Currently there is growing interest in unmanned shipping. In the case of unmanned ‘autonomous’ shipping, navigation is automated by on-board decision-making systems. Important motives for unmanned ships include the shortage of skilled mariners, the facilitation of slow steaming strategies, efficiency improvements in confined shipping areas, and increased safety. The aim of the present research is to simulate an unmanned ship through an Automatic Identification System (AIS) based traffic situation. In order to do this, the Maritime Research Institute Netherlands (MARIN) will use the existing simulation technology Dolphin, implement a new tool to read AIS data, simulate a large amount of ships, and develop an auto-captain. The real-time dynamic risk index developed by MARIN will be integrated in Dolphin, to monitor nautical safety of all ships with focus on the unmanned ship. The simulated unmanned ship will navigate according to the International Regulations for Preventing Collisions at Sea (COLREGS). In more complex situations, the auto-captain may use a dedicated decision support tool to find a more efficient solution to pass safely. This approach will be analysed using the real-time dynamic risk index, which will be updated based on latest insights. This paper will discuss the latest development and plans in the unmanned ship simulation project.