SimCity is becoming reality as dynamic simulation models allow cities to adapt their masterplans as political and social needs change and to trial what if’s to see the effect of potential future trends, says Elisabete Silva of the Cambridge Centre for Smart Infrastructure and Construction.
As hubs of economic and social power, our cities are constantly evolving as individual and collective decisions are made about their construction, adaptation and use.
Such changes present city planners, government departments and infrastructure owners and managers with their own unique set of challenges as planning must adapt accordingly.
"What happens if we subsidise certain housing or public transport? If we divert cars to a particular point, will there be a traffic bottleneck? What are the after-effects if we change an aspect of the local plan to intensify housing in a given area?"
Dynamic simulation models are required to deliver an adaptive planning policy that looks beyond static snapshots of analysis and scenarios. These models need to reflect the complex physical and social interactions to produce multiple simulation scenarios through time.
Collaborative research at the Cambridge Centre for Smart Infrastructure and Construction (CSIC) has resulted in a new generation of adaptive dynamic simulation models. These combine scientific methods and data collection/processing to show the consequences of planning decisions and link the physical environment with the socioeconomic, simulating the evolution of a city over time.
These models are already being used by organisations including Transport for London, Google and government departments, as well as by planners and decision-makers in cities including London, San Francisco, Shanghai and Lisbon.
This dynamic way of city planning is a quantum leap from what was once a very static process. Creating a master plan involved overlaying maps and then building from the plans over a long period. It didn’t work well because the static model was too simple – circumstances and conditions change quicker than a static master plan is able to respond.
For example, if a housing estate is planned today and tomorrow brings a sub-prime crash, or a demographic change pointing towards an increase of younger people who require smaller houses and flats, we are left implementing a plan – proposed years ago – on a reality that has changed; we cannot force reality to accommodate a static plan.
Today, access to big data feeds into our adaptive dynamic simulation models to analyse and create realistic and accurate simulations of what is happening in a city or region. The new wave of models, based on gaming technologies such as SimCity, flag up future trends and hot spots, enabling us to make more efficient, effective and sustainable planning decisions: Where is the incident that needs a traffic diversion and what are the best routes? Which locations call for a different type of housing?
"Adaptive dynamic simulation models produce plans that are sensitive to change and deliver greater economic and health efficiencies to society."
Data collected on a regular basis feeds the computer system to create scenarios that show consequences: What happens if we subsidise certain housing or public transport? If we divert cars to a particular point, will there be a traffic bottleneck? What are the after-effects if we change an aspect of the local plan to intensify housing in a given area? As well as showing the benefits, results may indicate traffic or health problems, allowing scenarios to be investigated, and the planning adapted based on the best outcome.
Demand for these advanced computer simulation models is high and the past five years has seen exponential growth in this area of research. I have specialised in simulation models for 23 years and contributed to the development of one of the most-used operational computer models for the simulation of land-use change, SLEUTH, and produced Countervailing Cellular Automaton (CVCA) to look at ecological landscape strategies.
Recently, in the research group I head at the University of Cambridge, there has been significant growth in the number of computer models that produce realistic scenarios of cities using real data to provide robust and useful results (CID-USST, for creative cities/industries and 4CMR's agent-based model of global carbon mitigation through bilateral negotiation under economic constraints).
Advanced computer simulation models such as these are able to deliver increasingly detailed information to decision makers for planning, managing, maintaining and regenerating our cities. We are now able to optimise, minute-by minute, the commuting routes for cars or people following an accident. We can also simulate land transactions over time on a parcel-by-parcel basis, taking into account how developers, local authorities, interest groups and buyers will react.
However, in order to answer specific questions tailored to the end user, we need diverse and detailed information, metrics against which to measure it and methods that include variation and change across years and along space. We need to understand the behaviours of those people and organisations involved in the process of changing space.
We need computers that can process vast amounts of data that allow us to visualise enormous quantities of information changing every minute, hour, month and year, and we need experts that understand new ways of planning. Researchers at CSIC are developing the tools and models to do this.
Adaptive dynamic simulation models produce plans that are sensitive to change and deliver greater economic and health efficiencies to society. An adaptive approach to the building and management of our cities is integral to a successful and sustainable economy.
Like the SimCity game, CSIC’s new generation of adaptive dynamic simulation models are designed to be easy to use by decision makers, planners, organisations and the general public. They can offer wide participation in the planning of the cities we live in and provide a responsive, pragmatic tool for the changing city environment.
Elisabete Silva is a CSIC co-Investigator and senior lecturer in spatial planning at the Department of Land Economy at the University of Cambridge.