Hybrid inspection-monitoring approach for optimal maintenance planning

Abstract

Optimal planning of Inspection, Maintenance and Rehabilitation (IM&R) of civil engineering structures has been the focus of specialised literature during the last decades. More recently, permanent structural health monitoring started to gain increasing attention. In this article, a methodology is proposed to optimally combine conventional inspection techniques and permanent monitoring. The proposed methodology integrates dynamic Bayesian update of the belief state of the structure, based on the sensor readings, in a decision analysis framework. Uncertainties resulting from the model, the measurements, the imperfect inspections and the imperfect maintenance actions are taken explicitly into account. A numerical application of the proposed methodology, for a steel offshore structure, is presented. A sensitivity analysis of the optimal cost to relevant parameters, such as sensor configurations and inspection result uncertainties, highlights the potential benefits of the proposed approach.