Visual Inspections and Key Performance Indicators: Bridging the Gap

The service quality of a bridge—measured by key performance indicators (KPIs)—is considered insufficient if they do not meet performance goals. Widely recognized KPIs are safety and serviceability. A bridge is regarded as structurally safe if the probability of failure during its service life is not expected to exceed some nominal value. Similar approach applies to serviceability. Whereas in the design phase the safety and serviceability concerns are addressed directly in quantitative manner, in the service phase, the condition state is determined based on inspection results, which is a qualitative performance indicator. The condition state is actually a vague measure for the deviation of inspected bridge from the “as new” condition. The direct assessment of safety and serviceability is regarded as not practicable. In this paper, the approach to determine KPIs based on inspection results is proposed. The KPIs are subject to observations obtained by visual examination and simple nondestructive testing. Typically, they are related to bridge components and indicate existing or expected bridge dysfunctionality that can result in insufficient KPIs. The challenge is to establish procedures that connect observations with KPIs. The papers suggest Bayesian networks to this end. The à priori values of KPIs, i.e., the ones of the intact bridge are assessed as a baseline for the assessment during the service life. The à posteriori assessment of KPIs based on inspection results is performed using Bayesian networks that model observations and its uncertainties. The proposed approach is illustrated in a simple example.


  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: pp 256-268
  • Monograph Title: Eleventh International Bridge and Structures Management Conference
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01672057
  • Record Type: Publication
  • Files: TRIS, TRB, ATRI
  • Created Date: Jun 14 2018 10:17AM