Expected utility theory for monitoring-based decision making

Cappello, Carlo and Zonta, Daniele and Glisic, Branko (2016) Expected utility theory for monitoring-based decision making. Proceedings of the IEEE, 104 (8). pp. 1647-1661. ISSN 0018-9219 (https://doi.org/10.1109/JPROC.2015.2511540)

[thumbnail of Cappello-etal-PIEEE-2016-Expected-utility-theory-for-monitoring-based]
Preview
Text. Filename: Cappello_etal_PIEEE_2016_Expected_utility_theory_for_monitoring_based.pdf
Accepted Author Manuscript

Download (683kB)| Preview

Abstract

The main purpose of structural health monitoring (SHM) is to obtain information about the state of a structure, in order to guide bridge management decisions. Nevertheless, in practice, once a rigorous estimate of the structural state is available, decisions are usually made based on the decision maker’s intuition or experience. In this paper, we present the implementation of expected utility theory (EUT) in those civil engineering decision problems in which decision makers have to act based on the output of SHM. EUT is an analytical quantitative framework that allows the identification of the financially most convenient decisions, based on the possible outcomes of each action and on the probabilities of each structural state occurring. The advantage of the presented implementation is the optimization of decision strategies in SHM. In the manuscript, we first formalize the solution of single-stage decision processes, in which the decision maker has to take only one action. Then, we formalize the solution of multi-stage decision processes, in which multiple actions may be taken over time. Finally, using an example based on a case study, we describe the variables involved in the analysis of SHM decision problems, discuss the possible results and address the issues that may arise in the application of EUT in real-life settings.