changes in emergent travel behaviour caused by changes in nodes, links, scheduling or capacity is based entirely on our modelling strategies and model resolution.

A lot of my time won't be spent on forecasting strategies for highly unpredictable systems: We'll be modelling individual decision making in multi-agent systems to better understand emergent travel patterns, using network theory to analyse the topology, properties & dynamics of transport infrastructure and gaining a better understanding of the mechanisms behind transport usage by analysing real data. To quote Jane Jacobs,

city designers should return to a strategy... of illuminating and clarifying life and helping to explain to us its meanings and order - in this case, helping to illuminate, clarify and explain the order of cities.

What's changed since this pronouncement is that we have increasing amounts of both big data and computing power at our disposal to better understand, analyse & model urban systems.

Some questions I might be trying to answer include: What is the relationship between transport networks and urban form? Which elements of this relationship are destructive and which positive for cities? How do we quantify robustness in our transport networks? How do you measure the efficiency of a particular bus route? What happens to car usage if energy costs rise? How do access times affect flow? Can we model how individuals make decisions about modes of transport? What do temporal models of travel patterns based on real usage data tell us about a city's transport needs and zoning? How can urban planning policy change these patterns in a way that reduces energy consumption? How can we distribute flows in a network to improve overall capacity without overloading any links? Can we measure the relative efficiency of transport systems in different cities? What makes some more efficient than others?

The outcome will be techniques & tools that will reach the hands of policy makers. TfL, amongst others, is affiliated with the project.