Model Order Reduction (MOR) techniques are interpolation methods exploiting existing data sets (input and output) delivered from an existing model or experimental setup. This paper presents the major idea behind reduction methods along with potential applications for crash and safety simulations.