In an industrial group like Safran, numerical simulation of physics phenomena is used in the majority of the design processes of its manufactured products. In the corporate research center of Safran, we develop technologies for improving these processes by deriving fast and reliable surrogates for various physics. We present some of the technologies developed in the recent years. First, a physical reduced-order modeling method for non-linear structural mechanics, applied to the lifetime computation of high-pressure turbine blades [1, 2]. Second, the learning of physics simulations under non-parameterized geometrical variability using classical machine learning tools thanks to non-linear deterministic dimensionality reduction (involving morphing and principal component analysis) and Gaussian process regression, with predictive uncertainties [3]. Recent improvement include automatic morphing and online efficient strategies [4]. Third, a physical reduced-order modeling method solving the incompressible Navier-Stokes equations on a manifold learned by a variational auto-encoder [5], with an extension to consider non-parameterized geometrical variability [6]. We also present our contributions to the open-source and open-data community. REFERENCES [1] F. Casenave, N. Akkari, F. Bordeu, C. Rey and D. Ryckelynck, A Nonintrusive Distributed Reduced Order Modeling Framework for nonlinear structural mechanics – application to elastoviscoplastic computations, Int. J. Numer. Meth. Eng. 121, 32–53 (2020) [2] T. Daniel, F. Casenave, N. Akkari, D. Ryckelynck and C. Rey, Uncertainty quantification for industrial numerical simulation using dictionaries of reduced order models, Mech. Ind., 23, (2022) [3] F. Casenave, B. Staber and X. Roynard, MMGP: a Mesh Morphing Gaussian Process-based machine learning method for regression of physical problems under nonparametrized geometrical variability, NeurIPS, (2023) [4] A. Kabalan, F. Casenave, F. Bordeu, V. Ehrlacher and A. Ern, Elasticity-based mesh morphing technique and application to reduced-order modeling, submitted, arXiv:2407.02433 [5] N. Akkari, F. Casenave, E. Hachem and D. Ryckelynck, A Bayesian non-linear reduced order modeling using Variational AutoEncoders, Fluids, 7(10), 334 (2022) [6] N. Akkari, F. Casenave, A consistent non-linear and geometrical reduced order modeling approach with uncertainty quantification applied to the convection dominated incompressible Navier-Stokes equations, WCCM2024, Vancouver, Canada