The concept of truck travel demand forecasting, internal to a region, has always been built upon modeling discrete truck trip ends, distributing truck trip ends to various origins and destinations using travel time impedances and land use characteristics, and allocating truck trip tables into distinct time periods using factors derived from observed counts. An innovative enhancement to this approach is to apply activity-based modeling (ABM) principles to truck tour characteristics and develop a tour-based truck travel demand model. For the Phoenix MPO, Maricopa Association of Governments (MAG), Cambridge Systematics (CS) acquired third-party truck GPS data, which captures all trips with origins and destinations inside the multi-county MAG region. This data had over three million GPS event records across 20,000 trucks for a one month period. After processing, these trucks yielded about 20,000 truck tours making over 62,000 stops at various land uses. This truck tour database forms a strong foundation to estimate robust tour-based models for various industry sectors.

The objective of the truck tour model is to develop truck trip chains (or tours) by industry sector and to model the tour stop pattern, where the stop pattern defines the number of stops made and the industry sector of each stop. The concept is similar in construct to a passenger ABM’s daily activity pattern model. Once stop patterns are simulated for each tour, location and time period are simulated, conditional on the sequence of truck stops by industry sector. The model system employs a series of choice models to accommodate each of these truck tour dimensions.

This paper focuses on two aspects – (a) processing of truck GPS data, and (b) developing tour-based truck model. The processing of truck GPS data is done for the MAG region to construct a truck tour database necessary for estimating tour-based models. The tour-based model system includes stop generation, stop purpose, stop location, and stop time period models to predict the occurrence of truck stops in space and time for each industry sector. This paper also discusses the calibration and validation of these linked discrete choice models, which output trip chains/tours for different industry sectors.