Electronic Travel Diary

Project Members

Executive Summary

Urban and regional planners use travel demand models to estimate changes in transportation activity over time. These models predict the number of trips generated by households as a function of various demographic and socioeconomic considerations and also predict the number of trips attracted to various employment and commercial centers. Estimates for vehicle mode choice, distribution of trip destinations across the metropolitan region, and traffic volumes on various roads also come from these travel demand models. Regional travel surveys, or travel diary studies, are used to collect the data necessary to build travel demand models. Randomly-selected survey participants record trip information over a one-day, three-day, or one-week period. Data collected from thousands of households across the region are then analyzed to develop travel demand models. These regional travel estimates are then used to predict emissions from motor vehicles and serve as primary input data for air regional quality analyses.

Emissions from motor vehicles are directly related to the amount of vehicle activity undertaken. Increased miles or hours of travel results in increased emissions. Emissions are also strongly related to the percentage of dirty vehicles operating in a region. However, even the emissions from "clean vehicles" that normally pass standard smog checks are a function of the way in which the vehicles are operated. Emissions are not only a function of miles driven, but a function of such trip characteristics as time lag between engine starts and the number of hard acceleration activities. Determining the operating characteristics of onroad vehicles is critical to understanding vehicle emissions behavior. Furthermore, some vehicle operating characteristics are likely to be correlated with driver behavior. For example, young male drivers may be more likely to undertake high-speed hard-acceleration activities which significantly increase vehicle emissions. Current emissions models do not take into account driver interaction with the vehicle and vehicle controls because such data are not available for analysis.

To date, vehicle activity studies and travel behavior studies have never been coordinated. This research effort focused on the development of electronic monitoring equipment that would allow such studies to be undertaken jointly. The new equipment will allow simultaneous monitoring of trip characteristics and vehicle and engine operating conditions. Hence, data collected with equipment developed in this research effort will help model developers understand the linkages between travel decisions and vehicle operating conditions. This, in turn, will lead to the development of improved algorithms for predicting emissions as a function of vehicle characteristics, onroad traffic conditions, and household/driver demographics.

A literature review was undertaken to investigate the various aspects of developing an integrated vehicle instrumentation package. Topic areas covered in this review include past and present travel survey methodologies, trends in the field of travel behavior research, results of other vehicle instrumentation studies, availability of useful technologies, survey bias and response rate issues, and traveler route choice studies. One of the greatest benefits of this review was the sharing of research findings among the limited number of researchers currently conducting automated travel diary and vehicle instrumentation studies.

To collect all required data streams for use in travel model development and validation, the integrated vehicle instrumentation package must contain a handheld travel diary, global positioning system, onboard engine monitor, interface computer, and power supply. Because many manufacturers and models within each equipment class had the potential to meet minimum design and performance criteria, the research team first developed a set of system goals in the form of functional specifications. These functional specifications served to narrow the field of potential equipment that would meet project requirements. Given a smaller set of potential equipment solutions, the research team reviewed the detailed technical specifications for the most promising equipment options. The most promising components identified and reviewed were then selected for purchase and field testing.

Each class of equipment was first tested separately to determine superior performers within each class. To facilitate competitive testing, the research team identified those factors that may influence the accuracy and ease of use of each equipment type. With this knowledge in mind, the team then developed standardized test scripts and testing procedures that would challenge the ability of equipment to perform under real-world test conditions. The descriptions of the test metrics and the performance of each individual component are summarized in the test plans and results document. The selection of individual components, final assembly of the prototype system, and software development were based upon these test results.

The final report summarizes the project in its entirety. Information presented in the four previous project volumes is summarized throughout this report. An review of each system component that is in the final prototype package is given, including the original specifications, test plans and results, and final recommendations. The testing of the final prototype unit is summarized in this final report as well. The functionality of the assembled unit is reviewed and shortcomings that must be overcome before such a system is fully deployed are identified.

The literature review and communications with other related research initiatives revealed several problem areas for instrumentation projects. Project development and testing exposed additional issues that should be taken into consideration in future electronic travel diary and vehicle instrumentation research efforts. These problem areas and issues include:

Manufacturer support for various equipment components is highly variable. Many of the candidate components identified for testing, and some of those actually procured for testing, were no longer being manufactured one year later. The replacement models (if any are available) will not always meet system specifications. In addition, several vendors went out of business in the midst of negotiating with the project team – this reflects the unstable nature of new, specialized technologies.
Technical expertise at many equipment companies is not readily available; in fact, the research team often found that there was only one person within a given company who could answered detailed questions regarding equipment specifications.
Route choice accuracy is a function of GPS accuracy. Field tests dramatically demonstrated the importance of using GPS units that provide accurate position under conditions of Selective Availability (purposeful degradation of the satellite signals by the military to reduce position accuracy). The research team determined that a unit capable of data post-processing (correction of the signal to compensate for military signal degradation) is required for future systems.
An alternative to post-processing GPS data, which is both time and resource intensive, is the use of GPS equipment that can receive position corrections via radio signal. For use in metropolitan area studies, these units need to be capable of receiving the correction signal throughout the region. The research team is currently testing two units in Atlanta under a separate research effort.
GPS units do not all initialize with the same time stamp when they are powered on, resulting in a potential clock offset between two separate units. Clock drift was also noted across the variety of GPS units tested, indicating that the units employ different internal algorithms for tracking the passage of time. Any research effort aimed at integrating data based on time (matching the location of an event and the environmental conditions associated with that event) must consider clock drift. An independent, single time stamp should be applied to all recorded data streams if second-by-second matching of these streams is a post-processing requirement.
The onboard engine diagnostics units currently available on the market do not yet provide an optimal solution. Currently, the OTC scanner provides the best capabilities for vehicles. However, the OTC has to be specifically configured for each vehicle. When operated in a generic onboard diagnostics (OBDII) mode, engine data are only reported every three-seconds, which is insufficient to provide accurate acceleration data. New systems are entering the market this year and should be considered for future studies.
Onboard computers and other components must be capable of performing under very high temperature conditions if they are to be located in the vehicle trunk. Few computers are available which can perform at temperatures exceeding 140 F.
Equipment durability is a critical issue. Component failure during data collection efforts will result in the loss of all data for the duration of the equipment outage. A warning interface to notify the driver of component or system failure should be built into future systems.
Off-the-shelf 12-volt deep cycle batteries are not capable of storing their rated charge. Batteries must be fully discharged and recharged 30 times, before their rated charge can be met. Field study teams need to purchase and "season" batteries well in advance of any studies being undertaken.
The power demand of the current prototype system is such that a single battery charge will only power the unit for three days. Adding a second battery will extend the duration of tests, but adds significant weight to the vehicle and may influence the response of the engine to various driving conditions by adding to vehicle load. Hence, future system research should continue to focus on reducing the power demand of the various components and on new battery technologies.

This final report also summarizes the linkage between this project, sponsored by the Federal Highway Administration and Georgia Department of Transportation and the ongoing research effort associated with the development of the year 2000 travel survey in the Atlanta metropolitan region (SMARTRAQ).

Back to Top

Copyright 1998-1999 GeorgiaInstitute of Technology
For problems or questions regarding this web contact [jwolf@trec.ce.gatech.edu].
Last updated: March 22, 1999.