This talk will present the recent progress made at Inria on anisotropic mesh adaptation for RANS simulations for aerospace simulations. We will focus on drag prediction, high-lift prediction and turbomachinery applications. I will first present the considered flow solver and the adaptive meshing strategy. Then, I will present several error estimates for RANS simulation (feature-based, goal-oriented) that will be compared. I will describe how in practice the mesh adaptation loop is performed and the benefits of doing it like this (multigrid-like effect, convergence study for free, ...). In the presented exemples, I will show the early capturing (convergence) of the mesh-adaptive solution process and that we obtain mesh-independent results without any human intervention ie automation and whatever the initial mesh prescription.