Exploring the Role of Robot's Movements for a Transparent Affective Communication

Robots operating in human-populated environments must be able to convey their intentions clearly. Displaying emotions can be an effective way for robots to express their internal state and a means to react to humans' behaviors. While facial expressions provide an immediate representation of the robot's ``feelings,'' there might be situations where only facial expressions are not enough to express the robot's intent appropriately, and multi-modal affective modalities are required. However, the characterization of the robot's movements has not been sufficiently and thoroughly investigated. In this work, we argue that transparent non-verbal behaviors, with particular attention to the robot's movements (e.g., arms, head, velocity), can be crucial for effective communication between robots and humans. We collected responses from N=967 people observing the robot during a science fair. Our results outline how movements can contribute to conveying emotions transparently. This is especially possible when no conflicting signals are present. However, facial expression is still the most dominant modality when other modalities are not aligned with the movement's intended emotion.