Designing Facial Expressions for Service Robots: A Context-Specific Approach

Abstract

While facial expressions are widely adopted in service robots as rendered faces on their screens to facilitate communication and interaction, it remains unclear how they should be designed according to specific service contexts. This paper argues that the appropriate level of realism and detail in robot facial expressions depends on a robot's task nature, level of social interaction, and service journey. We propose a design space constructed from two axes, realism and number of facial features, and map 24 screen-based service robot faces onto it. The mapping of restaurant, retail, and delivery robot faces reveals a lack of context-specific design patterns and an over-reliance on minimal expressions, with 75% of collected faces lacking a mouth. Drawing on service marketing literature, we construct a service journey map for restaurant server robots and identify emotions required along the service flow. These analyses expose a gap between research and practice, leading to a structured, three-stage design process for context-specific robot facial expressions.

Keywords: Human-robot Interaction, Service Robots, Facial Expressions, Interface Design

DOI: 10.54941/ahfe1007755