为什么值得关注

它将多个成熟的开源项目整合为可在真实机器人上运行的导航演示，并提供了详细的硬件部署、网络配置和分步操作指南，体现了短周期工程集成的价值。

适合谁使用

希望实现自主导航的 Unitree Go2‑W 用户
需要参考 ROS2 导航流程的机器人研究人员
将 Livox MID360 与 ROS2 集成的爱好者
评估四足机器人实时 3D 导航的工程师

典型使用场景

Go2‑W 在室内环境中的自主 3D 导航
使用 FAST‑LIO 进行实时建图与重定位
通过 jie_3d_nav 的 Web 界面设置目标点与路径规划
教学演示：将激光雷达、里程计和控制集成到真实机器人上

项目优势

完整的端到端流水线，经真实硬件验证
清晰的双子网 IP 配置，降低网络搭建难度
详细的构建步骤和故障排除章节，减少配置障碍
基于 FAST‑LIO、Livox SDK2、jie_3d_nav 等成熟开源库

使用前须知

非通用导航框架，紧密绑定特定硬件（Go2‑W、MID360、NUC）
假设机器人从地图原点启动，且需要 PC 端人工处理点云地图
存在定位漂移可能，建议以重启系统作为首要解决手段

README 快速开始

PermitNav for Go2-W

PermitNav is a rapidly integrated 3D navigation demo for Unitree Go2-W / Go2W, built with Livox MID360, ROS2 Humble, FAST-LIO, FAST_LIO_LOCALIZATION2, jie_3d_nav, and Unitree SDK2.

The name PermitNav comes from the original motivation of this project: building a working real-robot navigation demo in a very short time to prove that the system could run and to earn permission to leave for an internship.

This project is mainly intended to achieve a basic 3D navigation pipeline on a real Go2-W robot. It is an engineering integration project rather than a from-scratch navigation framework. This project is only meant to achieve a basic 3D navigation demo. My main focus is locomotion, and I have not worked much on navigation before. If bugs appear, restarting the system is probably the best first solution.

Video：【【开源】PermitNav，缝合速成的一种全局3d导航】 https://www.bilibili.com/video/BV12Y7j6iEHi/?share_source=copy_web&vd_source=b5c22f93b0d2fd36b1d6203208bd650b

The NUC side is responsible for:

MID360 data acquisition
  ↓
FAST-LIO mapping / real-time odometry
  ↓
FAST_LIO_LOCALIZATION2 global localization
  ↓
localization bridge publishing map -> base_link
  ↓
jie_3d_nav Web planning and controller
  ↓
go2w_cmd_bridge forwarding /cmd_vel to Go2-W

Default workspace:

~/go2w_nav_ws

Acknowledgements

This project is not a navigation framework implemented from scratch. It is a real-robot ROS2 navigation integration project built quickly on top of existing open-source projects.

We sincerely thank the authors and maintainers of the following projects:

Open-source project	Usage in this project
Livox-SDK2	Low-level Livox LiDAR SDK used to receive data from MID360 and other Livox LiDARs.
livox_ros_driver2	ROS/ROS2 driver for Livox LiDARs. In this project, it publishes MID360 point cloud and IMU data.
FAST_LIO_ROS2	ROS2 version of FAST-LIO. Used for LiDAR-inertial odometry, real-time mapping, `/Odometry`, and registered point cloud publishing.
FAST_LIO_LOCALIZATION	FAST-LIO based localization framework. In this project, it is adapted for relocalization in a pre-

项目描述

A rapid ROS2 3D navigation integration for Unitree Go2-W with Livox MID360.

PermitNav