Automated compositions: artificial intelligence aided conceptual design explorations in architecture

Autor: Milošević, Jelena, Đukanović, Ljiljana, Živković, Milijana, Žujović, Maša, Gavrilović, Marko
Přispěvatelé: Bajšanski, Ivana, Jovanović, Marko
Jazyk: angličtina
Rok vydání: 2023
Předmět:
Zdroj: Proceedings: Geometry, graphics and design in the digital age, The 9th International Scientific Conference on Geometry and Graphics MoNGeometrija 2023
Popis: The paper focuses on the challenges of the relationship between architecture and artificial intelligence (AI), in particular, the potential of this technology to support architects' creative design processes in the form of augmented intelligence. Conceptual architectural design is an intricate process that produces new concepts by using prior knowledge, experience, intuition, and creativity. Artificial intelligence should not be used during the conceptual design stage with the goal of solving a problem in a predefined search space. Instead, potential solutions and design requirements should be explored using this technology. The suggested approach avoids preconceived solutions and psychological inertia attributed to the designer's finite experience. The paper gives a brief critical review of the application of AI in an architectural context, especially in conceptual design. The review emphasizes changes in the design process brought about by cutting-edge strategies, methods, and tools based on AI. Further, research by design is presented to illustrate the possible use of biologically inspired algorithms in producing innovative design proposals. The applied strategy in architectural design production was based on the human-machine interface and interaction (HMI&I). Symbiotic human-machine interaction in the design process facilitates the emergence of automated compositions that display novel and unexpected forms, detail, materiality, structure, functionality, and aesthetics. This study allowed us to explore peculiar architectural design methodologies and the potential of digitally intelligent architecture.
Databáze: OpenAIRE