There exists many misconceptions about design computation in architecture and it is mostly because methodology is acquired from computer sciences through informal studies. Curvilinear geometry and generative methodologies are not intrinsic to computational architecture. The digital computer however provided significant processing capacity that elevated computational design into a field of its own.

Much of the innovation in computational design took place in 1970s. Only the recent developments in technology and software, especially the graphical algorithm editors, have provided new momentum and wider acceptance for computational design. There exists formal computational precedent models capable of generating floor plans. The scope of their quality varies from designs barely recognizable as a floor plan to designs that have succeeded in architectural competitions. There is a lineage of models that incorporate higher-level algorithmic constructs: Metaheuristics. The subject is introduced and evolutionary metaheuristic are explained in general since they are most widely used in architectural computations. A specific evolutionary algorithm, found in Galapagos, is analysed and applied as it used in the model demonstration.