State Study Investigated Vehicle Miles Traveled (VMT) and Car Ownership Shifts Using Agent-Based Model, Focusing on Technologies Like Shared E-Scooters.
Indianapolis, Indiana, United States
Assessing the Travel Demand and Mobility Impacts of Transformative Transportation Technologies in Indiana
Summary Information
Transformative transportation technologies, such as bike-sharing, shared e-scooters, and ride hailing systems, have been becoming increasingly widespread and thus reshaping the transportation landscape. These technologies have a high potential to significantly change travel behavior and travel demand and in turn, affect planning, operations, and decision-making processes of transportation agencies. This study proposed a modeling framework to estimate the impact of shared e-scooters on the local public transit system in Indianapolis, IN. Additionally, the study explored how transformative transportation technologies (bike-sharing, shared e-scooters, and ride hailing systems) could impact private vehicle usage. For this purpose, an agent-based model that integrated traditional and transformative transportation system was proposed to simulate the Vehicle Miles Traveled (VMT) and car ownership change under different development scenarios of transformative transportation technologies.
METHODOLOGY
For the analysis that investigated the relationship between each e-scooter trip and public transit, the study used a classification model to determine if shared e-scooters in Indianapolis either complemented or competed with the bus system, varying by time, area, and trip type. It analyzed over two million e-scooter trips from two companies (Sept 2018-Dec 2020) and GTFS data on the IndyGo bus system (Jan 2017-Dec 2019) with 3,425 stops and 7,085 trips. The model assessed the feasibility of e-scooters substituting or complementing bus trips for each historical trip.
For analysis that investigated the relationship between transformative transportation technologies and the use of private vehicles, the study used an agent-based model, with households as agents defined by attributes like age, gender, and income, to analyze the impact of transformative transport technologies on private vehicle use. It generated 372,000 household agents from 2018 Marion County data based on American Community Survey. Each member had utility functions for travel behavior and mode choice. The simulation model evaluated three scenarios: (i) traditional transportation modes, (ii) the current scenario with the current service availability and pricing of transformative technologies in Indianapolis, and (iii) the alternative scenario that varied the availability and cost using the current scenario as the basis.
FINDINGS
Results from the relationship classification model:
- The results revealed that, overall, 44 percent of the e-scooter trips could likely replace walking, while about 27 percent of the trips could potentially compete with bus trips, and 29 percent complement the existing bus system.
- Most of the complementary trips provided service outside of bus service time, shortening bus route distance, and reducing unacceptable waiting time, accounting for two, five, and 20 percent of the total trips, respectively.
Results from the agent-based simulation model:
- The results showed that, when the cost of shared mobility modes dropped by 60 percent compared to current levels, the VMT replaced by shared micro-mobility could breakeven with the VMT increase from ride hailing.
- Further decreasing the cost could bring up to a two percent net VMT reduction when the shared micro-mobility availability was increased by 600 percent. However, it should be kept in mind that expanding the shared micro-mobility system may also cause other issues, such as sidewalk crowding, intensive rebalancing, power consumption for battery charging.
- The results also showed that the deadheading and rebalancing of ride hailing vehicles could cause additional VMT by up to 30 percent.
