Scalability in External Communication of Automated Vehicles

Abstract

Automated vehicles will alter traffic fundamentally. While users can engage in non-driving-related tasks such as reading or even sleeping, the possibility to interact with other road users such as pedestrians via, for example, eye contact vanishes. Therefore, external communication of automated vehicles is currently researched with various concepts spanning dimensions such as anthropomorphism, technology, viewpoint, locus, message type, and others. However, the proposed concepts are mostly evaluated in simple scenarios, such as one person trying to cross in front of one automated vehicle. Therefore, we implemented a WebGL application of a four-lane road and conducted a within-subject study (N=46) to study the effects of nine concepts with and without the presence of other pedestrians and altering the yielding target of the automated vehicle. We found that all concepts were rated better than having no external communication. However, the effects were not uniform across the concepts.