An Exploration of Contemporary Issues in Highway Safety, Evolving Transportation Alternatives, and Activity and Travel Behavior Modeling

Abstract
Several critical issues have emerged in recent years in the fields of highway safety, alternative transportation modes, and activity and travel behavior modeling. Regarding highway safety, there is currently an ongoing methodological debate about the use of data-driven methods (machine learning, etc.), conventional statistics, statistical models that address unobserved heterogeneity, and causality models. The research team will provide an extensive review and assessment of these methodological alternatives and their potential application to highway safety. Next, there have been several recent studies that indicate that driver behavior is changing continuously over time in response changing vehicle technologies, changing behavior and utilization of social media and texting as well as other temporally shifting factors (Mannering, 2018). This has profound implications for highway safety and the development of safety policies and countermeasures. The intent of the safety portion of this study is to explore the temporal instability of driver behavior from various perspectives including the possible temporally shifting effects of aggressive driving and cellphone usage, two elements of driver behavior that are believed to be highly unstable over time. Statistical evidence of possible changes in the effects of these elements over time can help guide public policy and effect mitigation.

The study will also consider the effect of emerging transportation alternatives with regard to the following four options: 1) the market potential for shared autonomous vehicles, 2) the use of bike sharing as a potential auto-trip substitute, 3) the potential for peer-to-peer carsharing, 4) the role that socio-demographics and health-related factors play in ride sourcing behavior, and 5) the potential of subsidizing free electric vehicle charging infrastructure for household vehicle ownership. The first three of these options are part of the growing “sharing” economy concept, where limited resources are used more efficiently by sharing. While this seems to go against the grain of what has become standard American consumer behavior, there is evidence among the young that sharing may have potential to mitigate the adverse environmental effects of traditional transportation modes. Third sharing option, the renting of personal vehicles for monetary compensation (peer-to-peer car sharing) has become increasingly popular in the U.S., but yet surprisingly little is known about the attitudes, perceptions and decision process through which individuals decide to offer their car for rent in such peer-to-peer carsharing. The fourth option seeks to understand the recent growth in the popularity of mobility-on-demand (ride sourcing such as Uber and Lyft) which has already substantially disrupted the transportation market by providing a variety of new transportation options. The fifth options seeks to understand establish an early estimate of the value of free charging in the United States to aid in understanding potential for accelerating electric vehicle adoption in the United States.

The collection of interpersonal (social) network data and its incorporation into activity and travel behavior models is a growing area of travel behavior research. Prior research has found evidence of a link between strong and weak social connections and variations in activity behavior. In previous research, the research team has explored the use of Lin’s conceptualization of social capital and its division into expressive and instrumental resources (Lin 2001). These resources then lead to expressive and instrumental outcomes respectively. This study will propose to explore the theoretical significance of social capital on leisure activity behavior and residential choice. Using Lin’s conception, aspects of activity behavior will be classified into expressive and instrumental outcomes. These efforts will lay foundational steps into developing social network-based activity-based modeling frameworks capable of predicting travel patterns under the adoption of disruptive technologies and analyzing socially focused policy factors such as social isolation.

Research Products and Implementation

Scope of Work

​​Final Report