The work carried out in this Ph.D. thesis is part of a broader effort to automate industrial simulation systems. In the aeronautics industry, and more especially within Airbus, the historical application of simulation is pilot training. There are also more recent uses in the design of systems, as well as in the integration of these systems. These latter applications require a very high degree of representativeness, where historically the most important factor has been the pilot’s feeling. Systems are now divided into several subsystems that are designed, implemented and validated independently, in order to maintain their control despite the increase in their com- plexity, and the reduction in time-to-market. Airbus already has expertise in the simulation of these subsystems, as well as their integration into a simulation. This expertise is empir- ical; simulation specialists use the previous integrations schedulings and adapt it to a new integration. This is a process that can sometimes be time-consuming and can introduce errors. The current trends in the industry are towards flexible production methods, integration of logistics tools for tracking, use of simulation tools in production, as well as resources optimization. Products are increasingly iterations of older, improved products, and tests and simulations are increasingly integrated into their life cycles. Working empirically in an industry that requires flexibility is a constraint, and nowadays it is essential to facilitate the modification of simulations. The problem is, therefore, to set up methods and tools allowing a priori to generate representative simulation schedules. In order to solve this problem, we have developed a method to describe the elements of a simulation, as well as how this simulation can be executed, and functions to generate schedules. Subsequently, we implemented a tool to automate the scheduling search, based on heuristics. Finally, we tested and verified our method and tools in academic and industrial case studies.