Files
Abstract
This dissertation investigates element-level inspection data available in the National Bridge Inventory and proposes a novel coactive prioritization model for bridge asset management. The model accounts for time-dependent element interactions, referred to as coactiveness, in predicting bridge performance resulting from preventive maintenance, rehabilitation, or replacement (MRR) activities. The proposed coactive model hypothesizes that if one repairs one element, it should reduce the deterioration of other elements. Those improved elements, in turn, reduce the deterioration of the repaired element and so forth. Therefore, this study aims to enable data-driven time-dependent element interactions for MRR decision-making. The proposed model is used to analyze Georgia’s bridges at first. It is concluded that accounting for element interactions that are present in the element-data yields more realistic, and thus less overly conservative, performance predictions. The results also indicate that the overall Bridge Health Index (BHI) improves by 20% over the subsequent 20 years when expansion joints are repaired utilizing the coactive prioritization mechanism. In a subsequent study, it is concluded that coactive relationships exist among elements in the Alabama and Florida bridge inventories. In Alabama, MRR on bridge deck elements are more influential than MRR on the expansion joint for the long-term bridge performance. It is concluded from this study that early preventive maintenance implemented in Florida most leverages the coactive mechanism. However, most states that do not have as much resources as Florida for early maintenance should benefit from the coactive model. Therefore, three additional bridge inventories of Virginia, Pennsylvania, and New York are investigated to study the effectiveness of employing the coactive model. They are known to have an excellent bridge preservation program. In this last study, both state-owned and NHS bridges are investigated. A game theory model is applied to decision-making, and payoffs of two major players, the Federal Highway Administration and a state agency are evaluated. The analysis confirms that long-term bridge performance predictions leveraging a coactive mechanism are effective in prioritizing elements for MRR decisions.