Observational study on communication in mixed traffic scenarios: Autonomous vehicle versus pedestrian

• Main Authors: Thong, Eric Tze Xin, Zulklif, Abd Fathul Hakim, Karjanto, Juffrizal, Md Yusof, Nidzamuddin, Mohd Jawi, Zulhaidi, Abu Kassim, Khairil Anwar
• Format: Conference or Workshop Item
• Language: English
• Published: 2024
• Subjects: HE Transportation and Communications
• Online Access: http://eprints.uthm.edu.my/11757/1/P16825_b04a2af351ffd00410086d765faeb5ce%209.pdf; http://eprints.uthm.edu.my/11757/; https://doi.org/10.1063/5.0198562

This paper discusses the interaction between a pedestrian and autonomous vehicle during crossing the zebra lane by using inference statistical analysis. In this study, a video-based experiment was carried out at the university campus. On a straight road, two Perodua Myvi were utilized to test vehicle behaviour (yielding, non-yielding), driving style (assertive, defensive), and vehicle appearances (autonomous, manual) over distances of 100 m, 75 m, 50 m, and 25 m. The driver is dressed in a suit that hides the driver to signify that the driver is invisible to the pedestrian. Sixtysix (66) individuals were shown either manually or automatically operating the vehicle. Participants were asked to rate their decision to cross the road when the vehicle approached at the pre-defined distances. According to the findings, the vehicle’s driving style and appearance has no substantial impact on participants' willingness to cross the zebra lane. For both automated and manually driven vehicles, the vehicle's behaviour had a considerable impact on pedestrians' willingness to cross a road. There was also no statistically significant (p>0.05) three-way interaction between vehicle behaviour, driving style, and vehicle appearance. There was no statistically significant (p>0.05) simple two-way interaction between vehicle behaviour and driving style. Finally, the experiment's limitations were reviewed, followed by recommendations for future study