Towards meaningful computational descriptions of architectural form

Abstract

Computers have irrupted in the domain of architectural design, still with uncertain results. The impetus are driven by two major forces: First, the momentum originated by extensive application of computational tools to any imaginable human activity, not necessarily related to design. Second, the experience derived mainly from engineering fields, which has by-produced systems intended for computer-aided architectural design. Yet, architects complain. The magic is missing. The goals, mystified. Computerized design tools are product-oriented, not process-oriented. In Architecture, the how's are as significant as the what's. This thesis presents a computational environment for architectural design. The environment aims to overcome some critical limitations of current CAD systems. Architectural design is neither drafting, nor simulation, nor modelling. Architectural design is a process of progressive consciousness acquisition. The proposed envirorunent helps architects communicate better with their design objects. It is an environment where design worlds can be expressed and explored, where architects can manifest their preferences to approach each design problem, what instruments are to be used and how they are to be manipulated. It is intended to be a tool to design better, not directly a tool to produce better designs.