Traditional and Modern Architectural Design Generation for Kıyıköy in Türkiye Utilizing the DALL-E Artificial Intelligence Model

This study utilizes the DALL-E artificial-intelligence (AI) model to generate both traditional and modern architectural visuals that follow the Kıyıköy Conservation Master Plan for the historic coastal town of Kıyıköy, Türkiye. The process has two stages. First, key features of traditional architecture—such as morphology, materials, and design types—are extracted from the plan and used as text prompts; DALL-E then creates visual representations of heritage structures. Next, the prompts are adjusted to incorporate contemporary design goals while respecting constraints on height, façade materials, setbacks, roof shapes, and window-door ratios. A five-point compliance framework, based directly on the conservation plan, is used to evaluate each output: (1) building height and number of stories, (2) lot width and distance to neighbors, (3) roof type, (4) façade materials, and (5) window-door ratios and openness. Figure 4b meets all standards, while Figures 4a, 4c, and 4d meet four criteria, falling short only in fenestration proportions. The results demonstrate that DALL-E, guided by regulation-aware prompts, can quickly generate concepts that adhere to conservation rules, emphasizing its usefulness for heritage-sensitive, rule-based design projects. Incorporating zoning and morphological constraints directly into prompts provides a new, reproducible approach for integrating generative AI into preservation-focused architectural workflows.