The marine industry is facing many technical challenges; from reducing emissions to electrification to improving fleet management and operations. These are complex issues which translate to complex engineering tasks. In many cases these problems can best be addressed through modeling and simulation. Here we will examine the approach to developing digital twins for the marine industry to better understand overall vessel performance and the contribution of each subsystem. We will also explore how to best use these digital twins for design, manufacturing, and operation. The use of system simulation for developing vessel architectures rapidly will be discussed, including the different sub-systems which can be modeled and the behaviours which can be simulated. Vessel sub-systems such as propulsion, electrical network and climate control will be examined. In addition, the inclusion of realistic sea and weather conditions are critical to creating a model which is representative of the realistic operation of the vessel. Furthermore, the technology for generative design of vessel architectures will be examined. This will leverage the existing tools for generating rapid vessel architecture system models as well as the integration of data from numerous sources (operational data from the field, simulation data from other sources etc.) via the use of Reduced Order Models (ROMs). Usage of the models will also be covered, including the types of engineering tasks which can be addressed with digital twins. Conducting trade off analyses between these vessel models can allow engineers to gain insight into the benefits and drawbacks between architectures. This is especially useful in assessing electrification and what the potential effects are on multiple fronts such as performance, operational expenditures and capital expenditures. Finally, a summary of the main areas of vessel design which stand to benefit from the digital twins will be given.