Generative Design Software Development Process for Architecture

The existing design method of using traditional CAD (Computer Aided Design) software to explore architectural design alternatives is time-consuming and tedious. Generative Design ( hereinafter called the GD ) software is a potential solution that can streamline the design process and provide creative design alternatives for architects. However, current GD software is not widely accepted by architects because they are either difficult to use, or their generated results don't meet architect’s needs because the solutions it generates don’t always meet design criteria. This situation reflects a more fundamental reason that limits the promotion of GD software which the industry lacks a software development process that could guide developers to develop valuable GD software for architects.

Therefore, in this thesis, I propose a development process to fill the gap. The process consists of four major steps and it integrates the necessary architectural design knowledge and programming resources to follow and reference. To demonstrate the feasibility of the process, I developed a GD software for a basement parking layout design task. Based on the generative method of variational modeling with optimization, this tool can generate a variety of valuable

basement parking layouts in a very limited amount of time. According to the software test result, the generated solution reaches about 86.95% accuracy but reduces 36% modeling time when architects use it to reproduce an existing basement parking layout. The GD software developed under the guidance of the process proved to be valuable.