Expert judgement based maintenance decision support method for structures with a long service-life

Abstract

The optimal moment at which maintenance activities should be performed on structures with long service-life to guarantee the required quality of service is hard to define, due to uncertainties in their deterioration processes. Most of the developed methods and concepts use historical data to predict the deterioration process with deterministic values as a result. Some researchers recognise that probabilistic deterioration models are required for life-cycle models but in practice, however, historical data are often scarce. Moreover, the available data often only inform about a short period of time, while maintenance strategies, technologies, materials and external circumstances change over time. Therefore, the required probabilistic deterioration models cannot be retrieved and remain unproven in life-cycle modelling so far. Hence, this article introduces an expert judgement based Condition Over Time Assessment method that quantifies the uncertainty regarding the period that is required for structural assets to deteriorate to a given condition. The proposed method utilises Cooke’s classical model, which makes use of knowledge and experience of experts, who are weighed according to their performance in judging uncertainty, to assess this period. A bridge-based experiment shows that the proposed method has the potential to provide a means to effectively plan maintenance.

Keywords:

• Deterioration modelling
• engineering asset management
• expert judgement
• uncertainty quantification
• maintenance interval