Multi-objective optimum life-cycle management and decision-making for deteriorating structures: review and future directions

Abstract

Civil structures deteriorate under continuous exposure to external loads, mechanical and environmental stressors, and extreme events throughout their service life. The high uncertainty associated with these factors complicates the accurate prediction of structural performance and remaining service life. To address these challenges, timely inspection and monitoring are essential. This paper reviews and integrates recent advances to present a comprehensive probabilistic framework for the life-cycle management of deteriorating civil structures based on multi-objective optimisation (MOO) and indicates future directions in this research field. To assess structural condition and performance, various indicators are described, including condition-based, safety-based, damage-tolerance-based, and cost-based indicators, as well as resilience and sustainability metrics. The effects of maintenance on time-dependent structural performance are investigated using performance profiles, lifetime functions, damage propagation, and event trees. These models provide the foundation for probabilistic MOO-based service life management. The formulations of various objectives for MOO-based service life management of individual structures and infrastructure networks are summarised. Methodologies for solving MOO problems and selecting appropriate management strategies are reviewed. The integration of structural health monitoring and Bayesian updating is highlighted to continuously refine performance predictions and management decisions. Finally, future directions for more efficient and reliable MOO-based service life management are discussed.