Study Uses Augmented Reality to Test Social Robot Faces and Voices
New research from the University of Glasgow suggests that giving robot companions animal-like or anime-style faces and voices could maximize their emotional appeal. The work may help guide social robotics, a fast-growing field.
Social robots can provide companionship and comfort by mimicking human or pet behaviors. Yet surprisingly little academic research has examined how their designs can be optimized for expressiveness and emotional engagement.
In a paper to be presented at a conference this week, researchers from the university describe efforts to close that gap. With volunteers, they used augmented reality to test dozens of face-and-voice combinations overlaid on an existing social robot called Qoobo, which has a soft, cushion-like body and a swinging tail.
Dr Shaun Macdonald and Josh Yip of the School of Computing Science used this rapid prototyping method to explore how users felt about virtual robots and how well they could read the robot’s emotional states.
Dr Macdonald said social robots can provide valuable experiences for many people. Older adults have reported that they can help reduce loneliness or social isolation, and people who cannot keep a real pet have found companionship with zoomorphic robots that mimic animals’ appearance and behavior.
However, he said, the aesthetic design of zoomorphic social robots has largely relied on educated guesswork, with little research on what makes them appealing. He added that the team had taken a more systematic approach by using augmented reality for rapid, side-by-side design tests. These insights could help create user-customizable robots that tailor their appearances to owners’ preferences.
The research builds on Dr Macdonald’s previous work, including a software system called Augmenting Zoomorphic Robotics with Affect, or AZRA. AZRA can overlay augmented-reality sounds and animations on real social robots to prototype new appearances and functions.
The team created a mood board of appealing animal and robot images from real life and popular culture. They designed five prototype faces: animal-like, robot-like, emoji-style, anime-inspired, and one made simply of eyes.
They chose five vocalizations: cat noises, human-like vocal sounds, abstract electronic noises, music, and samples of “animalese,” the nonsense sounds used in Nintendo’s Animal Crossing series. Then 24 volunteers wearing AR headsets ranked different face-and-voice combinations on Qoobo for emotional clarity, empathy, and appropriateness.
Results showed that animal-like faces and vocalizations matched Qoobo’s animal-like body were well-liked and felt more engaging and immersive. Anime-like faces ranked just as highly, even though they were not based on real life.
Josh Yip said anime-like faces are more expressive than animal faces could be, letting the robot express itself beyond a normal pet without feeling uncanny. It still felt like a pet-like creature to participants.
The study also found that faces consistently outperformed sounds at conveying emotion and creating connections. Participants found human sounds easier to read emotionally but liked them less than animal sounds, even though they found animal sounds harder to identify with specific emotions.
The team was surprised that although participants said reading the robot’s emotions correctly was very important, the data showed readability had little influence on preferred designs. Similarly, “animalese” sounds ranked highly despite being unintelligible; some users felt they suggested greater intelligence.
Participants’ reactions were also affected by whether they owned a real pet. Pet owners were more likely to treat animal-like prototypes as real pets, but more likely to treat robot-like faces as machines.
Dr Macdonald added that no single face or voice has universal appeal, but the combination can greatly affect how well users feel robots communicate emotions. The findings could inform future zoomorphic social robot design and extend the lifespan of existing devices by letting users overlay new faces, voices, and functions using augmented reality.
The paper, “Anime, Animal or Animal Crossing? Comparing Aesthetic Styles for Zoomorphic Robot Faces and Sounds using Augmented Reality,” will be presented at the ACM International Conference on Mobile Human-Computer Interaction in Swansea, Wales, on Thursday, 3 September.
The project was supported by the University of Glasgow’s EPSRC Impact Acceleration Account and the Engineering and Physical Sciences Research Council. This article is republished from the University of Glasgow under a Creative Commons license.