01137527

Component

Despite decades of pavement and bridge management, improving the condition of transportation infrastructure in the United States remains a challenge. Agent-based modeling, a simulation technique that has been used broadly in the social sciences as well as in some aspects of transportation, has the potential to provide insight into how decisions at various levels impact infrastructure performance. This paper describes a series of models developed and experiments performed to explore ABM as a tool for improving infrastructure management. A prototype model and a second-generation model were developed in Repast, a freely available agent-based modeling tool. Experiments using the models employed both synthetic and real-world data to generate time-series of pavement condition, user costs, and other quantities of interest. While some of the simpler experiments generated the anticipated results, the results of others were more difficult to comprehend, suggesting evidence of emergent behavior at the system level. This also suggests the need for more rigorous model validation in preparation for expanding the types of experiments to include a broader array of policy options.

01120148

09-2681

English

15p

2009

Transportation Research Board 88th Annual Meeting

DVD

Figures (2) ; References (17) ; Tables (2)

Asset management; Behavior; Computer models; Government funding; Object oriented programming; Pavement management systems; Pavement performance; Simulation

Pavement Condition Index

Data and Information Technology; Highways; Pavements; I10: Economics and Administration

Transportation Research Board Annual Meeting 2009 Paper #09-2681

BTRIS, TRIS, TRB

Jan 30 2009 7:02PM