SAFEPEC - Innovative risk-based tools for ship safety inspection

SAFEPEC, the EU-financed research project, aims to make ship inspections more efficient with probability-based risk models and sensor-based inspection tools. With new technologies and risk models, the consortium creates innovative ready-to-use techniques making ship inspections on safety more reliable – and cheaper. The project promotes proactive safety and develops a unified risk-based framework built upon the analysis of historical data of casualties, near miss cases, deficiencies and non-conformities that are detected by various types of inspections. Based on this framework a software prototype is developed that enables the interoperability and coherent interpretation of those data sources; and can contribute to the early detection of failure, either in the ship structure or its equipment.

We, TUM, particularly work on two workpackages (WPs): WP4-Risk based inspection framework and WP5-Unified risk management models. The risk models are categorized into the 4 modules, as illustrated in the Figure below: Cause, Vulnerability, Consequence and Inspection, which are interlinked through the overall risk framework. Bayesian Network (BN) method is selected as a strategy of modeling the risk modules, not only for enabling Bayesian inference but also for the efficient representation of the random variables with considering their causal relations. The risk-based inspection framework is applied to selecting ships from a fleet for inspection and specifying the scope of an inspection for a particular object and deterioration mechanism.