SBIR Phase I: Synthetic Household Travel Data Using Third-Party Targeted Marketing and Mobile Phone Data

Period of Performance: 01/01/2014 - 12/31/2014

$150K

Phase 1 SBIR

Recipient Firm

Transport Foundry
3423 Piedmont Rd Ne
Atlanta, GA 30305
Principal Investigator, Firm POC

Abstract

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project includes improving the transportation planning process through better and more frequent data. An integral part of transportation planning at the federal, state, regional, and local levels of government involves testing infrastructure, policy, or population changes with travel demand models. Improved travel demand models enable more accurate predictions and more efficient funding to achieve important regional transportation goals, such as decreasing congestion on a specific corridor, enhancing regional economic growth, improving safety, and attainment of air quality standards. Unfortunately household travel surveys, the backbone data for these models, are increasingly expensive. They are so prohibitively expensive that sample sizes typically represent less than 0.5% of the population, and they are only collected about every ten years. Furthermore, only 5-10% of households respond to surveys, creating a significant nonresponse bias. This SBIR Phase I project aims to create a replacement for traditional household travel surveys with up-to-date, inexpensive, readily available, and representative data. The goal is to transform the practice of travel and urban modeling, helping the nation achieve important transportation and infrastructure goals for the future. This Small Business Innovation Research (SBIR) Phase I project will determine whether synthetic household travel data can be created by combining targeted marketing (i.e. consumer) data and anonymous mobile phone data from many different sources to create realistic synthetic household travel data for a full population in such a way that individual privacy is protected while near real-time travel data is provided to decision-makers. The process will use statistical weighting, imputation, simulation, and data fusion techniques to create synthetic household travel data at a fraction of the cost of current survey-based data collection methods. By fusing data together to create inexpensive, up-to-date, and detailed synthetic household travel data at regular intervals for nearly 100% of the population, researchers and decision-makers alike will be far better equipped to make important transportation decisions that affect our lives each day.