UAV dispatch systems gain new semantic control from large language models
LLM-Enabled UAV Dispatch: A System-Level Survey and Taxonomy
Artificial Intelligence
Summary
Scheduling and controlling fleets of drones is complicated, because it involves lots of planning and communication. The authors review how large language models (LLMs), like AI systems that understand human language, can be used as a smart middle layer that helps translate human instructions into drone actions and coordinates multiple drones safely. They group existing drone dispatch methods into four types and explain how LLMs contribute to each. They also discuss challenges like making sure the system works quickly and reliably in real-world conditions.
What this means in practice
- •For drone fleet operators: Design UAV dispatch systems that integrate language-based AI to better understand human task instructions and coordinate multiple drones safely.
- •For autonomous vehicle developers: Incorporate language-based semantic layers to improve task planning and coordination among autonomous agents beyond just drones.
A survey. It maps existing work.
Authors
Xiao Han, Aoyang Quan, Xiangyu Zhao, Xiangjie Kong, Guojiang Shen
Abstract
Unmanned aerial vehicle (UAV) dispatch is beginning to move beyond isolated path planning and optimization-driven resource allocation toward system-level coordination supported by semantic reasoning and LLM-based interfaces. This survey provides a unified characterization of LLM-enabled UAV dispatch systems that bridges semantic intent, symbolic decision-making, and physical UAV execution. Rather than treating LLMs as standalone add-ons, we conceptualize them as a cross-layer semantic orchestration layer connecting human instructions, external solvers, and distributed control modules. We organize the literature into four representative dispatch paradigms: pipeline dispatch, global assignment dispatch, decentralized agentic dispatch, and divide-and-conquer dispatch. For each paradigm, we analyze its decision logic, system structure, control flow, representative methods, and potential LLM roles. We further examine how LLMs support semantic parsing, retrieval-grounded planning, solver orchestration, local agent reasoning, multi-agent coordination, safety assessment, and human-facing explanation. We discuss the implications of these paradigms for scalability, robustness, coordination burden, and verification requirements, and identify open challenges including latency-aware reasoning, grounding reliability, physical feasibility guarantees, edge deployment, privacy protection, and distributed consistency. This survey provides a system-level taxonomy and design perspective for integrating LLMs into safety-critical UAV dispatch systems.