Robostral Navigate

2026-07-22

AI Summary

This work proposes a monocular vision-language navigation model with 8 billion parameters that operates solely on RGB images, eliminating the need for depth sensors, multi-camera setups, or pre-built maps—thereby enhancing hardware compatibility and reducing deployment costs. By directly predicting goal locations in image space, the model enables a unified navigation policy across wheeled, legged, and aerial robots and, for the first time, supports deployment without camera calibration. The approach leverages 2.4 million simulation-generated trajectories, prefix-cached sequence packing, tree-structured attention masking, and reinforcement learning to significantly improve training efficiency, visual grounding, and recovery from exploration failures. It achieves state-of-the-art success rates of 77.4% on R2R-CE and 75.1% on RxR-CE, outperforming existing monocular methods and even some multi-sensor systems.

Abstract

Deploying navigation systems at scale requires a recipe that minimizes sensor assumptions, generalizes across robot embodiments, and trains efficiently. Yet, today's best systems depend on depth sensors, multi-camera rigs, or pre-built maps, limiting the hardware they support and increasing deployment cost. We introduce Robostral Navigate, an 8B vision-language model built around this scalability objective. The model consumes only a stream of monocular RGB images - the most ubiquitous sensor across robotic platforms and predicts waypoints by pointing to the next target location in the current camera view. Operating purely in image space, rather than robot-specific coordinates, makes the policy naturally robust to changes in camera intrinsics and scene scale, enabling deployment across wheeled, legged, and aerial robots without recalibration. We generate 2.4 million trajectories across 350k simulated scenes to reduce the reliance on real-world data collection and scale easily. We further introduce a prefix-caching training recipe that packs entire episodes into single training sequences, reducing training tokens by 22x and cutting training time from months to days. A tree-based attention mask prevents conditioning on previous ground-truth actions, encouraging visually grounded action prediction, and reinforcement learning is used to further improve exploration and recovery capabilities. On the Room-to-Room and Room-Across-Room in Continuous Environments (R2R-CE and RxR-CE) benchmarks, Robostral Navigate sets a new state of the art. On R2R-CE, it achieves a 77.4% success rate, surpassing the best monocular method by 10.5 points and the strongest depth- or multi-camera system by 5.3 points despite using only a single RGB camera. On RxR-CE, it reaches 75.1% success rate, outperforming all monocular baselines.

Problem

• robot navigation
• scalability
• sensor assumptions
• embodiment generalization
• monocular vision

Innovation

• monocular navigation
• vision-language model
• prefix-caching training
• image-space waypoint prediction
• cross-embodiment generalization