U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

 
REPORT
This report is an archived publication and may contain dated technical, contact, and link information
Back to Publication List        
Publication Number:  FHWA-HRT-13-097    Date:  September 2014
Publication Number: FHWA-HRT-13-097
Date: September 2014

 

Analysis of Network and Non-Network Factors on Traveler Choice Toward Improving Modeling Accuracy for Better Transportation Decisionmaking

CHAPTER 1. INTRODUCTION

 

BACKGROUND

Transportation network flows and the associated performance of these systems are largely the result of choices made by travelers—choices of where to live and work; which activities to engage in and with whom; and where, when, by what mode to get there and along which path. These choices reflect travelers’ activity patterns (i.e., work and residence location, mandatory and discretionary activities, etc.) that motivate their desire to travel, situational and environmental variables (e.g., weather), as well as attributes of the transportation system, which determine the users’ experienced service levels. In addition to their own experience of the system’s attributes, particularly congestion and reliability, travelers are influenced by the information they obtain or otherwise receive about system conditions as well as various controls such as toll prices, access limitations, dynamic control, and other measures.

Researchers’ understanding of traveler choice behavior in transportation systems and the approaches used to capture its outcomes has undergone several paradigm shifts over the past 50 years, often through the involvement of different disciplinary perspectives. From aggregate models concerned primarily with total travel between traffic zones, to disaggregate models of household travel decisions, to the current interest in activity-based models that view travel choices in the context of the activity engagement decisions of individuals and households, the field has expanded, evolved, and matured considerably. Yet, while sociologists, geographers, and psychologists have provided valuable insights into many aspects of what people do as well as where and why they do it, the ability to operationally represent traveler choices and use this representation for the purpose of predicting how users will respond to various transportation system interventions remains rather limited.

Tools available to support operational analysis and strategic planning of transportation systems have reached considerable levels of sophistication in the past two decades. On the supply side, microscopic and mesoscopic simulation tools can now be applied to large networks. Most traffic microsimulation tools only consider driving decisions such as car following and gap acceptance but do not include tactical and strategic dimensions. On the demand side, activity-based models for strategic planning purposes are being implemented by several metropolitan planning organizations (MPOs). However, these models have generally lacked sensitivity to network performance attributes, particularly path-level attributes such as time-varying travel times and travel time reliability. As such, they cannot readily be used in the context of operational analysis tools to capture user responses to traveler information and traffic control measures.

Developments in simulation-based network models have provided a suitable platform for integrating user choices in operational analysis tools, especially with regard to incorporating route choice and response to traveler information in modeling traffic flows in highway networks. However, the observational/empirical basis for the behavioral models has been limited, often relying on small-scale laboratory experiments and stated choice methods to calibrate the models.

The needs of transportation agencies for methods to evaluate the relative impact of a growing array of system management measures and policies involving both aspects of the network as a well as non-network elements underscore the need for improved methods that can integrate a richer behavioral basis than is currently available in existing tools. The missing element in many situations tends to be in the representation of traveler choices in a network setting and the influence of both network and non-network variables on these choices.

This project is intended to address this gap in modeling capability to support a variety of initiatives that seek to improve traffic conditions, system safety, and sustainability by targeting user choices before and during travel.

PROBLEM STATEMENT

The main goal of this study is to illustrate the missing gaps in capturing traveler choices in the methods and tools intended for use in operational analysis and planning of a wide range of measures and policies aimed at improving the efficiency, reliability, sustainability, and safety of the transportation system. Another main goal is to present and demonstrate ways to overcome these gaps with available operational tools through case studies. Traveler choice behavior, particularly the dynamics of this behavior in interaction with network and non-network variables, is a challenging domain. Developments regarding various aspects of travel and activity behavior have multiplied over the past few decades, but researchers’ ability to predict these responses in conjunction with system planning and evaluation has not improved commensurately. This need is especially critical to the success of emerging program areas such as ATDM, ICM, and AERIS, as well as to much needed enhancements to traffic operations management in connection with adverse weather (weather-responsive traffic management (WRTM)). Similarly, the ability to predict the impact of planning interventions to non-network elements of the urban landscape, such as enhancements targeting walkability, neighborhood safety, and sustainable development density, is of much interest to the professional planning community.

OBJECTIVES AND APPROACH

The main emphasis of this effort is on travelers’ higher-level predictive strategic choices that determine when and how they might use the transportation system. These might be influenced by a range of variables, including experienced system performance (i.e., recurrent and non-recurrent congestion and travel time reliability), environmental factors (i.e., weather that affects both system performance and activity engagement opportunities, availability, and accessibility to alternative modes), availability and cost of parking, quality of the walking environment, and measures such as pricing, information supply, dynamic traffic management, etc. A thorough understanding of the determinants of travel choices and behavior and an operational ability to model their dependence on key attributes of the transportation system, network performance, and non-network factors will provide a foundation for designing effective interventions to improve system performance and for evaluating different policies and options by predicting how users will respond to these measures.

The approach adopted in this study is to provide a high-level discussion of all relevant issues in the context of a general, comprehensive framework and then demonstrate the targeted implementations of this overall approach through specific case studies. Each case study addresses one or more of the programs and policy interventions motivating the overall effort. For each case study, the relevant behavioral dimensions were identified, and, when applicable, integrated into a modeling framework that captures the interaction of these dimensions with the relevant supply-side elements. Where available, data were compiled to characterize these behaviors and, in selected cases, to develop and calibrate new models and specifications. For several case studies, the modeling framework developed for that application was applied to predict the effect of selected policy interventions and conduct sensitivity tests to various underlying behavioral parameters. This approach allowed researchers to address the study objectives through specific indepth applications, which, taken collectively, reflect the range of questions and interventions that require prediction of user behavioral responses as well as the range of methodologies and modeling perspectives in addressing these problems. As such, the study contributes to both its methodological and applied objectives.

In addressing the study objectives, the case studies, associated behavioral models, and integration of the models developed in this study are intended to satisfy the following requirements:

The developed models in this study represent a wide a range of analysis tools and demonstrate specific applications of the methods. The developed models also show that there is not one single model or tool capable of addressing all questions that arise in conjunction with management strategies such as ATDM, ICM, AERIS, or WRTM. In other words, the notion of a “one size fits all” condition in terms of modeling capability or resolution is not practical for the range of questions involving behavioral responses of travelers. To address this, the developed case studies use a range of models to achieve the right balance of detail, accuracy, computational tractability, and usefulness for the application under consideration.

REPORT ORGANIZATION AND STRUCTURE

This report begins with an overview and conceptual framework that organizes the wide range of issues and approaches in modeling traveler choices. The framework structures the discussion and leads to identifying the main gaps in terms of current modeling practice versus needs in terms of scope of coverage, usable tools, and relevance to the questions of interest. Based on this gap analysis, the road map is framed to illustrate how these gaps can be addressed at different levels in detailed case studies. Rather than address gaps in a generic fashion, intended for universal applicability but missing some essential aspects of specific applications, the development takes place in specific scenarios where methods and models are used and demonstrated. The different case study scenarios are ordered by decreasing time frame, with longer-term considerations and behavioral adaptations addressed first, progressively leading to short-term, day-to-day, and within-day responses of travelers. Some methodologies overlap. As a result, in order to avoid duplicate discussion, model discussion is organized to best highlight contrasts in advantages and drawbacks. The report organization is as follows:

 

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101