As autonomous vehicles edge closer to becoming commonplace, one major inquiry persists: Will individuals genuinely place their trust in them?

A recent research from the University of Michigan indicates that the response may rest in the vehicle’s vocal tone.

When the vocalization of an automated vehicle aligned with the driver’s—especially regarding gender—individuals expressed greater trust levels. Similarity in gender notably affected both logical confidence in the vehicle’s functionality and emotional connection, whereas age resemblance was primarily influential in emotional rapport.

The outcomes hint that even a straightforward aspect like voice configuration could significantly impact how AVs are perceived, making them not solely intelligent but also relatable.

Studies indicate that embracing widespread AV use encounters considerable obstacles, primarily due to public doubts concerning their safety and operational dependability. The gender of the voice remains a contentious topic in the landscape of voice design for AI technologies. Numerous voice assistants like Siri and Alexa commonly adopt female tones, often relying on anecdotal insights suggesting that individuals from diverse cultural backgrounds typically favor female voices.

The present study, unveiled at the Proceedings of the Human Factors and Ergonomics Society yearly assembly, examined how an AV’s speech affects two forms of trust: cognitive trust (the belief in the vehicle’s intelligence and reliability) and affective trust (the feeling of an emotional connection).

Researchers conducted an online analysis involving over 300 U.S. drivers, who registered online and grasped how the self-driving vehicle functioned prior to commencement. They viewed six brief clips of the vehicle navigating various locations, accompanied by the vehicle’s voice narrating its actions. Following each clip, participants rated their trust in the vehicle and shared personal details.

When the vehicle’s voice matched both the participant’s gender and traditional gender roles, trust levels increased—both logically (cognitive trust) and emotionally (affective trust).

However, when the voice only aligned with gender and did not reflect traditional gender expectations, it primarily enhanced emotional trust but not logical trust. These results indicate a challenge: while leveraging gender roles might foster trust, it can also inadvertently reinforce stereotypes; thus, developers should contemplate alternatives such as customizable voices, gender-neutral styles, or even nonhuman sounds.

“These findings underscore the significance of considering gender in the voice design of AVs for transmitting essential driving information and illustrate how societal stereotypes influence AV development,” stated Lionel Robert Jr., professor of information and robotics at U-M.

In essence, the manner in which a self-driving car “communicates” holds significant weight, as per researchers.

“Engineering AV voices to appear more personal and relatable might enhance people’s comfort in trusting them,” remarked the study’s primary author Qiaoning (Carol) Zhang, assistant professor of human systems engineering at Arizona State University.

This research also bears significance for those uninterested in acquiring AV vehicles.

“Even if you have no intention to own a self-driving vehicle, you will almost certainly encounter them on the roads,” noted co-author X Jessie Yang, associate professor of information and industrial and operations engineering at U-M. “Fostering trust in these automobiles is pivotal for public safety and the seamless incorporation of automated vehicles into our transportation framework.”