Predictive Group Maintenance Model for Networks of Bridges

Recent progress in the monitoring and prediction of the condition of infrastructure using sensing technologies has motivated researchers and infrastructure owners to explore the benefits of asset predictive maintenance, as an alternative to reactive maintenance. However, the application of predictive group maintenance for multi-system multi-component networks (MSMCN) has not received much attention in the literature or in practice. The paper presents an approach that prioritizes the maintenance of MSMCN of bridges, using a deterioration model of components with uncertainty, a lifecycle cost model, a predictive model for the optimal time for maintenance based on the latest inspection, a group maintenance model to reduce setup cost, and a scheduling model considering budget constraints. This model has been applied to a network of 15 bridges constituted by multiple heterogeneous components, and, compared with the Structures Investment Toolkit, it showed potential for a substantial decrease in maintenance costs, thus highlighting the practical significance of the presented approach.