ROS Index
No version for distro humble. Known supported distros are highlighted in the buttons above.
No version for distro jazzy. Known supported distros are highlighted in the buttons above.
No version for distro rolling. Known supported distros are highlighted in the buttons above.
|
|
Repository Summary
| Checkout URI | https://github.com/ei-nav/light-map-navigation.git |
| VCS Type | git |
| VCS Version | main |
| Last Updated | 2025-02-14 |
| Dev Status | UNMAINTAINED |
| CI status | No Continuous Integration |
| Released | UNRELEASED |
| Tags | No category tags. |
| Contributing |
Help Wanted (0)
Good First Issues (0) Pull Requests to Review (0) |
Packages
| Name | Version |
|---|---|
| building_entrance_recognition | 0.0.0 |
| icp_registration | 0.0.0 |
| classic_nav_bringup | 0.0.0 |
| classic_navigation | 0.0.0 |
| imu_complementary_filter | 2.1.3 |
| ipm_image_node | 0.0.0 |
| ipm_interfaces | 0.0.0 |
| ipm_library | 0.0.0 |
| ipm_service | 0.0.0 |
| linefit_ground_segmentation | 0.0.1 |
| linefit_ground_segmentation_ros | 0.0.1 |
| pointcloud_to_laserscan | 2.0.1 |
| custom_interfaces | 0.2.0 |
| delivery_bringup | 0.0.0 |
| delivery_executor | 0.0.0 |
| entrance_exploration | 0.0.0 |
| gazebo_simulation | 1.0.0 |
| ros2_livox_simulation | 0.0.0 |
| grounded_sam2 | 0.0.0 |
| llm_delivery | 0.0.0 |
| replanner | 0.0.0 |
| task_planning | 0.0.0 |
| utils_pkg | 0.0.0 |
README
OPEN: Openstreetmap-enhanced oPen-air sEmantic Navigation
OPEN is a system designed for efficient last-mile delivery using autonomous robots. It leverages OpenStreetMap (OSM) for scalable map representation, Large Language Models (LLMs) for understanding delivery instructions, and Vision-Language Models (VLMs) for localization, map updates, and house number recognition.
Installation
The current development environment is Ubuntu 22.04, ROS2 Humble, and Gazebo Classic 11.
Using Docker
This project includes a basic Dockerfile and can be used with Dev Container for simulation testing and development.
- Install Docker and NVIDIA Container Toolkit and configure them on the host machine:
# Docker installation
https://docs.docker.com/engine/install/ubuntu/
wget http://fishros.com/install -O fishros (optional)
# NVIDIA Container Toolkit installation
https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html
# nvidia-docker installation
sudo apt install nvidia-docker2
# Configure Docker container to use GPU
sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker
- Download and install the image:
# Download from Baidu Cloud
Link: https://pan.baidu.com/s/1xSDHBTYh8PTt4HNBLy7FWA?pwd=emgo
Password: emgo
# Load the image
docker load -i open.tar
- Clone the repository:
git clone https://github.com/EI-Nav/light-map-navigation.git --depth=1
git submodule init
git submodule update
cd src/classic_localization/FAST_LIO/
git submodule init
git submodule update
- Download Gazebo models and extract them to the
~/.gazebo/directory:
# Download from Baidu Cloud Drive (Choose one of the following):
# Option 1: Complete model library (includes unused models)
Link: https://pan.baidu.com/s/1Mkn4BHXaeWyvjxCMvw-gZQ
Password: pp8a
# Option 2: Minimal model library (recommended, contains only required models)
Link: https://pan.baidu.com/s/1CojOxnFrJjSLM_jpmkK_KA
Password: 49xs
# After extraction, place files in
~/.gazebo/
-
Install the
ms-vscode-remote.remote-containersextension in VSCode on the host machine. -
Open this project in VSCode, press
Ctrl+Shift+P, type and selectDev Containers: Rebuild and Reopen in Container. -
Install dependencies and compile:
# Set up your proxy
export https_proxy=http://127.0.0.1:7890 http_proxy=http://127.0.0.1:7890 all_proxy=socks5://127.0.0.1:7890
# Install dependencies
./build_dependencies.sh
# Compile
./build_project.sh
Source Installation
- Clone the repository:
git clone https://github.com/EI-Nav/light-map-navigation.git --depth=1
git submodule init
git submodule update
cd src/classic_localization/FAST_LIO/
git submodule init
git submodule update
- Install Livox SDK2:
sudo apt install cmake
git clone https://github.com/Livox-SDK/Livox-SDK2.git
cd ./Livox-SDK2/
mkdir build
cd build
cmake .. && make -j
sudo make install
- Set up the conda environment:
conda create --name dl_env python=3.10
conda activate dl_env
conda install -c conda-forge libstdcxx-ng
pip install "numpy<2.0" pandas matplotlib scikit-learn jupyterlab
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
pip install opencv-python pycocotools matplotlib onnxruntime
pip install git+https://github.com/facebookresearch/sam2.git
pip install git+https://github.com/IDEA-Research/GroundingDINO.git
- Install dependencies for ros package:
cd light-map-navigation
rosdep install -r --from-paths src --ignore-src --rosdistro $ROS_DISTRO -y
pip3 install -r requirements.txt
-
Install other dependencies according to the build_dependencies.sh.
