ARIES-Mission2: A Zero-Shot Vision-Language-Action Framework for Fast Large-Scale Aerial Mission Generation

2026-08-12

AI Summary

This work addresses the inefficiency of existing multimodal large models in spatial perception and path planning for low-altitude UAV mission generation. The authors propose ARIES-Mission2, a novel framework that achieves zero-shot, end-to-end vision-language-action task generation for the first time. Its front end integrates DeepSeek-V3 and Molmo-7B to perform zero-shot target localization and output GPS waypoints, while the back end formulates multi-target traversal as a Traveling Salesman Problem (TSP) and employs PSO, GPSO, and IPSO algorithms to generate an optimized closed-loop route. Evaluated on the UAV-VLPA-nano-30 benchmark, the system achieves a total trajectory length of 62.43 km—21.6% shorter than an unoptimized VLA baseline and 9.5% better than manual planning—with an average task execution time of 19.18 seconds, approximately 3.6× faster than human operators. The TSP module exhibits low computational overhead and strong scalability.

Abstract

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.

Problem

• UAV mission generation
• Vision-Language-Action
• spatial optimization
• route planning
• zero-shot

Innovation

• Zero-shot Vision-Language-Action
• UAV mission planning
• Traveling Salesperson Problem
• Geospatial interpolation
• Multimodal Large Language Models