Locomotives are a central resource for rail freight companies, and efficient and flexible locomotive control is therefore essential. Locomotive control is part of the freight operator's operational control together with the control of train drivers and wagons. Other aspects that are outside the freight operator's area of responsibility but play a major role in the operational activities are railtraffic management (handled by Trafikverket). Locomotive control has many similarities with personnel control and wagon control (in the sense that all these resources are dependent on traffic), and a deep reinforcement learning (DRL) agent who can support locomotive control should be able to form the basis for similar methods for these two other resource types as well. The goals of AI-based decision support include cost reduction and increased reliability through more efficient resource management and better adaptability to disruptions. The project contributes to the funding program Triple F's goals by improving the flexibility, efficiency and reliability of the railway, thereby facilitating the transfer of transport to the emission-efficient railway. It supports the transition to fossil-free alternatives and reduces CO2 emissions by increasing the competitiveness of rail for freight transport.