Abstract
The digital transformation of tourism, while enhancing efficiency and marketing, faces persistent challenges in delivering truly immersive and authentic remote experiences. While immersive technologies like VR mitigate temporal-spatial constraints, current implementations often lack real-time intelligence, personalized co-creation, and deep contextual understanding. This study introduces the Intelligent Travel Avatar, an embodied AI paradigm that combines an advanced LLM with a quadrupedal robot, which integrates physical environment perception, real-time data processing, and multimodal interaction to deliver context-aware, personalized guidance through natural language dialogue. Through the combination of embodied robotics and LLM-driven intelligence, this study offers both theoretical contributions and practical insights into the design of AI-powered tourism agents, while extending the application of GenAI in the tourism domain.
Recommended Citation
Xu, Tao; Gao, Manni; Liu, XuDong; Ye, Qiang; Ma, Lin; and Zhu, Bingyu, "Intelligent Travel Avatar: An LLM-based Tourism Quadrupedal Robot" (2025). ICEB 2025 Proceedings (Hanoi, Vietnam). 48.
https://aisel.aisnet.org/iceb2025/48