Abstract
The auditory experience of driving electric vehicles (EVs) has been studied as a means of conveying vehicle status information and improving user satisfaction. However, the driving experience and user preferences may vary depending on the driving context, including user characteristics. Thus, this study aims to 1) classify users based on their characteristics and 2) investigate their auditory experience preferences while driving EVs. For this purpose, 40 participants conducted questionnaires about their characteristics and performed a think-aloud task while driving 15.6 km in real EVs. As a result, three user characteristics and two user types were identified using factor analysis and K-means clustering, respectively. Text-formed think-aloud data were analyzed through network analysis to obtain insights for designing usercentered driving sound for EVs. These findings can contribute to the strategic management of EV sound design.
In this research, the auditory experience of electric vehicles (EVs) was studied considering various driving contexts and user characteristics (Kano, 1984; Bera & Maitra, 2021; Song, 2021; Kim, 2017). The auditory environment of EVs was examined, with emphasis on "driving sound," a comprehensive term for the various sounds produced by a vehicle during operation. Prior studies have used affective engineering approaches to evaluate auditory quality (Park et al., 2019), however, these did not fully account for diverse and detailed driving contexts. Thus, this research aimed to categorize user types based on their characteristics and derive user needs, evaluations, and ideas by user types, considering different driving speeds, road types, and user experiences with EVs (Maiberger et al., 2018; Gillibrand et al., 2011; Maguire, 2001; Harvey et al., 2011; Lee et al., 2016; Kim et al., 2013; Takada et al., 2019).
A real driving experiment involving 40 participants (27 males and 13 females) was conducted using actual EVs. Participants included experienced and inexperienced EV drivers. The experiment was divided into stages: preparation, practice driving, main experiment, and post-experiment questionnaire. Participants were asked to drive at different speeds and voice their thoughts aloud, focusing on their auditory experience during the drive (Maiberger et al., 2019).
Data was collected through questionnaires about personal driving style, opinions about EV engine sounds and noises, and preference for sound. During the drive, participants were instructed to think aloud, commenting on the sounds while driving the EVs. Recorded think-aloud data was converted to text for analysis. Data analysis was performed using factor analysis, k-means cluster analysis, ANOVA, exploratory factor analysis, Mann-Whitney's U test, and network analysis (Maiberger et al., 2019).
The methodology used here offers an immersive, real-world approach to understanding user perceptions of EV auditory experience. It combines quantitative and qualitative analyses to capture the complexity of driving contexts and user characteristics. The results from this study will potentially enhance the design of auditory experiences in EVs, thereby improving user satisfaction and safety. It underscores the need for future research in this area to account for a more comprehensive view of the driving context and its influence on user experience (Maiberger et al., 2018; Gillibrand et al., 2011; Maguire, 2001; Harvey et al., 2011; Lee et al., 2016; Kim et al., 2013; Takada et al., 2019).
This research aims to provide a user-centered approach in designing and evaluating auditory experiences in electric vehicles (EVs), considering user characteristics and their expectations. Factor analysis resulted in three significant user characteristics: driving style, sensitivity to noise, and acceptability of noise, with driving style and sensitivity to noise showing significant variances among users. K-means clustering further identified two main user types: Enthusiastic Driving & Sensitive to Noise (ES) type, and Safe Driving & Generous to Noise (SG) type.
The study also identifies four affective properties of EV engine sound: Sporty, Stylish, Comfort, and Calm. Each property was classified based on Eigenvalue and Cronbach's alpha values, ensuring their reliability. Differences in sound preferences were noted between the two user types, with the ES type showing a higher preference for "Sporty" and "Stylish" sounds compared to the SG type. However, no significant difference was observed in preferences for "Comfort" and "Calm" sounds.
Users' needs and ideas regarding EV sounds were gathered through a think-aloud task. The data revealed that both user types shared common needs and ideas about EV sound. However, the ES type viewed it more as an entertainment factor, while the SG type considered sound as an indicator of the car's condition. Moreover, both types showed a need for a sound mode selection feature.
The results underline the importance of considering user characteristics and preferences in designing EV sounds. Emphasizing the "Comfort" aspect could appeal to all users, while specific sound designs could be made more "Sporty" or "Stylish" for targeted user types. The study paves the way for creating user-centered auditory experiences in EVs, catering to various user types and expectations during different driving contexts. Future research could delve into exploring the relationship between the physical features of sounds and affective variables for a more comprehensive understanding of EV sound design.
In conclusion, the study offers a valuable contribution to the field by suggesting systematic and logical strategies for designing EV sounds based on user types and preferences. The research serves as a basis for automobile companies to design user-centered and customized sounds targeting specific user types, ultimately leading to a safe and satisfactory driving experience. Further investigation into the correlation between sound physical features and affective variables is proposed as a future research avenue.