The general license application (RBG) for the Swiss deep geological repository will be submitted in 2024. Furthermore, the decommissioning of the nuclear power plants (NPPs) is coming soon, with the Mühleberg NPP (KKM) scheduled to shut down in 2019. In support of the associated analyses and planning, availability of accurate descriptions of the nuclear waste streams is of paramount importance. The work described in this thesis represents a significant improvement of Nagra's most important radioactive waste characterization methodologies: for spent fuel, NPP decommissioning waste, and reactor waste. The resulting high-fidelity nuclide inventories will decisively support the future waste disposal research and development, as well as serve the NPP decommissioning planning in general. The spent fuel characterization demonstrated is based on Polaris. This approach is compared against the code sequence used in the industry (Studsvik CMS), in order to confirm that, when the necessary fuel data is available, either one of these codes can be used to provide detailed nuclide vector for the spent fuel. Afterwards, possible ways to produce the desired nuclide vectors in the future are identified. For decommissioning waste, every step of the old characterization methodology has been further developed and improved upon. This includes a new, detailed, fully three-dimensional approach to NPP modeling and the development of a new activation sequence (based on the ORIGEN depletion solver), which outputs a highly-resolved component-wise activation distribution and visualization. Additionally, the methodology is extended to apply the method's results for algorithm-optimized packaging concepts and waste volume minimization through decay storage and free release analyses. At the end, recommendations for the future improvements are presented. The decommissioning waste characterization methodology is then expanded to accommodate reactor waste, and the more complex nature of the non-stationary components of this waste stream (such as the control rods). A systematic general approach, based additionally on the component dose rate measurements, allows for a more efficient and flexible characterization of all reactor waste. These developments serve as an important foundation in the ongoing transition from a conservative approach towards a best-estimate plus uncertainty (BEPU) waste characterization, which is vital for accurate safety analysis studies, as well as waste and cost minimization. The results produced by these methodologies will form the basis for the next version of MIRAM (the Swiss Model Inventory of Radioactive Materials), MIRAM2020, which is to be used for the RBG. They will also be used as part of the next Swiss NPP decommissioning cost study, KS2021. At the same time, the obtained results are already being provided to the Swiss NPPs planning their decommissioning, specifically KKM and Beznau (KKB), allowing detailed component-wise segmentation strategies and packaging concepts decision making, finally leading to notable cost reduction for the utilities.