Hotspot and Sampling Analysis for Effective Maintenance and Performance Monitoring

In this project, the authors propose two sampling methods addressing “how much and where” the agencies need to collect infrastraucture condition data for accurate Level-of-Maintenance (LOM) estimation in maintenance network with single type or multiple types of infrastructures. The method for single type infrastructure integrates Fisher information with spatial sampling technique that can be customized based on local agencies’ requirements, such as station-balanced, spatially-balanced, or others. For infrastructure condition inspection in a network with multiple types of infrastructures, a high-dimensional clustering-based sampling method is proposed. The method is based on the fact that inspection activities are carried out on the roadway segment basis, and selects sample segments that contain multiple types of infrastructures for the accurate estimation of their respective LOMs. The sampling process consists of two components: current condition estimation and high-dimensional cluster analysis. The methods are implemented using the infrastructure inspection records in the State of Utah from September, 2014 to March 2016. The sampling results indicate that both methods outperform simple random sampling method which is widely used across agencies.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Edition: Final Report
  • Features: Appendices; Figures; References; Tables;
  • Pagination: 54p

Subject/Index Terms

Filing Info

  • Accession Number: 01641126
  • Record Type: Publication
  • Report/Paper Numbers: UT- 17.12
  • Contract Numbers: 16-8506
  • Files: NTL, TRIS, ATRI, USDOT, STATEDOT
  • Created Date: Jun 21 2017 11:51AM