In the era of digital twins, computational software plays a crucial role in enabling more efficient projects that complement physical testing. This study, motivated by the development of a digital model aimed at diagnosing failures in roller bearings of transmission systems, proposes the application of the Verification and Validation (V&V) process, following ASME standards, to ensure that the computational model accurately represents the real behavior of these components. The procedure involves two main steps: verification and validation. Verification is subdivided into model verification, which compares the code outputs with analytical theoretical models to confirm consistency with established theory, and calculation verification, which assesses the numerical accuracy through mesh refinement analysis. Validation involves comparing the computational results with experimental data to determine whether the model produces values consistent with reality. Only when both verification and validation yield consistent outcomes can the model be considered reliable for practical use. A progressive approach is adopted for model verification, starting with a simple theoretical model of roller bearings based on Hertzian contact theory under static equilibrium and dry contact, evolving to an elastohydrodynamic (EHD) lubrication model, and finally advancing to a multi-level computational model. Experimental tests are conducted to confirm practical validation, ensuring that the implemented digital twin is both reliable and representative of real-world applications.