Impact of prior perception on bridge health diagnosis

We use Bayesian logic in reproducing how a rational agent, called Ernest in the paper, analyses monitoring data and infers structural condition. The case study is Adige Bridge, a 260 m-long statically indeterminate structure with a deck supported by 12 stay cables. Bridge structural redundancy, possible relaxation losses and an as-built condition differing from design suggest that long-term load redistribution between cables can be expected. Therefore, the bridge owner installed a monitoring system, including fiberoptic sensors that allow measurement of deformation with an accuracy of a few microstrains. After 1 year of system operation, which included maintenance of the interrogation unit, the data analysis showed an apparent contraction of the cable lengths. This result is in contrast with the expected behavior. We analyze how a rational agent analyzes the observed response, and, in particular, we discuss to what extent he is prone to accept the sensor response as a result of the real mechanical behavior of the bridge versus a mere malfunction of the interrogation unit. In this analysis, we consider four psychological profiles, which vary based on their personal trust in the reliability of the instrumentation and on their knowledge of the structural behavior of the bridge. Using Bayesian logic as a tool to combine prior belief with sensor data, we explore how the extent of prior knowledge can alter the final engineering perception of the current state of the bridge and we demonstrate how the engineer’s posterior judgment is predictable with a mathematical model. Formal reproduction of the human decision-making process can have strong impact in the field of structural health monitoring, as it may enable: (1) quantification of probabilities that engineers attribute to various events based on their subjective experience (which is currently an important challenge); (2) better understanding and improvement of the decision-making process itself; (3) embedding of decision making into structural health-monitoring methods for the full benefit of the latter.