Exploring the Social Perception of a Robotic Arm with Personality

Abstract

Robotic personality plays a key role in shaping interactions between humans and social robots. However, the effects of implementing synthetic personalities in industrial, non-humanoid robots remain largely unexplored, despite evidence that social cues and cognitive capabilities can enhance mutual understanding between industrial robots and human co-workers—an essential aspect for the co-existence paradigm of Industry 5.0. This work focuses on implementing artificial personalities in an industrial manipulator. In the proposed approach, personality forms the core of a cognitive architecture, influencing action selection, execution, and the robot's responses to environmental stimuli. The implementation was validated using a Kinova Jaco2 arm with human participants engaged in a collaborative task. The task was conducted under two between-subject conditions: a system that interacted solely through movements and a system that was additionally capable of verbal communication. Within each condition, a partially within-subject factor was included, consisting of different combinations of robot personality traits. Specifically, personality profiles varying in Conscientiousness, Extroversion, and Agreeableness, based on the Big Five personality traits model, were tested. Results from questionnaire analyses indicate that users accurately perceive the social norms associated with different personalities. When verbal communication is employed, these distinctions are even more pronounced. Furthermore, users' behavioral data suggest that verbal interaction may lead to higher engagement while reducing cognitive load and visual attention demands.