The proposal tackles the use of AI applied to urban and architectural design. More specifically we put emphasis on the implementation of a genetic algorithm interlinked with environmental parameters, in order to explore, characterize and promote generative design tools and methods. We consider this work from a sustainable standpoint and we expect that advanced digital tooling could contribute to design more efficient cities. We lean our experiment on the local context, with the objectives to build a parametric model and a genetic evaluation engine based on local energetic priorities and policy. Moreover, we enhance the proposal with the contribution of a local architectural office. We would like to measure the appropriation degree of the generative and computational design and we want to identify the interests, limits and brakes of these tools in order to challenge further developments in the field.
The digital turn goes along with deep transformation in the field of architecture and urban design. It could support the emergence of new solutions in the climatic urgency context, allowing more efficient building, more sustainable architectural and urban morphologies, impacting the creative, evaluation and analysis processes. It involves new environmental opportunities and if numerous researches have been done in the field of digital architectural design, we should note that advanced digital tooling doesn’t meet the interest of the practitioners and professional offices. This is especially true when we consider artificial intelligence, machine learning or genetic algorithms and broadly what we usually gather under the umbrella of the generative design. Algorithms and artificial intelligence could help us in high-performance design, by increasing the amount of information and contributing to adding new information to the design process. But they involve new skills, new methodologies and the cultural appropriation and acceptation by designers and engineers including professional architects, project owners and contractors.
First, we aim to implement a genetic algorithm to the urban and architectural design, through the use of parametric design tools and visual programming in order to conceive new building typologies. Our parametric model will include environmental performances and we extract these criteria from the Plan Air Climat defined by the Grenoble Alpes Métropole.
Second, we seek to evaluate the effectiveness of this generative design by confronting its uses to a real practice and context. We will evaluate the appropriation of these technics by the practitioners and thus contribute to promote these methods.
Philippe Marin : philippe.marin(at)grenoble.archi.fr
The project is supported by the Multidisciplinary Institue in Artificial Intelligence (MIAI)