# SatNav: A Scalable Benchmark for Long-Horizon UAV Vision-Language Navigation from Satellite Imagery

arXiv: https://arxiv.org/abs/2609.31507 · HTML: https://mlsanity.com/paper/2609.31507/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 43 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Builds on (most cited first, 9 of 9)

- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [Gemini: A Family of Highly Capable Multimodal Models](https://mlsanity.com/paper/2312.11805/) (2023)
- [StreamVLN: Streaming Vision-and-Language Navigation via SlowFast Context Modeling](https://mlsanity.com/paper/2507.05240/) (2025)
- [UAV-VisLoc: A Large-scale Dataset for UAV Visual Localization](https://mlsanity.com/paper/2405.11936/) (2024)
- [VLNVerse: A Benchmark for Vision-Language Navigation with Versatile, Embodied, Realistic Simulation and Evaluation](https://mlsanity.com/paper/2512.19021/) (2025)
- [LongFly: Long-Horizon UAV Vision-and-Language Navigation with Spatiotemporal Context Integration](https://mlsanity.com/paper/2512.22010/) (2025)
- [AirNav: A Large-Scale UAV Vision-and-Language Navigation Dataset with Natural and Diverse Instructions](https://mlsanity.com/paper/2601.03707/) (2026)
- [OnFly: Onboard Zero-Shot Aerial Vision-Language Navigation toward Safety and Efficiency](https://mlsanity.com/paper/2603.10682/) (2026)
- [HUGE-Bench: A Benchmark for High-Level UAV Vision-Language-Action Tasks](https://mlsanity.com/paper/2603.19822/) (2026)

## Abstract (arXiv)

Urban uncrewed aerial vehicle (UAV) vision-language navigation (VLN) requires agents to follow instructions across extended urban spaces, inherently demanding long-term memory and geospatial grounding. However, scaling existing benchmarks remains difficult because of their reliance on costly reconstructed 3D assets, limiting geographic diversity and episode scale. To address this, we introduce SatNav, a scalable, long-horizon UAV VLN benchmark built from high-resolution satellite imagery. SatNav targets city-level navigation missions and uses satellite crops as approximations of UAV nadir views for visual observations. Through an automated cue-to-episode pipeline, SatNav constructs 118K episodes from 59 scenes across 18 cities, with an average trajectory length of 379 m. To stress-test long-horizon memory and geospatial reasoning, SatNav defines three task families: Boundary, Landmark, and Route, targeting loop progress tracking, landmark-based spatial grounding, and route following with counting cues. Benchmarking classical VLN agents and recent agents based on large vision-language models (LVLMs) on SatNav shows that city-scale navigation remains challenging. We further introduce SwiftVLN, a modular framework with switchable memory components, and conduct systematic memory-design ablations. Finally, satellite-to-UAV transfer experiments show that satellite-trained navigation models can operate on real-flight UAV observations, showing the practical relevance of SatNav. Our project page: https://eku127.github.io/SatNav/