-
Compile:
./build.sh
Configuration
To use this repository, you’ll need to configure the following credentials:
VLM API
# /workspace/light-map-navigation/src/building_entrance_recognition/config/api_config.yaml
vlm_api:
key: "your_key"
base_url: "your_base_url"
model_name: "your_model_name"
LLM API
# /workspace/light-map-navigation/src/task_planning/config/api_config.yaml
llm_api:
key: "your_key"
base_url: "your_base_url"
model_name: "your_model_name"
GSAM2 Config
grounded_sam2_node:
ros__parameters:
gsam2:
scale: 1.0
do_sharpen: false
base_path: "/workspaces/light-map-navigation/src/grounded_sam2/grounded_sam2/grounded_sam2/"
text_prompt: "ground . car . building"
sam2_checkpoint: "./checkpoints/sam2.1_hiera_small.pt"
sam2_model_config: "configs/sam2.1/sam2.1_hiera_s.yaml"
grounding_dino_config: "grounding_dino/groundingdino/config/GroundingDINO_SwinT_OGC.py"
grounding_dino_checkpoint: "gdino_checkpoints/groundingdino_swint_ogc.pth"
box_threshold: 0.3
text_threshold: 0.25
area_threshold: 80
processing_rate: 10.0
Run
# Allow the container to access the host's X server
xhost +
Simulation Mode Example
# Run the robot's navigation system and simulation environment
ros2 launch classic_nav_bringup bringup_sim.launch.py \
world:=MEDIUM_OSM \
mode:=nav \
lio:=fastlio \
localization:=icp \
lio_rviz:=False \
nav_rviz:=True \
use_sim_time:=True
# Run delivery-related nodes
ros2 launch delivery_bringup delivery_system_sim.launch.py
# Run client to send delivery requests
ros2 run delivery_executor delivery_executor_action_client
GSAM2 Node
If you want to use Grounded-SAM-2 for open-vocabulary instance segmentation, you can run the following commands:
# Set up your proxy
export https_proxy=http://127.0.0.1:7890 http_proxy=http://127.0.0.1:7890 all_proxy=socks5://127.0.0.1:7890
# Download SAM2 checkpoints
cd src/grounded_sam2/grounded_sam2/grounded_sam2/checkpoints
bash download_ckpts.sh
# Download GroundingDINO checkpoints
cd src/grounded_sam2/grounded_sam2/grounded_sam2/gdino_checkpoints
bash download_ckpts.sh
# Run Grounded-SAM-2 server
cd /workspaces/light-map-navigation
ros2 launch grounded_sam2 gsam2_node.launch.py
Robot Model
Four-wheeled vehicle using a differential drive model.
Known Issues
- There are coordinate system issues, with the vehicle appearing below the map.
TODO
- Delete old delivery code, models, and tests.
- Improve documentation.
Acknowledgments
This project builds upon and is inspired by several excellent works:
We thank all the contributors of these open-source projects for their valuable work.
Citation
If you find our work useful for your research, please consider giving this repository a star ⭐ and citing our paper as follows:
@misc{wang2025openbenchnewbenchmarkbaseline,
title={OpenBench: A New Benchmark and Baseline for Semantic Navigation in Smart Logistics},
author={Junhui Wang and Dongjie Huo and Zehui Xu and Yongliang Shi and Yimin Yan and Yuanxin Wang and Chao Gao and Yan Qiao and Guyue Zhou},
year={2025},
eprint={2502.09238},
archivePrefix={arXiv},
primaryClass={cs.RO},
url={https://arxiv.org/abs/2502.09238},
}
CONTRIBUTING
No CONTRIBUTING.md found.
No version for distro noetic. Known supported distros are highlighted in the buttons above.
No version for distro ardent. Known supported distros are highlighted in the buttons above.
No version for distro bouncy. Known supported distros are highlighted in the buttons above.
No version for distro crystal. Known supported distros are highlighted in the buttons above.
No version for distro eloquent. Known supported distros are highlighted in the buttons above.
No version for distro dashing. Known supported distros are highlighted in the buttons above.
No version for distro galactic. Known supported distros are highlighted in the buttons above.
No version for distro foxy. Known supported distros are highlighted in the buttons above.
No version for distro iron. Known supported distros are highlighted in the buttons above.
No version for distro lunar. Known supported distros are highlighted in the buttons above.
No version for distro jade. Known supported distros are highlighted in the buttons above.
No version for distro indigo. Known supported distros are highlighted in the buttons above.
No version for distro hydro. Known supported distros are highlighted in the buttons above.
No version for distro kinetic. Known supported distros are highlighted in the buttons above.
No version for distro melodic. Known supported distros are highlighted in the buttons above.