Harnessing Publicly Available Information with Data Science to Extract the Operational Profile of a Vessel
The optimal design of an efficient and cost-effective vessel requires extensive knowledge about its intended operation. However, this information is not always available or accessible for a ship designer/yard. This often results in a vessel which is less well suited for the job than it should be. The vessel is over specified for the required task and as a result more expensive than it could be for the client to buy and operate. A platform was developed which can extract the entire operational profile of a trailing suction hopper dredger with as little information as possible. The information used consists of publicly available data, such as that of the automatic identification system, weather information and sea charts. The platform uses machine learning algorithms to determine the vessel task, time spent in the task and the vessel uptime. Combining these results with additional knowledge of dredgers and their drive systems allows for an estimation of both the dredger production and the power and fuel consumption. The paper discusses the methods used in the platform to extract the operational profile from the publicly available data and how this results in a power and fuel consumption estimation. The results of the platform will be validated with information available from two trailing suction hopper dredgers.