Dear Diary

March 1, 2008 By: Ayalew Adamu, Bahram Ravani, Kin S. Yen, Ty A. Lasky GPS World

Travel Behavior Gathered with High-Sensitivity GPS

How do you increase capacity of transportation systems without new construction? Through improved planning. Gathering data on who is traveling where, when, and why is being made possible through the use of the Global Positioning System Automated Travel Diary (GPS-ATD). The intuitive user interface captures trip information, with minimal user input and burden, and wireless ties personal and vehicular GPS-ATD units together. The result? A unit that could prove useful for the 2010 California Statewide Household Travel Survey.

Most current transportation systems are at or have exceeded their maximum capacity, and adding capacity through building is no longer a viable option due to space or environmental constraints. Better planning — at the city, regional, and state levels — can reduce the demand on the transportation system.

But improved planning requires a better understanding of traveler behavior. Traditional cross-sectional survey methods measure traffic at a single point of a road or intersection, providing traffic loading at a specific location over time. While this may support conclusions regarding capacity at a point, it does not provide needed information on traveler behavior such as trip purpose, trip frequency and schedule, route selection, and speed profiles throughout the entire route. In comparison, longitudinal surveys directly measure traveler behavioral change at the level of the individual traveler, and provide information leading to a better understanding of the factors that influence and direct personal travel behavior. Such data are critical in:

1. developing travel demand models and forecasting future demand;
2. predicting the number of trips generated by households as a function of demographics, socioeconomics, and location relative to employment and commercial centers;
3. estimating travel-mode choice and traffic volumes on specific roads;
4. assessing the impact of changes in the transportation system or policies;
5. predicting emissions from motor vehicles and input for air-quality analysis; and
6. calibrating regional travel models.

Household longitudinal travel surveys collect three categories of data: household information, household member personal information, and travel activity information for particular days. Travel diaries are filled out by survey respondents or the surveyors after questioning the respondents; diaries are the standard method used to capture participating household travel activity information.

But gathering complete information in this way has been problematic. The drawbacks of self-administered paper-based survey designs are well known, and this approach is not suited for long-term mobility pattern observations. Computer-assisted self-interview methods, in which respondents input information into a computer, are an improvement over paper-based surveys, but respondents need access to a computer and the Internet. Moreover, multi-day personal surveys often suffer from responents' survey fatigue and low response rates typical in longer survey durations. It is common for respondents to underreport or to provide incorrect data due to poor memory, misunderstood instructions, or carelessness. Short or infrequent day trips are the type most often not reported.

To accurately capture the full spectrum of travel behavior, surveys need to be carried out over several weeks while maintaining data accuracy and minimizing the respondents' burden. Earlier longitudinal surveys using GPS have shown great potential. GPS-based surveys are more accurate and minimize the respondent burden. In one study, 75 percent of respondents took less than one minute to enter all required trip information into the travel diary, compared to 10 minutes on a paper diary or 20 minutes on a follow-up phone call. In addition, GPS digital data can be readily imported into computer analysis programs. This approach captures route choice, path, and speed profile information — items not feasible with traditional paper surveys. Such data may be used to measure the level of congestion of a particular highway, for instance.

Previous GPS travel diaries may be classified into two types: interactive and passive. Some previous GPS travel diaries were interactive, requiring the respondent to input survey information such as marking trip start and end, trip purpose, trip cost, and travel mode. The majority of previous GPS travel diaries were passive GPS data loggers requiring no interaction with the respondent. The respondents only needed to carry and turn on the device whenever they traveled. Other trip information is collected through paper diaries or follow-up phone call, or is estimated by computer-aided software based on the GPS data.