Contribution
Localization of Multiple Electric Vehicles
* Presenting author
Abstract:
Electric vehicles typically emit little noise at low driving speeds and must be equipped with acoustic vehicle alerting systems (AVAS), radiating artificial warning sounds. Which sounds are best suited for this application has been a topic of recent research, investigating both warning capabilities and the environmental noise impact of AVAS. While previous studies focused primarily on single vehicles, we present a laboratory experiment on the simultaneous localization of up to three electric vehicles, simulating a static parking lot scenario. In a laboratory experiment with 55 participants, three common AVAS signal types were compared to a combustion engine sound, utilizing a circular array of 24 concealed loudspeakers. The results show that AVAS consisting of two amplitude-modulated pure tones leads to a significantly larger localization error and slower localization time than combustion engine noise, narrowband noise AVAS, and an AVAS sound of 25 harmonic tones. Introducing multiple vehicles with the same kind of sound drastically increased localization error and time for all three AVAS signals compared to combustion engine noise. These findings indicate that, especially in multiple-vehicle scenarios, the auditory localization of existing AVAS solutions is more challenging than for combustion engine noise, with the amplitude-modulated pure tone AVAS performing particularly poorly.