Background: With global population aging, social robots are increasingly valued for providing emotional companionship to the elderly. However, most current products still lack warmth and emotional care, especially in recognizing implicit cues like ton...
Background: With global population aging, social robots are increasingly valued for providing emotional companionship to the elderly. However, most current products still lack warmth and emotional care, especially in recognizing implicit cues like tone or silence. This limits emotional connection and weakens the user experience.
Methods: Focusing on elderly users, this study used surveys, user journey maps, and in-depth interviews to identify key emotional needs and design pain points. Based on the findings, an optimized solution was developed integrating four elements: task predictability, clear operation paths, emotional care prompts, and process tolerance. A design experiment was conducted for initial validation.
Results: The optimized interaction model actively recognizes and responds to users' concerns, offering psychological support. Its human-like responses shift the interaction from task-based to relationship-based, making users feel understood and cared for, and enhancing trust and emotional connection.
Conclusion: This study proposes an optimized Emotional Language Interaction Process integrating four core elements, using the Xiaodu Health Smart Screen as a case. Initial validation showed positive user feedback and high trust. Results demonstrate that combining multimodal emotion recognition and large language models effectively addresses current limitations in emotional care, offering a practical path to enhance elderly users’ long-term engagement and overall experience.