Notes on the potential of simulation for architectural conception

In recent projects EZCT Architecture & Design Research ( hereafter EZCT ) investigated the potential of simulation applied towards architectural conception. What does it imply to know the structural, thermodynamic capacity as well as the daylight penetration of a building a priori for architectural conception? EZCT initiated the CCD project in order to understand this epistemological shift. The CCD project started the collaboration with Marc Schoenauer, who is a leading researcher in the domain of evolutionary computing and machine learning. For the CCD projec1 a genetic algorithm was develop that evolves a volume such that the forces inflicted upon it are dissolved as efficient as possible, while reducing the volume. One of the aims of the project is to arrive at a mechanization of the design process, such that instances of the concept can be produced fully automated. This is indispensable to fully develop the potential of non-serial fabrication of objects. In his article “Towards a fully associative architecture”2 B.Cache confirms this: “the design process itself would need to be automatized and we cannot have a piece of software written for each type of design problem”. Instances of an idea; J.Holland too defines evolutionary computing as the mechanization of the scientific method, appropriated to architecture the mechanization of the conceptual method.