In 2021, almost 70,000 pedestrians were injured or killed in traffic accidents in the United States [1]. Level-5 Automated Driving Systems (ADSs) have the potential to create safer roads by eliminating human errors [2] system. While the development and deployment of level-5 ADSs has been improved, the interactions between pedestrians and ADSs are not fully understood, despite pedestrians being the most vulnerable road users. This study investigated the factors that affect pedestrians’ acceptance of level-5 ADSs and design features for safe and efficient external human-machine interfaces (eHMIs) to facilitate communication. A survey was conducted with 37 participants to investigate the impact of pedestrians’ background, behaviors, and personal innovativeness on ADS acceptance. A follow-up lab study was performed with 70 participants to determine effective eHMI design features. It was found that there was no effect of background information on the acceptance factors or behavioral intention to cross in front of level-5 ADSs, though pedestrian behaviors and personal innovativeness had a significant effect. In the eHMI lab study, both visual and auditory features were used to create eHMIs including external speedometers, audio cues giving advice to pedestrians, and a method of indicating the driving system was level-5 ADSs. This study gives recommendations about the effect of pedestrians’ background, behaviors, and personal innovativeness on eHMI acceptance and intention to cross the street in front of level-5 ADSs as well as several key visual and auditory features pedestrians included in eHMIs.
Pedestrian Interaction with Automated Driving Systems: Acceptance Model and Design of External Communication Interface
Lect.Notes Computer
International Conference on Human-Computer Interaction ; 2024 ; Washington DC, USA June 29, 2024 - July 04, 2024
2024-06-01
20 pages
Article/Chapter (Book)
Electronic Resource
English
Pedestrian Environment Model for Automated Driving
IEEE | 2023
|Analysing pedestrian-vehicle interaction to derive implications for automated driving
TIBKAT | 2019
|