Improving early OSV design robustness by applying 'Multivariate Big Data Analytics' on a ship's life cycle

Typically, only a smaller portion of the monitorable operational data (e.g. from sensors and environment) from Offshore Support Vessels (OSVs) are used at present. Operational data, in addition to equipment performance data, design and construction data, creates large volumes of data with high veracity and variety. In most cases, such data richness is not well understood as to how to utilize it better during design and operation. It is, very often, too time consuming and resource demanding to estimate the final operational performance of vessel concept design solution in early design by applying simulations and model tests. This paper argues that there is a significant potential to integrate ship lifecycle data from different phases of its operation in large data repository for deliberate aims and evaluations. It is disputed discretely in the paper, evaluating performance of real similar type vessels during early stages of the design process, helps substantially improving and fine-tuning the performance criterion of the next generations of vessel design solutions. Producing learning from such a ship lifecycle data repository to find useful patterns and relationships among design parameters and existing fleet real performance data, requires the implementation of modern data mining techniques, such as big data and clustering concepts, which are introduced and applied in this paper. The analytics model introduced suggests and reviews all relevant steps of data knowledge discovery, including pre-processing (integration, feature selection and cleaning), processing (data analyzing) and post processing (evaluating and validating results) in this context.