The project goal is to implement an AI-based Energy Management System (EMS) for a hybrid propulsion system in a commercial vessel. The EMS is to operate the propulsion system optimally with respect to fuel consumption. First, using the Dynamic Programming (DP) algorithm, the optimal operating strategy is calculated for a given driving route. Then, in the second step, three causal AI-based approaches are developed and, in the third step, compared with the solution of the DP approach and also with each other. The comparison among each other serves to compare the feasibility of the approaches, considering practically relevant aspects such as the required computing power. The comparison with the DP solution is used to benchmark the AI approaches to ensure that the resulting EMS effectively operate the drive as fuel-optimally as possible. The best EMS will be implemented on a commercial vessel as part of the project.