Up to now two kinds of graphene membranes for gas filtration applications have been made: very selective membrane by adding graphene oxide on porous material or membrane with an excellent permeance due to the atomic thickness of CVD grown graphene decorated with nanometer size pore holes. GATES project aims at making the laboratory demonstration at a TRL level 4 of a new membrane made of CVD graphene that will be both highly selective (103) between H2 (He) and CO2, N2, O2, Ar gases and with a high permeance (10-5 mol m-2 s-1 Pa-1 ). The project concept lies on technological innovations on the device and pore holes realization. The developed processes will be compatible with a CMOS technology to insure manufacturability. The device fabrication will escape one of the roadblock of graphene technology, which is its transfer step after the growth of the Single Layer Graphene (SLG). This step is hard to industrialize, and induces many macroscopic defects that can degrade largely the yield of membrane production. The ambition of GATES is to develop and demonstrate a technology without transfer, CMOS compatible. Ultimately, two devices will be stacked to increases the selectivity by adding tortuosity. To achieve a very high selectivity, it is mandatory to have sub nanometer pore diameter. In GATES, initial pyridinic-N or pyrrolic-N vacancies on the SLG will be created by nitrogen doping using different plasma technologies. If needed, these vacancies will be further etched. An alternative approach, to be evaluated in GATES, is to make nano-pores by using heavy ions produced at GANIL (Grand Accélérateur National d'Ions Lourds) facility. In both cases, the shapes and pores size statistics will be evaluated by advanced TEM observations performed in the consortium. These statistics will be used as input parameters for the modelling of the membrane performances. Furthermore, other complementary characterization techniques, as Raman spectrometry and X-ray photoelectron spectroscopy (XPS), will be employed for having deep knowledge of the produced membranes. In the project the filtering properties of undoped, N-doped and functionalized nano-porous graphene will be modelled. Ultimately, double-layered functionalized graphene will be also modelled. This modelling work will support the experimental developments and will be used to define the optimum structure for the best permeability/selectivity ratio. GATES will also assess the porosity due to grain boundaries by simulating different local environments, e.g. pentagons-heptagons. The development of such a graphene membrane and its application to leak detection, will offer a significant innovative product for users by providing a highly portable/integrated device with a very short response time.