Can My Comfort Reflect Your Preferences? An Exploratory Study on Comfort-Driven Architecture in Human-Robot Interaction*

Developing robots that behave in a more humanlike manner requires endowing them with the capability to perceive, interpret, and appropriately respond to human actions and intentions. To achieve this goal, we propose a cognitive comfort-driven architecture inspired by a developmental approach and the theory of attachment. Our framework equips robots with basic cognitive capabilities and a motivational system based on comfort that drives their behavior. Using a mixed-methods approach, we evaluate our architecture by exploring how participants interacted with two different robot profiles inspired by human attachment styles and how their behavior evolved over time. Our results suggest that comfort motivation may indicate how well participants could interpret and respond to the robot’s implicit demands. This research offers new insights into the role of social motivation in influencing human-robot interaction. By equipping robots with cognitive models inspired by human brain processes, we can improve their social understanding and create more natural and intuitive human-robot interactions.