FLY-LLM Sim: A Novel Integration of UAV and LLM Lab Platform

Recent developments in Large Language Models (LLMs) have created new possibilities to integrate integrate these models into robotics platforms, allowing unmanned aerial vehicles (UAVs) to have enhanced control and decision-making capabilities. The UAV market has experienced significant growth in recent years, accompanied by a surge in UAV-related job opportunities. The need for LLM-powered UAVs is expected to increase significantly as LLM advances. Therefore, it is essential to provide the upcoming generation of professionals, students, and educators with the knowledge and abilities they need to develop, implement, and operate LLM-driven UAV systems efficiently. Nevertheless, there is a absence of training materials and education on LLM-powered UAVs, particularly for practical, hands-on learning. We proffer a novel Fly-Llm Sim lab platform that provides effective and efficient hands-on practice. Our lab platform comprises different simulation environments and preconfigured lab modules. The outcome of these lab modules is to educate and train users on LLM-powered UAV to perform advanced aerial route planning and perform vision-based cyber attack analysis. The lab platform will be open source, allowing other developers to customize it to their needs. Hence, our platform adopts a plug-in-based design to support the customization of the lab modules. Our evaluation results show that our platform is effective and efficient in training and educating students and professionals in operating LLM-powered UAVs via natural language commands and developing custom task-specific prompts using prompt engineering techniques.