# ScaleNav uses internet video training to guide robot navigation

_Published Tuesday, October 20, 2026 at 12:05 AM EDT · Science, Robotics · Latest · Tier 2 — Notable_

Researchers report that ScaleNav, a robot navigation system pretrained on internet videos, outperformed evaluated baselines on two navigation benchmarks and was tested on a real quadruped robot. The preprint reports success rates of 86.5% on a single-view adaptation of Short-Horizon OVON and 94.9% on the researchers' HabitatGS-LN benchmark.

The system learns from first-person videos how a view changes when approaching a target described in text. It then learns executable actions from labeled simulated data before joint fine-tuning. Its task is limited to reaching and stopping at a text-specified target visible in the robot's initial observation; the reported benchmark percentages do not describe its real-robot performance.

## Sources

- [arXiv Query: search_query=cat:cs.RO&id_list=&start=0&max_results=30](https://arxiv.org/abs/2610.25690v1)

---
Canonical: https://techandbusiness.org/newswire/wUQNtq4LAxANu9AWaI88Wk
Published: 2026-10-20T04:05:34.164Z
Story chronology: 2026-10-19T02:37:24.000Z
Retrieved: 2026-10-20T07:24:49.582Z
Publisher: Tech & Business (techandbusiness.org)