Real-time interaction of physical models and generative AI in the architectural design process (hybrid design studio experiments – HDS)

This study aims to experimentally examine the simultaneous real-time use of physical models and generative AI tools in architectural design. Distinguishing itself from prior sequential hybrid studies, this research introduces a real-time reciprocal loop between tactile and digital tools. The studio included two sequential design experiments: one with contextual constraints and one without. Methodologically, the study adopted a Design-Based Research framework, utilizing student surveys to evaluate the effectiveness of the hybrid approach. In addition, expert opinions were obtained to support pedagogical inferences and to evaluate student outcomes. Findings indicate that while physical model-making remains the dominant driver for design realization compared to prompt writing, both tools function supportively. Furthermore, most students reported that predefined contextual constraints were more productive. Ultimately, the study contributes to design pedagogy by demonstrating that real-time hybridization fosters reflection-in-action, while structured constraints act as essential catalysts for creative focus in AI-integrated workflows.