Central aspects of the design include both the distribution of material to optimize mechanical stress resistance and the selection of suitable materials. In practice, these two tasks are often carried out sequentially. However, this approach can not only lead to longer development cycles but also promote suboptimal solutions, as early decisions may limit the design flexibility.

Considering material selection and distribution simultaneously as a unified optimization problem opens up new possibilities in product development. The design space remains flexible for a longer period, allowing for the exploration of more efficient and high-performance design variants. A key requirement for this is the ability to add or reduce material automatically and variably, without compromising the structural integrity of the design. Deep learning-based approaches have already demonstrated their ability to effectively assess mechanical stresses within components.