Methodology for Multicriteria Decision Making in Highway Asset Management

Abstract

A methodology is proposed for multicriteria decision making involving trade-off analyses between candidate projects as well as project selection and programming in highway asset management under certainty, risk, and uncertainty. A set of system goals in highway asset management structure was first identified, and relative weights of the system goals were determined. Performance indicators under each goal were identified. Benefits achieved under various system goals as a result of implementing a project are typically measured with noncommensurable units; they need to be converted into nondimensional units so that trade-offs between projects can be measured under equal parameters. Where such conversion processes involve certainty and risk, this paper develops systemwide multiattribute utility functions for individual asset management programs to form the basis of trade-off analyses. Because of the limitation of utility theory for situations under uncertainty, an alternative approach based on Shackle's model was introduced, and corresponding functions for project trade-offs were calibrated. The two sets of functions were collectively termed "systemwide trade-off functions" and used to deal with cases under certainty, risk, and uncertainty. A case study was conducted for systemwide highway project selection using information on candidate projects in past state highway programming in Indiana. Case study results were further compared with actual highway programming practice in Indiana to validate the proposed methodology. High matching rates of at least 85% were achieved for decision making under certainty, risk, and uncertainty. The approach could be adopted by other state transportation agencies for asset management practice.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 76 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.