Recent Advances in Robotic Navigation via Large Language Models

摘要: Recently, with the advances in Large Language Models (LLMs), robot navigation models have demonstrated superior generalization capabilities, including environment perception, decision-making, reasoning, planning,  instruction understanding, and human-robot interaction.In this paper, we systematically review recent LLM-based robot navigation research papers, categorizing existing studies into a novel taxonomy comprising perception, planning, control, interaction, and coordination.We also present an overview of the principal datasets and metrics used in robot navigation, analyzing the distinctive characteristics of the datasets and the performance of the main LLMs-based methods.Furthermore, we discuss the challenges hindering the integration of LLMs into robot navigation and provide opportunities and potential directions for future development.