Robots are envisioned as a new breed of embodied computer interface that will provide intellectual, social, and physical assistance in people’s daily lives as information booth attendants, museum guides, shopkeepers, storytellers, educational assistants, companions for children and the elderly, and so on. These tasks require robots to interact socially as well as physically—to understand and respond appropriately to human social behaviors, to effectively employ social and communicative cues, and to be aware and responsive to different social norms and contexts.
The social sciences have studied human social behavior and interaction extensively, but transferring this knowledge to the design of robots poses novel challenges for designers and researchers. As roboticists, how do we design social skills into robots? What behaviors are important or appropriate for different tasks and contexts? How can we use social-scientific theories as the basis for designing these behaviors? How can empirical research aid us in designing behavioral mechanisms for robots and assessing their effectiveness in interaction?
The goal of this proposed tutorial is to present participants with a novel framework, which draws on principles and methods from the social sciences and interaction design, and its associated methods for designing social behavior for robots. This framework combines top-down, predictive models of social behavior from social-scientific theory and bottom-up design variables extracted from empirical data using cognitive and computational modeling.
There will be no prerequisites for attending this tutorial; all members of the robotics community are welcome. This tutorial may particularly appeal to robotics researchers interested in developing real-world applications for robots, as well as scholars who want to evaluate and study the social aspects of human interactions with robots. The tutorial will provide conceptual content from social research as well as hands-on methods for their application to robot design.
As part of the documentation of the tutorial, we will prepare a “participant kit” that includes the materials the participants will use during the tutorial, an overview of key concepts in social-scientific theory and interaction design, and further readings on the materials covered in the tutorial. We also plan to create, maintain, and update a tutorial website that includes the participant kit and serves as a future resource for designing social behavior for robots.