Fu, Yongjie (Columbia University), Fang, Bowen (Columbia University), Liu, Mengxuan (Columbia University), Di, Xuan (Columbia University)

MetaLLMetro: LLM-Centric Virtual Reality Navigation Platform for Metro Stations

Scheduled for presentation during the Video Session "On-Demand Video Presentations" (VP-VP), Saturday, November 22, 2025, 08:00−18:00, On-Demand Platform

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on April 2, 2026

Abstract

This paper introduces MetaLLMetro, a Virtual Reality (VR) subway station system that integrates Large Language Models (LLMs) to enhance user navigation and real-time traffic information acquisition. Navigating underground subway stations remains challenging, particularly when the stations are complex and difficult to locate the correct directions or platforms. To overcome these limitations, we propose a VR platform where a fine-tuned LLM interprets users’ natural language requests, accurately infers their intended destinations, and dynamically generates navigation paths within a virtual 3D subway environment. By fine-tuning the LLM on transit-specific tasks and queries, MetaLLMetro delivers a more accurate and intuitive navigation experience, effectively bridging the gap between ambiguous user input and precise, actionable spatial directions. Furthermore, our system leverages LLM-driven queries to interface with the Google Maps API, offering users a routing assistant that provides real-time geographic context and step-by-step navigation guidance.