With ever-increasing demands on building performance, the early stage of architectural design becomes an even more significant factor for harnessing a potential building's performance. As decisions done during the early stages have the most impact on the entire project, architects and engineers must consider every possible design choice. Generative Design can help achieve this by automatically generating a large amount of designs, for example with the help of Generative Adversarial Neural Networks. One of the important aspects to consider during the early stages is regulation-compliance. Following guidelines from the beginning eases the design process by eliminating the need for adjustments later on. This thesis focuses on building a script with which regulation-compliant models of the U9-station in Munich can be generated using Autodesk Revit and Dynamo, with regulations originating from official DB (German Railways) guidelines. Furthermore, the possibility and success of translating regulations into code scripts are analyzed statistically and the state of the art of generative design in the early stages is presented.     «

With ever-increasing demands on building performance, the early stage of architectural design becomes an even more significant factor for harnessing a potential building's performance. As decisions done during the early stages have the most impact on the entire project, architects and engineers must consider every possible design choice. Generative Design can help achieve this by automatically generating a large amount of designs, for example with the help of Generative Adversarial Neural Networ...     »