Abstract
Amid the integration of AI into special education, establishing trust between AI educational robots and children with autism through emotional design is a critical challenge in HCI. While emotional design benefits general educational robots, systematic frameworks and empirical evidence for autistic children remain scarce. This study applies the three-level emotional design model (Instinctive, behavioral, reflective) to examine how multimodal elements—voice (volume, speed), appearance (color, form), and emotional expressions—affect attention, interaction willingness, and trust. A controlled experiment (N=12) revealed that higher volume, slower speech, warm colors, and positive expressions significantly improve engagement and trust. Based on the results, a design evaluation index hierarchy model was developed, incorporating perceptual adaptation, behavioral guidance, and emotional resonance. This model offers theoretical and practical guidance for developing trust-building AI robots, supporting inclusive, human-AI collaborative educational environments.
Keywords
AI in Special Education，Emotional Design，Trust Building，ASD
DOI
https://doi.org/10.21606/drs.2026.1680
Citation
Fang, S., Yu, H., and Yuan, X. (2026) Design for Trust: A Study on the Emotional Design Method of AI Educational Robots for Children with Autism, in Simeone, L., Gray, C. M., Verhoeven, A., de Götzen, A., Bakırlıoğlu, Y., Zohar, H., Stead, M., and Buwert, P. (eds.), DRS2026: Edinburgh, 8–12 June, Edinburgh, United Kingdom. https://doi.org/10.21606/drs.2026.1680
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Included in
Design for Trust: A Study on the Emotional Design Method of AI Educational Robots for Children with Autism
Amid the integration of AI into special education, establishing trust between AI educational robots and children with autism through emotional design is a critical challenge in HCI. While emotional design benefits general educational robots, systematic frameworks and empirical evidence for autistic children remain scarce. This study applies the three-level emotional design model (Instinctive, behavioral, reflective) to examine how multimodal elements—voice (volume, speed), appearance (color, form), and emotional expressions—affect attention, interaction willingness, and trust. A controlled experiment (N=12) revealed that higher volume, slower speech, warm colors, and positive expressions significantly improve engagement and trust. Based on the results, a design evaluation index hierarchy model was developed, incorporating perceptual adaptation, behavioral guidance, and emotional resonance. This model offers theoretical and practical guidance for developing trust-building AI robots, supporting inclusive, human-AI collaborative educational environments.