Generative AI-driven design optimization: eight key application scenarios

The traditional surrogate-based design optimization is reaching a new turning point with the development of generative AI. In order to effectively harness generative AI-driven optimal design, a clear understanding of its underlying concepts is essential, and for this, key application scenarios for generative AI-driven design optimization are defined and presented in this paper. Additionally, to enhance intuitive comprehension, each scenario is illustrated with representative visual examples of the inputs and outputs of AI. Design and analysis data are classified into 1D, 2D, and 3D, while AI models can be divided into generative, predictive, and optimization models. By combining these data types and models, we have identified eight representative scenarios. This study explores how generative AI-driven approaches can overcome the limitations of traditional design optimization and examines their practical applicability through the proposed key scenarios.