Objective: Personality is a cognitive construct that shapes individuals’ characteristic patterns of thought, emotion, and behavior, influencing how they interact with others. Consequently, implementing personality in robots can significantly affect Human–Robot Interaction (HRI), shaping key dimensions such as trust, engagement, attention, empathy, and enjoyability. Although psychology lacks a univocal definition of human personality, several theoretical models, such as the Big Five, PEN, HEXACO, Cattell’s, and Myers–Briggs, aim to describe personality along trait dimensions. Psychological and physiological evidence further indicates that personality traits are deeply intertwined with memory processes, internal simulation, action selection, emotion generation, emotional intelligence, and action execution. However, in the robotics literature, personality implementation is often reduced to isolated behavioral aspects, typically focusing on a single trait. In contrast, approaches in affective computing attempt to capture the complexity of human personality but often remain impractical for real-world HRI applications. To address this gap, this thesis investigates the following research question: How should the personality of an artificial agent be designed to benefit Human-Robot Interaction? Accordingly, the main objective of this research is to design and implement a cognitive architecture capable of managing the interplay between personality and cognitive and affective processes, and to study how this interplay can be exploited to enhance HRI in real interaction scenarios. Approach: This research proposes a taxonomy of robotic personality based on the traits of Conscientiousness, Extraversion, and Agreeableness. A multidisciplinary approach, integrating insights from psychology, cognitive science, and robotics, was adopted to model and implement the influence of these traits on a set of cognitive and affective capabilities. Each component of the architecture was progressively evaluated through real HRI experiments. Once the overall architecture was validated, its integration was exploited in natural and challenging interaction scenarios to assess how personality-driven cognition can enhance HRI. Main results: The proposed cognitive architecture demonstrates adaptability across different robotic platforms, ranging from humanoid robots to industrial platforms, and across diverse tasks. Each architectural component proved fundamental in supporting an increasingly human-like concept of robotic personality, and consequently to enhance the intuitiveness of the interaction. Experimental results showed that personality plays a crucial role in enhancing empathy, trust, enjoyability, and sociability, key dimensions for fostering companionship with social robots. Moreover, the architecture demonstrated effectiveness in specific application domains, such as education, where it improved task performance in middle-school settings. The obtained results have been the starting point to investigate how to design robotic personality in preschool environments to support multicultural integration and collaboration. Significance: This work highlights the potential impact of robotic personality in complex and socially demanding environments. Educational applications illustrate how personality-driven robots can support learning and social inclusion. Beyond education, empathic robotic behavior holds promise in healthcare and elderly care, where long-term interaction and emotional sensitivity are essential. Finally, the integration of robotic personality and cognitive capabilities can contribute to smoother and more effective HRI in Industry 5.0 scenarios, where humans and collaborative robots are expected to share the same workspace.
From Design to Real-World Impact: Enhancing Human–Robot Interaction through Personality-Driven Robots
NARDELLI, ALICE
2026-09-16
Abstract
Objective: Personality is a cognitive construct that shapes individuals’ characteristic patterns of thought, emotion, and behavior, influencing how they interact with others. Consequently, implementing personality in robots can significantly affect Human–Robot Interaction (HRI), shaping key dimensions such as trust, engagement, attention, empathy, and enjoyability. Although psychology lacks a univocal definition of human personality, several theoretical models, such as the Big Five, PEN, HEXACO, Cattell’s, and Myers–Briggs, aim to describe personality along trait dimensions. Psychological and physiological evidence further indicates that personality traits are deeply intertwined with memory processes, internal simulation, action selection, emotion generation, emotional intelligence, and action execution. However, in the robotics literature, personality implementation is often reduced to isolated behavioral aspects, typically focusing on a single trait. In contrast, approaches in affective computing attempt to capture the complexity of human personality but often remain impractical for real-world HRI applications. To address this gap, this thesis investigates the following research question: How should the personality of an artificial agent be designed to benefit Human-Robot Interaction? Accordingly, the main objective of this research is to design and implement a cognitive architecture capable of managing the interplay between personality and cognitive and affective processes, and to study how this interplay can be exploited to enhance HRI in real interaction scenarios. Approach: This research proposes a taxonomy of robotic personality based on the traits of Conscientiousness, Extraversion, and Agreeableness. A multidisciplinary approach, integrating insights from psychology, cognitive science, and robotics, was adopted to model and implement the influence of these traits on a set of cognitive and affective capabilities. Each component of the architecture was progressively evaluated through real HRI experiments. Once the overall architecture was validated, its integration was exploited in natural and challenging interaction scenarios to assess how personality-driven cognition can enhance HRI. Main results: The proposed cognitive architecture demonstrates adaptability across different robotic platforms, ranging from humanoid robots to industrial platforms, and across diverse tasks. Each architectural component proved fundamental in supporting an increasingly human-like concept of robotic personality, and consequently to enhance the intuitiveness of the interaction. Experimental results showed that personality plays a crucial role in enhancing empathy, trust, enjoyability, and sociability, key dimensions for fostering companionship with social robots. Moreover, the architecture demonstrated effectiveness in specific application domains, such as education, where it improved task performance in middle-school settings. The obtained results have been the starting point to investigate how to design robotic personality in preschool environments to support multicultural integration and collaboration. Significance: This work highlights the potential impact of robotic personality in complex and socially demanding environments. Educational applications illustrate how personality-driven robots can support learning and social inclusion. Beyond education, empathic robotic behavior holds promise in healthcare and elderly care, where long-term interaction and emotional sensitivity are essential. Finally, the integration of robotic personality and cognitive capabilities can contribute to smoother and more effective HRI in Industry 5.0 scenarios, where humans and collaborative robots are expected to share the same workspace.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
