In 1978, A. Getis and B. Boots attempted to develop a spatial pattern theory using spatial process and geometric primitives as an organisational framework. Digital geographic data are becoming increasingly available. This availability together with electronic data highways is creating a data-rich environment which will require new methods of looking for pattern. Spatial autocorrelation is a mathematical construct which describes the degree to which close objects are similar or dissimilar. Remote sensing’s contribution to pattern recognition comes in the form of classification techniques. Unsupervised clustering is one of the most common image-classification techniques and in intent comes close to achieving the unbiased pattern recognition required for inductive geographic investigation. In 1994, White and Engelen used an urban growth and land use Cellular Automata (CA) model to explore the general principles of ‘urban spatial organisation’. Artificial life (AL) models differ from CA models in that AL models have separate rule sets for each entity, whereas CA models have uniform rule sets.