Developing key performance indicators for the assessment of existing bridges

Abstract

Highway authorities worldwide are responsible for the management of vast road networks with numerous bridges. While traffic numbers and loads continuously keep growing, the condition of the existing structures is increasingly deteriorating. In general, the authorities have only limited funds at their disposal to maintain the functionality of the crucial transport infrastructure. Consequently, they are faced with the difficult task of prioritizing investments in bridge repairs and replacements based on the limited information at their disposal. These decisions are commonly made based on the condition of the structure as observed for example in bridge inspections carried out in regular intervals. Meanwhile in some cases several additional performance indicators for bridges may have emerged, such as reassessment results, monitoring data, results of non-destructive testing etc. which may shed additional light onto a structureﾒs condition and influence its level of urgency regarding replacement or strengthening. While there have been investigations into the influence of monitoring or materials testing onto the reliability of a specific structure, these methods typically require extensive modelling and expert knowledge. Although these methods are effective at the level of a selected bridge, they cannot be applied on the transport infrastructure as a whole in order to facilitate the prioritization of investments. This paper investigates an approach where different indicators of structural performance are linked and weighted in order to generate a few key performance indicators which then may serve as the basis for the prioritization of funds. The initial definition of such KPIs may be carried out based on previous experience and expert assessment. The results of such a prioritization may then be compared to the actual choices made hitherto and will most likely show significant differences. While this may also indicate that some choices should be reinvestigated, it is more probable that the definition of the KPI needs to be adjusted. By this procedure, possibly with the use of algorithms based on artificial intelligence, the definition of the key performance indicator may be improved gradually with time. As a result, highway authorities may obtain a valuable tool, which on the long run could provide an increasingly accurate assistance in the optimal allocation of scarce funds.