The future will be more and more CAE. We agree. It is therefore strategic to rely on a robust scientific research know-how, as years and years of university laboratory can provide. Proven ability to develop methods for validating numerical models is needed, as well as the availability of adequate computing power. We are on all these aspects.