Analysis of Morning Commute Route Choice Patterns Using Global Positioning System-Based Vehicle Activity Data

Abstract

This paper examines morning commute route choices of 182 drivers, with the use of disaggregated Global Positioning System–based vehicle activity data collected during a 10-day period. This paper attempts to describe how these commuters tend to behave in the real world. A binary logit model of morning commuters’ choice between a single commute route and multiple routes was established on the basis of evidence of drivers’ varying valuations of a number of route and trip characteristics as well as commuters’ sociodemographic characteristics. The research results of this paper indicate a strong relationship between the morning commute route choice decision (single versus multiple routes) and commuters’ work schedule flexibility, sociodemographic characteristics, and commute route attributes.