TRB 2016 Blue Ribbon Committee
16th National Transportation Planning Applications Conference

Tour-Based Truck Travel Models using Truck GPS Data

Corresponding Author: Arun Kuppam, Cambridge Systematics, Inc.

Presented By: Arun Kuppam, Cambridge Systematics


As part of SHRP2 C20 freight model grant, which Maricopa Association of Governments (MAG) successfully secured, calls for the development of a next generation freight demand model that involves synthesizing firms, linking suppliers with buyers, creating supply chains, estimating truck tours by industry sector, and integrating it with regional activity-based model (ABM). In order to develop truck tour-based models, MAG acquired and processed truck GPS data from two vendors – StreetLight Data, Inc., and American Transportation Research Institute (ATRI). The processed data from StreetLight Data yielded a database of 266,832 tours and 1,216,754 trips from over 17,000 single-unit trucks. On the other hand, the processed ATRI data resulted in about 81,090 tours from 39,080 combination-unit or heavy trucks. These two truck tour databases formed a strong foundation to estimate robust tour-based models for various industry sectors for three truck types (light, medium and heavy).

The objective of the truck tour model is to develop truck trip chains by industry sector by truck type. These truck trip chains are then grouped into the major linkages based on land uses the trucks make stops at and the probability of making another stop based on the number of previous stops. The tour-based model generates the number of stops by industry sector, number of stops on a tour, stop purposes, and the location and time of day of stops.

All the tour model components were coded in R, and each component was individually assessed and calibrated. The reasonability of explanatory variables were determined by their magnitude, t-statistic and their relation to the dependent variable. The individual model outputs were also compared against the truck GPS data to assess the model performance. These comparisons indicated that the model components are predicting very closely to the observed data.

This presentation will focus on two aspects – (a) processing of truck GPS data from two different sources, and (b) developing tour-based models by truck type. This presentation will also discuss the calibration and validation of these models.


Discuss This Abstract