Design Methodology integrating artificial intelligence in the schematic design phase.

Abstract

Artificial intelligence (AI) has arrived and established itself as one of the most influential technological tools of recent years, with applications in diverse fields of the exact and social sciences, as well as in design and creation. This article proposes a design methodology that integrates AI during the creation process, specifically in the schematic design phase. The development of housing projects within the Colombian context is used as a case study. Based on an analysis of various theoretical and methodological frameworks for design and AI—including fundamental AI concepts, the difference between weak and strong AI, machine learning, deep learning, and generative AI—the potential of AI applied to the schematic design phase is also explored. Furthermore, several design methodologies are examined, such as Design Thinking, Double Diamond, the Bruno Munari method, and human-centered design. These methodologies are then compared with architectural variables (functional, formal, contextual, user-related, and technical).

Finally, a hybrid design methodology is obtained, where AI is incorporated as a support tool to enhance creativity in architecture, improving execution times and technical validation of projects, without replacing the fundamental role of the architect.