생성형 AI와 더블 다이아몬드 모델의 협업 구조를 활용한 디자인 프로세스 연구 -스마트 소파 디자인을 중심으로-

(Background and Purpose) As the aging population continues to grow rapidly, the importance of designing smart products tailored to the needs of senior users is increasingly emphasized. At the same time, the advancement of generative AI technologies has opened new possibilities for integrating AI into creative processes, including product design. In this context, this study proposes a collaborative design process based on the Double Diamond model by combining a text-based generative AI (ChatGPT) and an image-based generative AI (Midjourney). The objective is to examine how designers can effectively collaborate with generative AI systems, particularly in designing a smart sofa customized for active seniors, using the proposed process as a framework. (Method) The study first reviews the concepts and operating principles of ChatGPT and Midjourney, as well as the structure and function of prompts, which play a crucial role in generative AI systems. It also examines the structure and characteristics of each phase—Discover, Define, Develop, and Deliver—of the Double Diamond model, along with previous research on AI-assisted design processes. Based on these theoretical considerations, the study proposes an integrated design process combining generative AI with the Double Diamond model. This process was then applied to the design of a smart sofa for active seniors, with the roles of both the designer and generative AI analyzed in detail. (Results) Applying the proposed process, a smart sofa was designed. During the Discover stage, based on insights from prior studies, ChatGPT was considered a potential tool to assist in analyzing large-scale data and deriving design concepts and directions. In the Define stage, design concepts and directions were established, and persona-based scenarios and prompts were created using ChatGPT. In the Develop stage, Midjourney generated initial design drafts based on the prompts, which were refined through analysis and iterative prompt adjustments. The final design, selected in the Deliver stage, reflects both physical and emotional needs of seniors through its form, atmosphere, and color scheme, while incorporating core keywords used in the prompt. (Conclusions) This study suggests that generative AI can play a meaningful role in assisting designers, particularly in the Define and Develop phases of the design process. ChatGPT helped formulate clear, structured prompts, which supported the use of Midjourney and improved the connection between different AI tools. Midjourney, in turn, allowed faster exploration of visual ideas by generating multiple design options efficiently. Overall, the findings indicate that generative AI holds value not only as a practical tool but also as a potential creative partner in human-centered design. Further studies are needed to explore how this collaborative approach can be adapted to various design fields and how well it works across different product types and user needs.