In this research, large streams of truck-GPS data from the American Transportation Research Institute (ATRI) will be utilized to analyze the route choice behavior of long-haul trucks. Specifically, the following aspects will be analyzed.

(1) Diversity of observed truck route choices: Availability of large streams of GPS data provides an unprecedented opportunity to observe and analyze the diversity in route choices. This is because, for any given origin-destination (OD) pair, the analyst could observe and compare the route choices of a large number of trips, as opposed to observing only one or a few trips. Doing so helps in quantifying the number of different routes taken by trucks between an OD pair and paves way for a systematic analysis of the “diversity” in route choices between any OD pair. From a practical standpoint, such analysis of the diversity in observed route choices helps in improving the existing methods on choice set generation for route choice models.

In this study, GPS data from ATRI is used to analyze the diversity in long-haul truck route choices between several TAZ-to-TAZ OD pairs. Preliminary analysis with over 2000 trips between about 20 different OD pairs suggests that the routes used by most trips between any given OD pair overlap considerably, suggesting a smaller extent of diversity in route choices. Further analysis will be conducted by observing the route choices between a larger number of OD pairs and by quantifying the diversity there in. In addition, the factors (e.g., network structure) that explain the diversity in route choices will be explored.

(2) Factors influencing long-haul truck route choice: In addition to analyzing the diversity in truck route choices, this study analyzes the factors influencing truck route choices. Preliminary results suggest that, contrary to what is typically assumed in traditional truck traffic assignment models, travel time is not the only factor influencing truck route choice; a variety of other factors such as travel costs, tolls, and variability in travel time can potentially influence route choice. This study will further explore the role of such factors in explaining truck route choice using GPS data from ATRI.