Service robot design for work tasks: Assessing feature importance across domestic and restaurant service contexts

BackgroundAs service robots increasingly enter diverse contexts such as homes and restaurants, users' perceptions of their appearance, interaction modalities, and functional capabilities have become critical determinants of usage intention.ObjectiveThis study develops a multidimensional classification framework for service robot features and examines how user preferences vary across contexts.MethodsA service robot feature classification framework was constructed using HIEs decomposition, a literature review, and focus group discussions. A questionnaire survey was then conducted to examine feature importance and usage intention across service contexts, with PLS-SEM and statistical analyses used for validation and contextual comparisons.ResultsThe study established a four-dimensional framework for service robot features: Morphology-Color-Material-Texture, Use Interaction, Capability Features, and Additional Features. Users showed both common and context-specific perceptions of feature importance. Context moderates the impact of feature importance on usage intention. In home contexts, the perceived importance of "information reception" capability has a positive impact, while the perceived importance of "touchscreen operation interface layout" has a negative impact. In restaurant contexts, the perceived importance of "head size", "body articulation" and "tactile signal recognition" has a positive effect, while the perceived importance of material- and color-related appearance features has a negative effect.ConclusionService robot features can be structured into four dimensions. More importantly, the effects of these features on usage intention are context-dependent. Different features play distinct roles across home and restaurant contexts. These findings highlight the importance of context-aware design strategies and provide practical guidance for optimizing service robot features across diverse environments.