Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Darcy Bullock


This dissertation describes a methodology for conducting travel time studies based on global positioning system (GPS) receivers and geographic information system (GIS) technology. Compared to traditional approaches, the new methodology provides consistency, automation, finer levels of resolution, and better accuracy in measuring travel time and speed. These characteristics enable the detection of localized traffic effects much more efficiently than with traditional techniques. The new methodology automates the data collection and data reduction procedures, and provides improved procedures for documenting and analyzing travel time and speed data. As a result, large amounts of reliable travel time and speed data can be collected and processed. The new methodology was used to process 28,000 miles of travel time runs on 4,000 highway segments (nominally 0.2-mi long) along 660 miles of highways in three metropolitan areas in Louisiana (Baton Rouge, Shreveport, and New Orleans). Nearly 183,000 segment travel time and speed records based on GPS data collected every 1 second were obtained. Using 0.2-mi segments was justified by an analysis which showed that segment lengths shorter than 0.5 mi were needed to detect localized traffic effects. These traffic disturbances became visible only when segment lengths were at most half the length of the associated disturbance. The analysis also showed that traditional link-based segments, which are typically longer than 0.5 mi, were not sufficient to characterize localized effects properly. GPS time intervals larger than one second were simulated. The analysis showed that all segments had GPS data only when the GPS time interval was at most half the shortest segment travel time observed for the run. The analysis also showed a direct relationship between segment speed uncertainties and GPS time intervals. Median speeds were shown to be more robust estimators of central tendency than harmonic mean speeds. Segment speed coefficient of variation maps were shown to be powerful indicators of traffic flow variability along highways. The ITE Manual of Transportation Engineering Studies was shown to seriously underestimate required sample sizes for travel time studies (or alternatively, to overestimate confidence levels associated with specified sample sizes). A new formulation was then developed to correct the shortcomings.