[News] JAS Menton 2024: Generative design and AI for the hospital of the future

 

For several years, Architecturestudio has been involved in generative design research, using digital tools based on artificial intelligence algorithms such as genetic optimization, agent-based models and machine learning to assist architects in the design phase.

After defending a doctoral thesis on eco-design, we are now applying these tools to the generation of functional 3D sketches for complex projects such as hospitals, whose layout represents a combinatorial optimization challenge. Although hospital design cannot be reduced to algorithmic solutions, a proximity matrix derived from a detailed technical program can generate 3D functional organization proposals, helping architects and enabling hospitals to visualize their space.

Generative AI can also be used to answer written queries through automatic machine learning based on data from the hospital project in question, using available tools such as Mistral AI. The method and constraints remain rigorous.

Presented by Laurent-Marc Fischer, Associate Architect at Architecturestudio, and Claire Duclos-Prévet, Architect DE and R&D Researcher at Architecturestudio, this conference aims to spark debate on the opportunities and limits of using these advanced digital tools in hospital design.