Using virtual environments as a tool to evaluate electric vehicle sounds

• Published in: 5th IET Hybrid and Electric Vehicles Conference (HEVC 2014) p. 4.1
• Main Authors: Singh, S, Payne, S.R, Mackrill, J.B, Jennings, P.A
• Format: Conference Proceeding
• Language: English
• Published: Stevenage, UK IET 2014; The Institution of Engineering & Technology
• Subjects: Electrical engineering computing; Traffic engineering computing; Transportation; Virtual reality; electrical engineering computing; virtual reality; electric vehicles; electric vehicle exterior sound evaluation; town T-junction; pedestrians; car arrival time; emotional evaluation; subjective detectability; residential crossing; car detectability; electric car; safety; participant feedback; pedestrian safety; listening evaluations; virtual environments; legislations; legislation
• ISBN: 9781849199117; 1849199116
• DOI: 10.1049/cp.2014.0948

Legislations mandate that electric vehicles should emit sounds to alert pedestrians of the vehicle's approach to ensure pedestrians' safety. Additionally, manufacturers want these sounds to promote positive impressions of the vehicle brand. A reliable and valid methodology is needed to evaluate electric vehicles' exterior sounds on the criteria of safety and brand. This paper summarizes the framework and findings of a research project aimed at developing such a methodology. The project firstly examined literature. This helped propose a methodology for evaluating an electric vehicle's exterior sounds which tests its detectability and emotional evaluation in a virtual environment. Experiment 1 tested the methodology. Thirty-one participants detected an electric car, emitting 15 sounds, and evaluated the car as powerful and pleasant in a virtual environment depicting a town T-junction. The car's arrival time, direction of approach, and thus distance to pedestrian, varied across conditions. The vehicle's arrival time affected its detectability and powerfulness evaluation. Similarly, its direction of approach, while travelling at different distances to the pedestrian, affected the powerfulness evaluation. Thus, these are important elements to manipulate during listening evaluations. The methodology was improved using the results and participants' feedback. Experiment 2 tested the external validity of the revised methodology. Fourteen participants evaluated an electric car, emitting 3 sounds, in both the real-world and the virtual environment of a residential crossing. The car's detectability, emotional evaluations and the sounds' recognisability as a vehicle were measured. No significant difference was found in subjective detectability and emotional evaluation in the virtual and realworld. Thus, these results may be generalised to a similar real-world situation. Overall the methodology increases the realistic context and experimental control compared to existing listening evaluations. It benefits by combining two competing elements necessary for assessing electric vehicle exterior sounds, namely pedestrians' safety and vehicle brand.