In conventional computational fluid dynamics, various mathematical models are used in conjunction with numerical methods. For example, solving non-linear partial differential equations requires a considerable amount of computing resources in order to achieve adequate results.
In other fields of application, methods from the field of machine learning are being used more and more. They are now partially or completely replacing conventional methods. This potential should now also be used in this new context. Skilful use of machine learning could, for example, open up an entirely new category of solvers for differential equations.