Assisting an architect with alternative automated space layout designs using order crossover genetic algorithm in AutoCAD

Space layout planning is the first phase in the process of designing a building. Though a large number of possible layout plans may be generated, a set of requirements and related constraints limits the number of possible options. In this paper, an automated floor plan is presented using genetic algorithm, AutoCAD as an implementing tool and Autolisp as a programming language. The topological and geometric constraints such as adjacency between the layout subspaces and position and size of the subspaces are taken into consideration. The algorithm is evaluated not only for the uniqueness of solution but also for the generation of alternative designs, thereby offering the user alternative options to choose from.