The Multiscale Heat & Fluid Flow Lab (MFL) developed an AI model to evaluate the influence of changes in design factors affecting aerodynamics by using wake flow. The accurate assessment of how design changes influence aerodynamic performance is a critical challenge in the automotive design process. However, results of the traditional assessment vary depending on the experience of the evaluators performing this process when multiple designs change simultaneously. Specifically, MFL trained ResNet18 model with wake flow and its corresponding design factors using two different approaches – one is Multi-label Classification to identify whether design elements changed, enhancing interpretability through grad-CAM visualization and the other is Multi-target Regression model to quantitatively measure the extent of design element changes. As a result, both models effectively analyzed design elements and could comprehend their impact on aerodynamics. This research is being conducted in collaboration with HYUNDAI NGV.