In order to recommend where roads, transit, buildings, sidewalks, and other infrastructure are needed, planners need to understand people, jobs, other destinations, and the movements between them. Traditionally, local surveys collect this information; however, surveys are increasingly biased, expensive, infrequent, and onerous. The aim of this IDEA project was to apply a prototype in Seattle, Atlanta, and Asheville that fuses passive, big data—including household-level data, firm-level data, and location and speed data from mobile phones—into individual-level synthetic travel diaries. Synthetic travel diaries reveal where, when, why, and how individual people travel, with each person's demographic data appended (e.g. household income, age). The project focused on making the prototype consistent nationally, rapidly deployable for any size city, and systematically updatable over regular time periods. The prototype uses a simulation framework to fuse the passive data with National Household Travel Survey data. The method produces locally sensitive synthetic populations with individual-level travel diaries using the same code in the three metropolitan regions. The validations of time use, tours per day, and geographic distribution of trips were comparable to other validation datasets in each region, and the differences discovered in these measures appear to be reasonable considering the variability in the regional travel estimates. These results suggest that this passive data approach to analysis will allow engineers and planners to investigate travel behavior in a way that is not feasible today, including improvements to travel demand modeling, tolling studies, before-and-after studies, and congestion mitigation studies.

The final report is available.