GSCE: A Prompt Framework with Enhanced Reasoning for Reliable LLM-driven Drone Control

• DOI: 10.48550/arxiv.2502.12531
• Publication Date: 2025-02-17
• Authors: Wenhao Wang; Yanyan Li; Long Jiao; Yuan Jiawei
• Keywords: Drone

The integration of Large Language Models (LLMs) into robotic control, including drones, has the potential to revolutionize autonomous systems. Research studies have demonstrated that LLMs can be leveraged support operations. However, when facing tasks with complex reasoning, concerns and challenges are raised about reliability solutions produced by LLMs. In this paper, we propose a prompt framework enhanced reasoning enable reliable LLM-driven control for drones. Our consists novel technical components designed using Guidelines, Skill APIs, Constraints, Examples, namely GSCE. GSCE is featured its constraint-compliant code generation. We performed thorough experiments drones wide level task complexities. experiment results demonstrate significantly improve success rates completeness compared baseline approaches, highlighting drone