Predicting Detectability and Annoyance of EV Warning Sounds using Partial Loudness

At low speeds, electric vehicles (EV) emit less noise compared to internal combustion engine vehicles, making them more difficult for other road users to detect. In this study, the use of a computational partial loudness model for predicting detection and perceived annoyance of EV warning sounds was investigated and compared with the signal-to-noise ratio (SNR). In two experiments, detection thresholds were obtained for 4 different warning sounds in 5 different background noises. In a third experiment, subjective perceived annoyance ratings were obtained for 4 warning sounds in 5 noise conditions. For all experiments, the partial loudness and SNR measures were compared with the listener data. Overall, the detection thresholds in terms of partial loudness were similar for stationary warning sounds. This was not the case with SNR. The perceived annoyance increased with partial loudness as expected and the slope of the increase was similar across different warning sounds in different noise conditions. This was not the case with SNR. Thus, a partial loudness model is a better objective method to predict detection and annoyance of stationary warning sounds. However, more work is needed to improve the prediction of detecting non-stationary warning sounds.