MultiUAV-Plat: An LLM-Oriented Platform, Benchmark and Framework for Multi-UAV Collaborative Task Planning

2026-06-29

AI Summary

This work addresses the absence of evaluation platforms tailored for large language models (LLMs) in multi-drone cooperative mission planning, where challenges such as partial observability, spatial coverage, and multi-agent coordination are critical yet underexplored. To bridge this gap, we propose Agent4Drone—a structured framework that integrates LLM-based agents with multi-drone collaborative planning through modular components encompassing memory, observation, comprehension, planning, execution, and verification. We also develop a lightweight simulation environment supporting RESTful APIs, role-based information access, and real-world tool interaction. Accompanying this framework is a benchmark dataset comprising 75 task sessions and 1,500 natural language instructions. Experimental results demonstrate that Agent4Drone achieves a task success rate of 57.9%, substantially outperforming the ReAct baseline, and reduces the failure rate from 32.4% to 12.9%, establishing the first standardized evaluation protocol for this emerging domain.

Abstract

Large language models (LLMs) provide a promising interface for high-level robotic task planning, but their use in multi-UAV collaboration remains difficult to evaluate systematically. Existing UAV simulators mainly emphasize dynamics, perception, or low-level control, while existing LLM-agent benchmarks rarely capture aerial-robotics constraints such as partial observability, spatial coverage, UAV assignment, and multi-vehicle coordination. To bridge this gap, we present MultiUAV-Plat, a lightweight, easy-to-use, LLM-agent-oriented simulation platform for multi-UAV collaborative task planning. The platform exposes concise RESTful APIs, agent-facing observations, role-based information access, hidden validation logic, and optional 2D/3D visualization, allowing agents to solve missions through realistic tool interaction rather than privileged simulator access. Built on this platform, the MultiUAV-Plat Benchmark contains 75 mission sessions, 1500 natural-language tasks, and 9396 validation checks across target assignment, area search, and area assignment and patrol scenarios. We further propose Agent4Drone, a task-specific LLM agent framework that structures multi-UAV behavior into memory, observation, task understanding, planning, execution, and verification. In a full paired benchmark comparison, Agent4Drone achieves a 57.9% task pass rate, a 74.6% average task check pass rate, and a 72.0% global check pass rate, substantially outperforming a ReAct baseline at 30.6%, 47.9%, and 43.1%, respectively. Agent4Drone also reduces the total failed task rate from 32.4% to 12.9%. These results demonstrate that MultiUAV-Plat and MultiUAV-Plat Benchmark provide a reproducible foundation for studying LLM-driven multi-UAV autonomy under realistic information and execution constraints.

Problem

• multi-UAV collaboration
• LLM-based task planning
• aerial robotics constraints
• benchmark evaluation
• partial observability

Innovation

• LLM-agent
• multi-UAV collaboration
• task planning
• simulation platform
• benchmark