Auditory Warnings for Electric Vehicles: Detectability in Normal-Vision and Visually-Impaired Listeners

Abstract Electrical vehicles operating at low speed are often too quiet to be detected by pedestrians in time. In order to study the efficiency of additional auditory warning signals they might be equipped with, a sample of 100 sighted and 53 blind listeners was exposed to a virtual road-crossing scenario in which they had to detect whether an approaching vehicle came from the right or left. Nine warning signals, designed to differ in particular sound features such as FM, AM or the number of harmonics were studied and compared with the recording of an unfitted electrical vehicle (EV) and a conventional diesel car. The responses measured in the scenario in which cars approached at irregular intervals over two 20-min periods showed no reaction-time differences between blind and sighted participants, and a significant advantage when listening under dry weather conditions as opposed to recordings mixed with the sound of rain. Most importantly, however, regardless of listening conditions and the population studied (sighted or blind), the additional warning signals differed greatly in efficiency. Some signals facilitated detection of the EV as much as making it as noticeable as a control diesel car of significantly higher sound pressure level. Other signals were largely ineffective compared with the unfitted EV. Analysis of the signal characteristics suggested a relatively low number of harmonics, absence of frequency modulation, and irregular amplitude modulation to be the most salient features facilitating timely detection.