A study on the comparison and utilization of generative design methodologies : Focusing on parametric, topological optimization, and generative AI approaches

In recent years, generative design methodologies utilizing artificial intelligence (AI) have gained significant attention in the design field. Integrating topology optimization, parametric design, and generative AI design has become a major research focus in practical applications. This study analyzes these generative design methodologies and examines their roles and limitations in the design process. Furthermore, a new framework, PAX, is proposed by merging the PERT system with the AI Experience (AX) design methodology, providing a systematic approach to AI-driven design processes. The research methodology includes literature review and case studies to examine how each generative design methodology is applied in real-world design workflows. Based on this analysis, their effectiveness and complementary utilization strategies were derived. The findings indicate that topology optimization excels in weight reduction and structural stability, while parametric design is effective in iterative and adaptive modifications. Additionally, generative AI design plays a crucial role in design exploration and creativity but requires improved controllability and reliability. The proposed PAX framework aims to effectively integrate these generative design methodologies, ensuring optimal utilization of AI within the design process.