The purpose of this study was to acquire cellular-based origin-destination (OD) travel data from AirSage and evaluate the trip patterns by comparing it with other independent data sources. These included the Phoenix MPO’s regional travel demand model and the 2008 NHTS Add-on dataset from the Maricopa Association of Governments (MAG).

A “30 zone by 30 zone” trip matrix data-set provided by AirSage includes the following specifications:

1. Coverage – 30 x 30 zones from aggregated TAZs for the MAG region;

2. Day of Week – Average weekday (Tuesday, Wednesday and Thursday) in one month specified by MAG;

3. Time of Day – AM Peak, PM Peak, Off-peak, and Daily;

4. Residence Class Attributes – Residents and Visitors.

Some of the key findings from the comparisons include: (1) The daily total trips matched well between AirSage and the validated travel demand model; (2) The number of trips by peak period varied in different parts of the region; (3) The number of trips by zones were close enough in AirSage and the model; (4) The trips going to downtown were different from AirSage and the 2008 NHTS dataset; (5) The zone to zone differences are minimal between AirSage and the model, but the intrazonal differences are pretty significant; (6) The AirSage data seem to capture a very low number of intra-zonal trips compared to the model; (7) The AirSage data does not have many variations in trip lengths within each time period while the MAG model provides a bell shaped curve for each of the peak periods; (8) An assignment of AirSage data showed a good match between volumes when compared against the model, while the AirSage data was consistently higher in the individual time periods.

This study explores potentials of using AirSage cellular-based OD data in travel demand modeling. Specifically, this paper focuses on providing several key summaries from the comparisons of AirSage data with the MAG travel demand model and the weighted 2008 NHTS datasets. These include total number of trips by time period, at the zonal level and to/from select zones (e.g., downtown); trips by trip lengths for each time period; and screenline analyses of AirSage versus modeled trips.