Automating the conceptual design of residental areas using visual and generative programming

Abstract

The design processes based on parametric design and generative design allows multiple alternatives to be generated, analysed and corrected. The aim of the developed program was to automate the urban design concept development process for selected land units with a local development plan. This document, ordinances and other sources were the guidelines for the program to be developed. The solution-generating program starts by taking in the data, and then geometrical objects are created that have the right relationships to each other. Finally, results are evaluated on the basis of calculated indices and further generations are created taking into account the best options. This process must be overseen by a human, but with the development of machine learning and artificial intelligence, automation will progress, reducing human effort. The results of the study were collated and compared with other experiments in an in-depth literature review.