The mesh adaptive direct search (MADS) algorithm is designed to solve blackbox optimization problems for which the structure of the involved functions is unknown. This presentation focuses on recent applications for which MADS was applied via the NOMAD software. These problems include the optimal positioning of snow monitoring devices over a large territory, the calibration of hydrological models in a climate change context, the optimization of alloys with the FactSage thermodynamic database, and finally biobjective optimization of aircraft trajectories.