# Preprint demonstrates differentiable learning for quadruped obstacle avoidance

_Published Wednesday, October 14, 2026 at 10:07 PM EDT · Robotics, Science · Latest · Tier 2 — Notable_

Researchers report deploying an end-to-end differentiable reinforcement learning system for three-dimensional obstacle avoidance on a Unitree Go2 robot in static and dynamic environments. Their arXiv preprint presents this as the first transfer of that approach from simulation to a physical robot for this task.

The system learns obstacle interactions using soft penetration penalties, then transfers a simulated teacher's behavior to a controller using depth observations. A terrain-aware module blends it with a base walking policy. The paper reports better sample efficiency and obstacle avoidance than its comparison methods in simulation; the physical demonstrations establish a narrower deployment result.

## Sources

- [arXiv Query: search_query=cat:cs.RO&id_list=&start=0&max_results=30](https://arxiv.org/abs/2610.19627v1)

---
Canonical: https://techandbusiness.org/newswire/4NU4b6qGsaoKtF9ZA-fcif
Published: 2026-10-15T02:07:25.148Z
Story chronology: 2026-10-14T05:54:25.000Z
Retrieved: 2026-10-15T04:37:56.634Z
Publisher: Tech & Business (techandbusiness.org)