Figure by Pete Ferguson, Mechanicity team Ph.D. fellow
Figure by Pete Ferguson, Mechanicity team Ph.D. researcher

Main Themes

Since cities are essentially collective entities, a critical aspect of all urban agglomerations is the nature of the networks through which their social and economic activities take place. One only has to think of the way biological systems such as trees and plants grow to seek resources in the air or the ground by establishing branches to efficiently capture the food they need, to be aware that such principles also dominate the way we colonise our own human space. Hence the role of roads, transportation systems, and communications and logistical networks is of particular interest to any research into short-to-medium term policy considerations affecting, for instance, economic growth and energy efficiency, as well as longer term questions about urban morphology and sustainability.

Urban networks have properties that relate to the specific constraints of physical geography. For example, spatial networks tend to be planar in that a vertex exists at the intersection of every edge. This leads to a very peaked degree distribution for each vertex where the vast majority of vertices have low degree values (eg often just 1 to 4 streets segments join at each intersection)

Spatial networks can be collapsed into different, more complicated topologies by various means of aggregation. In transport networks, properties such as road names, lines of sight, transport category and geometry can be used individually or in combination to define aggregations of this planar graph. The resulting aggregated networks have very different properties of connectivity, and community structure.

Alternatively, weights can be applied to represent the varying cost of traversal from one edge or vertex to another in the underlying planar graph. For example, in spatial networks, physical geometry is particularly relevant and many edge / vertex properties can be calculated by measuring the metric distance or other geometric properties between individual vertices or edges through the network. These kinds of network are implicit in the majority of transportation modelling in that communication is made through a transportation network with particular traversal costs from each origin zone to each destination zone.

Changing network centrality under London flood conditions, by Pete Ferguson
Changing network centrality under London flood conditions, by Pete Ferguson

Network properties can be used to understand how specific social communities are reinforced by the constraints and opportunities networks provide for interaction. Properties of network centrality can also be used to understand how the underlying network potential of a location can heavily influence the distribution of land use and transport flow across the city and the likely location of future urban growth.

We are developing new operational land use transport models that build on developments in allometric and network morphology. These contains new insights into the way energy is featured in these models and we already have web 2 prototypes under development that let researchers and policy makers explore future cities and what if style predictions under different scenarios relating to energy use and distribution.