ARIES-Mission2：用于快速大规模空中任务生成的零样本视觉-语言-动作框架
ARIES-Mission2: A Zero-Shot Vision-Language-Action Framework for Fast Large-Scale Aerial Mission Generation
浏览论文内容
中文总结 AI 辅助
ARIES-Mission2是解耦视觉语义感知与路径优化的零样本VLA框架，在UAV基准上，其飞行距离更短、任务生成速度更快，且TSP模块可扩展性佳。
中文摘要 AI 辅助
多模态大语言模型（MLLMs）已展现出强大的语义理解能力，但将其直接用于低空无人机（UAV）任务生成时，会受限于空间优化能力弱、路径规划效率低的问题。为解决该问题，本文提出ARIES-Mission2，这是一种零样本视觉-语言-动作（VLA）框架，将视觉-语义感知与物理路径优化解耦。给定自然语言指令和卫星图像后，ARIES-Mission2首先使用DeepSeek-V3进行任务解析，使用Molmo-7B进行零样本目标定位，再通过地理空间插值将检测到的像素坐标转换为GPS航路点。为减少原始视觉语言模型（VLM）生成的访问顺序导致的冗余回溯，后端将多目标无人机遍历问题建模为旅行商问题（TSP），并对比4条候选路径，包括原始VLM顺序、经PSO、GPSO、IPSO优化后的路径，最终选择最小成本闭环路径用于任务生成。在UAV-VLPA-nano-30基准上的实验显示，ARIES-Mission2的总飞行距离为62.43km，较未优化的VLA基线（79.66km）缩短21.6%，较人工规划（69.00km）缩短9.5%；完成全部30项任务的工作流耗时575.40s，单任务平均耗时19.18s，约为人类专家规划速度的3.6倍。组件级计时结果显示，VLM推理是运行时间的主要部分，单任务耗时19.02s，而TSP求解器单任务仅需0.16s；可扩展性分析进一步表明，随着目标数量增加，TSP模块的计算时间增长幅度更小。
英文摘要
Multimodal Large Language Models (MLLMs) have shown strong semantic understanding capabilities, but their direct use in low-altitude Unmanned Aerial Vehicle (UAV) mission generation remains limited by weak spatial optimization and inefficient route planning. To address this issue, we propose ARIES-Mission2, a zero-shot Vision-Language-Action (VLA) framework that decouples visual-semantic perception from physical route optimization. Given natural-language instructions and satellite imagery, ARIES-Mission2 first uses DeepSeek-V3 for task parsing and Molmo-7B for zero-shot target grounding, and then converts detected pixel locations into GPS waypoints through geospatial interpolation. To reduce the redundant backtracking caused by raw VLM-generated visiting orders, the back end formulates multi-target UAV traversal as a Traveling Salesperson Problem (TSP) and compares four candidate routes, including the raw VLM order and the routes optimized by PSO, GPSO, and IPSO. The minimum-cost closed-loop route is then selected for mission generation. Experiments on the UAV-VLPA-nano-30 benchmark show that ARIES-Mission2 achieves a total flight distance of 62.43 km, reducing the route length by 21.6% compared with the unoptimized VLA baseline (79.66 km) and by 9.5% compared with manual human planning (69.00 km). The complete 30-task workflow takes 575.40 s, averaging 19.18 s per task, which is approximately 3.6 times faster than human expert planning. Component-level timing shows that VLM inference dominates the runtime with 19.02 s per task, while the TSP solver requires only 0.16 s per task. Scalability analysis further indicates that the TSP module maintains lower growth in computation time as the number of targets increases.
