Using a virtual environment to study how pedestrians and bicyclists respond to warning and intent communication for autonomous vehicles
Abstract
Details
- Title: Subtitle
- Using a virtual environment to study how pedestrians and bicyclists respond to warning and intent communication for autonomous vehicles
- Creators
- Lakshmi Devi Subramanian
- Contributors
- Joseph K Kearney (Advisor)Jodie M Plumert (Committee Member)Daniel V McGehee (Committee Member)Kasturi R Varadarajan (Committee Member)Juan Pablo Hourcade (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Computer Science
- Date degree season
- Summer 2022
- DOI
- 10.25820/etd.006620
- Publisher
- University of Iowa
- Number of pages
- xii, 77 pages
- Copyright
- Copyright 2022 Lakshmi Devi Subramanian
- Language
- English
- Description illustrations
- illustrations (chiefly color), graphs, tables
- Description bibliographic
- Includes bibliographical references (pages 68-77).
- Public Abstract (ETD)
Vehicle automation technologies are developing at a furious rate. Many automated systems are now standard equipment such as adaptive cruise control, blind spot monitoring, lane assist, and automatic braking. For these driver assistance technologies, the human driver is still present in the car to intervene when any automation fails. However, fully autonomous vehicles that will not require the presence of driver are on the horizon. But traditional driving requires communication with other road users such as warning of potential danger and communicating an intent to yield. Highly functional autonomous vehicles are expected to reduce human error, but they face a new challenge on how to communicate warning and intent with pedestrians and bicyclists.
My dissertation is focused on two research problems: 1) How do pedestrians and bicyclists respond to a vehicle warning system which sends safety alerts during nighttime conditions, and 2) How do pedestrians cross streams of traffic in the presence of an external Human Machine Interface(eHMI) that communicates the intent of an autonomous vehicle to yield or no yield at a pedestrian crosswalk in daytime conditions. I used 3D immersive virtual reality and developed virtual environments that simulate the real-world traffic conditions for my dissertation research. Our results showed that the nighttime vehicle warning system has the potential to draw the attention of pedestrians and bicyclists about oncoming cars and also influence them to behave safely. The results of the intent communication study showed that the pedestrians’ road-crossing behavior was influenced by the timing of the eHMI as well as the way the yielding vehicle slowed down in daytime traffic conditions.
The findings from my dissertation studies help to understand the behavioral changes made by pedestrians and bicyclists towards the latest vehicle automation technologies thereby contributing to the design and development of new vehicle technologies aimed to increase the safety of vulnerable road users.
- Academic Unit
- Computer Science
- Record Identifier
- 9984285050602771