This dissertation is a culmination of work spanning several modes of travel, multiple datasets, and different contexts. Because the proliferation of new mobility services disrupted the transportation ecosystem, I aim to understand travel behavior and investigate how new and traditional modes intermingle. I focus my attention on Mobility-on-Demand which encompasses ridehailing services such as private ridehailing, ridesplitting, and microtransit. In the order listed, they represent an increasing degree of sharing; an increasing number of passengers will share the same vehicle at the same time. While each of them is distinct, there are interactions amongst them and other modes of travel. Therefore, this dissertation also examines the interactions of Mobility-on-Demand with other modes such as public transit, micromobility, and the private automobile. At the start of this research endeavor, the City of Chicago made available a comprehensive and novel dataset of ridehailing trips. I deploy an unsupervised machine learning technique to uncover patterns of utilization. The clustering of trips reveals identifiable trip categories that reveal the spatio-temporal dynamics of ridehailing demand. Examples of clusters include trips servicing the Chicago airports, trips in the evening likely for recreational activities, and trips to avoid bad weather. After understanding trip types, I use an econometric approach to study the determinants of ridehailing demand with an emphasis on socio-spatial community area differences. During this task, I discover a divergent relationship between private ridehailing and ridesplitting based on community vulnerability. I find that more vulnerable communities, identified via a novel index, are correlated with higher ridesplitting demand whereas the opposite is true for more privileged communities. More vulnerable communities may be taking advantage of the tradeoffs that are in favor of ridesplitting, where sharing a vehicle with a stranger, losing privacy, and increasing travel time are compensated with lower fares. These studies use a ridehailing trip dataset filled with millions of observations, yet these analyses cannot ascertain individual-level contexts and choices because the data does not include rider information such as gender, age, and income. Consequently, I turn to survey-based data to understand individual mode choice. Microtransit is next in the evolution of ridehailing services. It begins to blur the lines between ridehailing and public transit by incorporating a mix of their attributes. For example, it represents an on-demand service and can operate as a curb-to-curb service. I design an efficient choice experiment with microtransit alternatives and accompany it with questions about respondent sociodemographics, attitudes, and current travel behavior. Using the respondents’ current commute mode, the choice experiment seeks to observe the tradeoffs between travel time, cost, and novel features when choosing between the respondent’s current mode and microtransit. By utilizing a discrete choice model that recognizes latent attitudes, I find not only differences between transit and car commuters when it comes to the effects of travel time and cost, but also differing effects of the COVID-19 pandemic: namely, that the pandemic increases the probability of choosing modes that are more private. This highlights the necessity of understanding the short-term impacts and long-term implications of COVID-19 on travel behavior. To investigate the effects of the pandemic, I use survey-based data collected from Chicagoland transit users. Represented in the data are users of the Chicago Transit Authority, Metra, and Pace transit agencies with information on past transit ridership, COVID-19 ridership, priorities for transit investments, and intent to use transit after the pandemic. Also covered in this data is the respondents’ travel behavior involving other modes, which connects to the Mobility-on-Demand theme of this research. I model their ridership status, their intent to return when all health concerns are alleviated, and the possibility of increased ridership with fare integration of multiple mobility services, which is a key feature of Mobility-as-a-Service. The model results indicate the importance of teleworking, the need to explore strategies that will bring back transit ridership equitably, and an opportunity for integrated transportation systems to attract ridership. What is clear from the pre-pandemic studies, microtransit, and transit ridership study is that Mobility-on-Demand plays an increasing role in the transport ecosystem. After all this research, I synthesize the results to understand ridehailing utilization, determinants of demand, the effects of the pandemic, and transportation equity. With a broader perspective on Mobility-on-Demand, I provide policy recommendations and discuss avenues for future research.

Creator

• Soria, Jason Malinay

DOI

• https://doi.org/10.21985/n2-r1gv-aw35

Subject

• Transportation

Language

• en

Alternate Identifier

• etdadmin_upload_891799
• http://dissertations.umi.com/northwestern:15986

Keyword

• mobility-as-a-service
• microtransit
• ridehailing

Date created

• 2022-01-01

Resource type

• Dissertation

Rights statement

• In Copyright