Risk-based asset management for hierarchical dynamic multiobjective systems: Theory, methodology, and application

Asset management for highway systems is addressed in terms of multiobjective dynamic decision problems for hierarchical systems. The dissertation explores the problems of management and resource allocation across different highway assets, programs, and functional units. A theoretically-based asset management methodology has been developed in three parts. The first is a classification system for assets to determine priority according to urgency of needs. The second is a dynamic sequential decisionmaking process for analyzing individual highway assets. The third part consists of a hierarchical multiobjective aggregation procedure that is useful in coordinating the resource needs and requirements of various functional units. This integrated approach to asset management covers short- and long-range decisionmaking that is applicable for different asset types, and includes extreme-event analysis. The research characterizes asset management objectives and highway assets by identifying decision metrics and developing a classification scheme for the filtering and assessment of assets, while incorporating the potential for extreme events. The research develops the theoretical foundations for relating two sequential decisionmaking tools---multiobjective decision tree and multiobjective risk impact analysis method---to analyze hierarchical dynamic distributed systems. The multiobjective decision tree is graphical tool that is used to handle noncommensurate objectives while the multiobjective risk impact analysis method applies linear dynamics to model the system and is used in the analysis of risk and decision impacts in a dynamic framework. The viability of the methodology is demonstrated through a case study of pavement assets. The asset management approach developed in this research is sufficiently general that it can be applied to the analysis of large-scale dynamic systems that exhibit the same characteristics as highway systems.