Simulation has been a competitive differentiator for engineering-driven businesses, available at all stages of the development process and lifecycle, used by the various domains within an organization, not necessarily simulation experts. It requires discipline integration, scalability, reduced-order model, and democratization. The concept of digital transformation involves new approaches for data and lifecycle management, the understanding of the digital thread, digital twin, predictive and cognitive capabilities, including improvement of model complexity, integration of physics, increase of knowledge. These trends require bringing the physical and virtual worlds closer together and also the adoption of cyber-physical model at all stages of design, production, and operation. To overcome the drawback of simulation and the need to balance the computational effort with accuracy and efficiency, new modelization strategies are adopted with ML and AI, which use a combination of virtual and physical data for training ROM, with an order of magnitude faster than the multiphysics one.