# Preprint reports world-model locomotion on Unitree G1

_Published Thursday, September 3, 2026 at 2:06 AM EDT · Science, Robotics · Latest · Tier 2 — Notable_

A robotics preprint reports that a world-model-assisted locomotion policy was deployed on a physical Unitree G1 humanoid using onboard proprioception and one depth stream. The authors say the robot traversed gaps, stepping stones, and stairs with an average success rate of 93.3%. The system jointly trains a recurrent world model and PPO policy so a predictive feature can guide foot placement without explicit foothold labels.

## Sources

- [arXiv Query: search_query=cat:cs.RO&id_list=&start=0&max_results=30](https://arxiv.org/abs/2609.02542v1)

---
Canonical: https://techandbusiness.org/newswire/RQT2vZ6asMmGX1_LafvYv5
Published: 2026-09-03T06:06:06.561Z
Story chronology: 2026-09-02T12:57:01.000Z
Retrieved: 2026-10-18T11:01:26.619Z
Publisher: Tech & Business (techandbusiness.org)