Information impacts on route choice and learning behavior in a congested network

Xuan Lu, Song Gao, Eran Ben-Elia

Research output: Contribution to journalArticlepeer-review

43 Scopus citations

Abstract

Every traveler makes route choices in an uncertain environment that includes random disruptions to the traffic system such as incidents, bad weather, and random behavior of fellow travelers. The premise underlying the development of advanced traveler information systems-that better-informed travelers make better route choices-should be tested. This paper studies en route real-time information about the occurrence of an incident and ex post information on forgone payoffs (FPs) (i.e., travel times on nonchosen routes). Data were collected from an interactive experiment in which subjects made multiple rounds of route choices on a hypothetical network subject to random capacity reductions, and travel times were determined by performance functions of route flows from the previous round. En route real-time information increased the network's travel-time savings and reliability under the experimental setting, yet FP information had the opposite effect. The most efficient information structure in terms of travel-time savings is a combination of real-time information and no FP information. Real-time information at downstream nodes encourages participants' strategic behavior at the origin. FP information appears to increase risk-seeking behavior; it encourages route switching without real-time information and suppresses it with real-time information. These results could be valuable for policy evaluations of further developments of advanced traveler information systems.

Original languageEnglish
Pages (from-to)89-98
Number of pages10
JournalTransportation Research Record
Issue number2243
DOIs
StatePublished - 1 Dec 2011
Externally publishedYes

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'Information impacts on route choice and learning behavior in a congested network'. Together they form a unique fingerprint.

Cite this