Trade Space Wonders: Expanding Horizons with Graph Embeddings to understand large trade space from generative methods.

trade space wonders;generative methods;graph embeddings

In systems engineering, automated trade space generation surpasses manual methods, offering efficient and innovative solutions. But it also poses challenges in analyzing the vast trade space generated. In this work, we study three systems engineering cases in mission, system, and subsystem design across automotive and aerospace industries. We demonstrate how graph embeddings created by unsupervised machine learning can capture structural information, similar to human understanding of systems.