TY - GEN
T1 - Design and Feasibility of an LLM-Powered Humanoid Robot as a Reading Companion to Support Children’s AI Literacy
AU - Xiao, Feiwen
AU - Li, Zhaohui
AU - Jiang, Shiyan
AU - Xiong, Jinjun
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.
PY - 2026
Y1 - 2026
N2 - Artificial intelligence (AI) is increasingly prevalent in children’s everyday experiences, creating a growing need for developmentally appropriate AI literacy learning opportunities. Recent advances in large language models and social robotics enable new possibilities for embodied, interactive learning. We present an LLM-powered humanoid social robot as a reading companion for AI literacy books for children aged 5–8, integrating speech, vision, language generation, and gesture to support multimodal interaction. We implement and compare three interaction policies (i.e., heuristic, dynamic, and guided) representing different levels of control in LLM-mediated dialogue. Through controlled interaction evaluation, we demonstrate the feasibility of deploying this system on real hardware and identify a key insight: a structured interaction framework with bounded LLM support yields more robust and pedagogically coherent interactions than fully scripted or open-ended approaches. These findings highlight design directions for safe and effective embodied AI learning companions for young children.
AB - Artificial intelligence (AI) is increasingly prevalent in children’s everyday experiences, creating a growing need for developmentally appropriate AI literacy learning opportunities. Recent advances in large language models and social robotics enable new possibilities for embodied, interactive learning. We present an LLM-powered humanoid social robot as a reading companion for AI literacy books for children aged 5–8, integrating speech, vision, language generation, and gesture to support multimodal interaction. We implement and compare three interaction policies (i.e., heuristic, dynamic, and guided) representing different levels of control in LLM-mediated dialogue. Through controlled interaction evaluation, we demonstrate the feasibility of deploying this system on real hardware and identify a key insight: a structured interaction framework with bounded LLM support yields more robust and pedagogically coherent interactions than fully scripted or open-ended approaches. These findings highlight design directions for safe and effective embodied AI learning companions for young children.
KW - AI Literacy
KW - Reading Companion
KW - Social Robot
UR - https://www.scopus.com/pages/publications/105043930056
U2 - 10.1007/978-3-032-29788-4_71
DO - 10.1007/978-3-032-29788-4_71
M3 - Conference contribution
AN - SCOPUS:105043930056
SN - 9783032297877
T3 - Communications in Computer and Information Science
SP - 498
EP - 504
BT - Artificial Intelligence in Education. Late Breaking Results, WideAIED, Practitioners, Industry and Policies, Blue Sky, Doctoral Consortium, FoL Workshops and Tutorials, FoL Invited Papers - 27th International Conference, AIED 2026, Proceedings
A2 - Blanchard, Emmanuel G.
A2 - Chen, Guanliang
A2 - Chi, Min
A2 - Isotani, Seiji
PB - Springer Science and Business Media Deutschland GmbH
T2 - 27th International Conference on Artificial Intelligence in Education, AIED 2026
Y2 - 27 June 2026 through 3 July 2026
ER -