Developing a Hierarchically Structured Action Catalogue for an Elderly Care Robot through Participatory Design

As companion robots become more and more common in our everyday life, the ability to perform contextually appropriate and emotionally meaningful actions becomes essential. Unlike humanoid robots, many service robots rely on constrained forms of expression, minimalist movement, gestures, or visual signals to communicate with intent. Designing such behaviors requires not only technological precision but also deep empathy and user-centered insight. Therefore, actions that can accurately reflect robot’s behavior should be carefully extracte d 1. Lemmy, the focus of this study, is a non-humanoid, twowheeled companion robot intended to support elderly individuals within home environments (Figure 1). Lemmy's design emphasizes subtle emotional expression and gentle behavioral cues through movement, screen-based animations, and sound. It is designed to provide reminders, emotional support, and assistive presence, promoting wellbeing without relying on overtly human-like traits. This paper presents the development of Lemmy’s action catalogue through a structured, participatory approach. Rather than engineering a predefined set of actions, we engaged designers in co-creating an intuitive and emotionally rich action catalogue grounded in user stories. The result of the overall process was not only a working foundation for Lemmy’s interaction design but also a transferable methodology for deriving behavioral actions of socially assistive robots.