CSM SLAM

CSM SLAM is a ROS 2 package that implements Simultaneous Localization and Mapping (SLAM) using the Correlative Scan Matcher algorithm as the scan matcher and GTSAM as the backend optimizer. This package provides offline and Online SLAM capabilities for processing laser scan data and building maps of unknown environments.

How to Run

Prerequisites

• ROS 2 (Humble or later)
• Python 3

Build the Package

Create a virtual environment and install the dependencies.

conda create -n csm_slam python=3.12
conda activate csm_slam
pip install -r requirements.txt

Build the package.

mkdir -p csm_ros_ws/src
cd csm_ros_ws/src

Clone the repository in the src folder.

cd csm_ros_ws
conda activate csm_slam
source /opt/ros/<ros_distro>/setup.bash
colcon build --packages-select csm_slam
source install/setup.bash

Set the venv_path in config/venv_settings.yaml to the path of the virtual environment you created. Adjust the other config file to your needs.

Run Online SLAM

ros2 launch csm_slam csm_slam.launch.py bag_path:=/path/to/your/rosbag2

Run Offline SLAM

ros2 launch csm_slam csm_slam_offline.launch.py bag_path:=/path/to/your/rosbag2

Both launch files will automatically start RViz2 for visualization

Dataset tested with this package can be found at https://google-cartographer-ros.readthedocs.io/en/latest/demos.html#revo-lds. To download the dataset, run the following command:

wget -P ~/Downloads https://storage.googleapis.com/cartographer-public-data/bags/revo_lds/cartographer_paper_revo_lds.bag

Output of the slam tested in ROS kilted, Ubuntu 24.04 and Conda environment:

If something is not working, please open an issue.

TODO

• Add tests.