Screening Cycle Decision Modeling for Infrastructure Management

Those responsible for the supervision of a collection, or group, of infrastructure-related structures often create programs to screen those structures for safety. These programs systematically move through the group, identifying some structures for further investigation and others as adequate. Because such screening programs take time and effort, there is a time lag for a structure between one screening and the next. This paper examines two major issues: the choice between available tests in a screening system, and the identification of structures as adequate. The effect of the time lag between screens is also examined. The proposed screening model assists the user in deciding when to move to another structure, allows the iteration of a present analysis on the present structure using different assumptions, and aids in deciding whether to accept a "safe" judgment for a structure. The latter, a recognition of the significant potential effects of even a small probability that an unsafe structure has been declared "safe," is analyzed in detail through sensitivity analysis. It is demonstrated that carefully estimated values for the variables underlying a screening model are critical when designing such a model. A design assumption that screening system tests should decrease in conservatism with increasing complexity and cost is challenged and found to be false in one case, thus requiring investigation when screening systems are designed.