Autonomous ship control requires a robust and resilient delivery of necessary services. The reduction of manning and increase of technology onboard results in an increased demand of automating onboard platform systems. Platform systems can be represented as a graph of interconnected components. The components are modelled as transfer functions between demand and supply. A method is proposed where this graph is traversed using straightforward algorithms to find all configurations able to deliver the required service. The different configurations are scored using a utility function to select the best one. We demonstrate this approach is feasible. Limited effort is needed to implement even if the number of components in the platform system is high where traditional automation methods struggle to deliver.