The inspection of marine vessels is currently per formed manually. Inspectors use sensors (e.g. cameras, devices for non-destructive testing) to detect damaged areas, cracks, and corrosion in large cargo holds, tanks, and other parts of a ship. Due to the size and complex geometry of most ships, ship inspection is time-consuming and expensive. The EU-funded project MINOAS develops concepts for a Marine Inspection Robotic Assistant System to improve and to automate ship inspections. As one example of a physical system implementation, we introduce our magnetic wall-climbing robot. This remotely operated lightweight system is able to climb a vessels steel frame and is able to deliver visual inspection data on-line. For any type of surveying process, the raw and meta data are mandatory in order to compare inspection data over time. In this paper, we describe our approach of how the magnetic climbing robot is localized and controlled. Additionally, we describe the design of the robot as well as the localization device which is able to provide a real-time 3D pose of the inspection robot.