About this book:

This book provides a comprehensive framework for developing cognitive architectures specifically tailored to social robots intended for long-term human interaction. It argues that for a robot to move beyond a simple tool and become a persistent social partner, it must integrate symbolic reasoning (for norms and logic) with subsymbolic neural networks (for perception and pattern recognition). By creating a hybrid system, designers can equip robots with the necessary infrastructure to manage the complexities of human relationships, including the maintenance of episodic and semantic memory, the regulation of affective states, and the presentation of a consistent yet adaptive personality.

The text details the essential components of these "social minds," emphasizing the importance of modeling the self, the user, and shared intentions. It covers the technical requirements for social perception—such as gaze, gesture, and prosody—and explains how dialogue management must incorporate pragmatics and grounding to ensure mutual understanding. Central to the book’s thesis is the role of learning; it explores how robots can use reinforcement, imitation, and self-supervision to evolve alongside their users. This continual learning is presented as the bedrock of personalization, allowing the robot to adjust its behavior to individual quirks and group dynamics while strictly adhering to privacy and ethical standards.

Beyond architectural design, the book provides a roadmap for the practical deployment and evaluation of social robots in "the wild," such as homes and classrooms. It introduces specialized metrics for assessing social intelligence and relationship quality, moving beyond task efficiency to measure trust, rapport, and psychological impact. The final chapters address the critical engineering challenges of robustness and failure recovery, stressing that a robot’s ability to transparently explain its errors is vital for maintaining human trust. Ultimately, the work calls for standardized tooling and reproducible methodologies to transition social robotics from isolated laboratory experiments to reliable, life-enhancing technologies.