Use of agent based modelling to investigate the dynamics of slum growth

Abstract

Informal settlements arise as a result of the urgent need for shelter by the urban poor. Urban planners and policy makers face challenges in effective management of slum settlements as they do not fully understand their dynamics and extents. Advances in Geomatics research have recently offered growing results in slum characteristics using various remote sensing and artificial intelligence approaches. The main objective of this research is to propose a conceptual model for the implementation of an empirically informed agent based prototype that can simulate future patterns and trends in land cover change over time specifically with reference to informal settlement proliferation in the city of Cape Town in South Africa. The study incorporates physical, environmental, social and economic factors specific to Cape Town in structuring behavioural rules for agents in a predictive environment. Input data is extracted from a time series study of remote sensing imagery, ancillary data and statistics. The resulting concept model for the prototype incorporates a static model, a dynamic and an interactive behaviour model that collectively form a combo for successful implementation of the physical agent based model. On implementation the model is expected to simulate city wide slum growth patterns and trends in Cape Town over time, highlighting likely areas of new settlement and revealing the eventualities of existing ones. Urban planners can use pattern information for proactive slum management and in preventing risk prone settlement especially in some areas of near coast Cape that are flood prone.