Underground resources (groundwater, heat, oil, or minerals) are difficult to access. The estimation of reserves and the design of optimal exploitation strategies rely on incomplete and indirect underground structural data (among others drill holes, conceptual geological models) and their numerical virtual representations. The CHYN is at the forefront of the research in this field. It develops CAD methods to convert field data into 3D volumetric numerical representations of the underground from the reservoir (kilometres) to the sample (micrometres) scale. Such algorithms include interactive 3D modelling techniques and optimization tools to design the best possible exploitation method for each case (e.g. optimal pumping well placement). Big Data analysis is used to analyse large data sets such as hyperspectral images of geological outcrops. Such 3D numerical models could be printed in 3D either for communication purposes or to conduct physical experiments. The algorithms behind these 3D models can be used for a variety of problems linked to additive manufacturing in which the creation of 3D numerical models is an issue.