Mobility On-Demand exists in many forms in the Dallas-Fort Worth region.  For some transit agencies, mobility on-demand service has been added to provide first and last mile services to existing fixed routes.   In other cases, mobility on-demand service has begun to supplant fixed route service.  Some cities without any existing transit have developed mobility on-demand service to satisfy the needs of the area.  In 2024, Dallas Fort-Worth had over 12,000 annual daily ridership in mobility on-demand service.  Throughout the nation, mobility on-demand services are growing, but a deeper understanding of the users is still needed.

In Fall 2022 and Spring 2023, North Central Texas Council of Governments (NCTCOG), Dallas Area Rapid Transit (DART), Denton County Transportation Authority (DCTA) and Trinity Metro partnered together to conduct an on-board, Origin-Destination Transit Survey, aimed to examine the transit in North Central Texas. The purpose of the study was to collect the travel patterns and trip-making decisions of weekday transit passengers to assist the transit agencies in their planning process and for use in NCTCOG’s regional travel model. In addition to surveying fixed route services, this project also included surveys on Mobility On-Demand services of four local transit agencies. An understanding of how riders make use of the mobility on-demand services in conjunction with the rest of the transit system can provide insights into how to better plan for transit in the future.

The Mobility On-Demand survey was conducted on Arlington Transportation, DART GoLink, DCTA GoZone and Trinity Metro ZIPZONE.  The survey was administered in two ways: through a self-administered survey available on an app or website, and as an in-person survey completed with a surveyor on tablets.  The survey consisted of questions about their origin and destination, mode of access and egress, trip purpose, routes used, fare and payment, and household and person demographics.  Over 1,500 surveys were collected from riders.   The results of the survey provided insights into the mobility on-demand trips as well as allow comparisons to the fixed route survey trips.

The survey results included information about trip characteristics.  86% of one-way trips surveyed only included mobility on-demand on their trip and did not include any other form of public transportation. 82% of the trips were home-based.  47% of trips were work-related. 

Responses from the survey also provided details about the ridership.  11% of trips were made by riders who have a disability that affects their mobility.  Approximately 32% of trips were made by students.    35% of trips came from households with an annual household income of less than $20,000. 

Abstract Background

Electric vehicles (EVs) are set to transform the automotive industry by reducing reliance on fossil fuels and lowering carbon emissions. However, their higher upfront cost continues to deter many potential buyers, prompting government agencies to implement financial incentives—such as rebates—to stimulate adoption. California’s Clean Vehicle Rebate Project (CVRP) is a key example, offering rebates to offset the initial cost of buying or leasing EVs. However, EV adoption rates vary widely, raising the question of whether a uniform incentive program can be equally effective across diverse communities. This study addresses this question through a case study of the CVRP, examining how demographic, economic, geographic, and infrastructural attributes shape the program’s uptake.

Description of Abstract

This research outlines an advanced framework to evaluate effectiveness of statewide EV rebate programs. To do so, California’s CVRP data from 2010 through 2023 was utilized, supplemented by numerous county-level data from various sources, such as U.S. Census for demographic statistics and Department of Transportation for infrastructure details. Once the dataset was compiled, a two-phase machine learning (ML) approach was employed. First, dynamic time warping (DTW) in conjunction with K-Means clustering was applied to group counties based on temporal patterns in rebate participation for battery electric vehicle (BEV) and plug-in hybrid electric vehicle (PHEV). Subsequently, a Random Forest (RF) model was developed to predict county-level rebate participation using demographic, economic, geographic, and infrastructural characteristics. 

Results showed that counties with greater racial diversity, higher educational attainment, stronger economic conditions, and more urbanized built environments, including denser charging station networks and advanced public transit systems, exhibited notably higher rebate participation, particularly for BEVs. Conversely, rural counties and those with older populations, lower income, and limited infrastructure tended to engage much less with the rebate program. The RF model demonstrated strong predictive accuracy (R² = 0.82 for BEVs; R² = 0.78 for PHEVs), underscoring the utility of ML for policy planning.

Based on these findings, we recommended geographically targeted infrastructure investments, enhanced financial incentives for underserved communities, and tailored outreach campaigns to more effectively promote the adoption of EVs and improve the impact of rebate programs.

Statement on Why Abstract is Noteworthy

This study leverages advanced clustering and predictive modeling techniques to derive actionable insights that enhance the effectiveness of EV financial incentives, thereby preventing budget waste. These insights contribute to a replicable framework that can be applied to similar EV incentive programs. The codes developed in this study are available upon request and will be hosted on GitHub, facilitating broader application. Planners and policymakers can use this framework to improve resource allocation efficiency and accelerate the transition to clean transportation across diverse communities.

Next Steps

Future work should explore causal inference methods to better identify the direct impact of incentives and infrastructure, and integrate post-2023 data to capture the influence of emerging technologies, policies, and funding structures on EV adoption.

Sara Sohaee Urbina, ssu2@pdx.edu

Liming Wang, lmwang@pdx.edu

Abstract Title

Bridging the Mobility Gap: Connecting Older Adults with Shared Autonomous Vehicles (SAVs) in Transportation Planning

Topic

Automated Vehicles

Abstract Background

The United States faces a growing challenge in providing equitable mobility, accessibility, and safety for its rapidly aging population. Shared Autonomous Vehicles (SAVs) offer a potential solution to maintain older adults' independence, quality of life, and access to essential services. However, successful implementation of SAVs hinges on understanding this demographic's perceptions, needs, and potential adoption barriers. Current research in this area remains limited, creating a critical gap in transportation planning knowledge.

Description of Abstract

This mixed-methods study directly addresses this gap by investigating the factors influencing older adults' willingness to use SAVs. Guided by the Technology Acceptance Model (TAM), we analyze quantitative and qualitative survey data collected from older adults across Washington, Oregon and Texas. The research assesses perceived usefulness, ease of use, safety concerns, and social influences. We identify key benefits and barriers to SAV adoption. Preliminary findings suggest the potential benefits of SAVs, including increase in road safety, and enhanced mobility and accessibility, while also identifying challenges to their acceptability, such as technological concerns. 

Statement on Why Abstract is Noteworthy

This application is noteworthy because it bridges a critical gap in transportation planning literature by focusing on an often-overlooked demographic—older adults. Unlike traditional transportation studies, which generally prioritize younger, tech-savvy populations, this research offers nuanced insights into the specific needs and concerns of aging individuals. It contributes to the planning applications community by proposing actionable strategies for enhancing SAV accessibility and trust among older adults. Furthermore, it provides evidence-based recommendations for policymakers, urban planners, and technology developers to design more inclusive transportation systems that accommodate the diverse requirements of older populations.

Project will be complete by Fall 2025, Expected Milestones:

• Data Collection will be done by March 2025.

• Data Analysis will be completed by April 2025.

By the year 2050 Connected and Automated Vehicles (CAVs) have the potential to significantly disrupt travel demand across the country. As such, there is an urgent need to better understand the potential impacts and benefits of CAVs on travel demand and travel forecasts. To address this need, NCDOT funded a research study to better understand these potential impacts and benefits, and to develop guidance on how travel demand models can be modified to best capture the supply and demand side changes that will likely result from different levels of CAV deployment.

This presentation will provide an overview of research findings and case study analysis. It will include recommendations and guidelines for considering CAVs within a strategic scenario planning context using a travel demand model considering both a basic and advanced approach.