Cenaero’s expertise in Multi-disciplinary Design Optimization (MDO) has been largely built and demonstrated on turbomachinery applications, whether in a collaborative framework of within direct support to industry. In surrogate-assisted optimization “useful accuracy” is the motto and tailored physics-based design methodologies and specifications (parameterization, objectives and constraints formulation, . . . ) need to be set up. Multi-level design loops are developed as a backbone to robustly automate all the steps of the multi-disciplinary evaluation process, including the mesh generation steps, the computational setups, the link with the HPC infrastructure, the computation monitoring steps, the co-/post-processing and the design steering capabilities. There shall be no weak link in the computational setup so that all its constituents are fit for efficient design space exploration. Multi-disciplinary design methodologies need to be tailored to the needs of the specific components and systems. The trend is towards further integration of the design cycles, with an increasingly holistic approach: more physics and disciplines, integrating materials and processes, from components to systems. Reduced Order Modeling and Machine Learning capabilities, as key assets to multi-scale modeling and design space exploration and analysis, constitute a strategic development axis. After an overview of the research themes and activities at Cenaero, selected aeronautical applications within collaborative projects will be presented and perspectives drawn.