Swarm-SLAM: Sparse Decentralized Collaborative Simultaneous Localization and Mapping Framework for Multi-Robot Systems

Look up our Documentation and our Start-up instructions!

Swarm-SLAM is an open-source C-SLAM system designed to be scalable, flexible, decentralized, and sparse, which are all key properties in swarm robotics. Our system supports lidar, stereo, and RGB-D sensing, and it includes a novel inter-robot loop closure prioritization technique that reduces inter-robot communication and accelerates the convergence.

To clone Swarm-SLAM:

sudo apt install python3-vcstool
git clone https://github.com/MISTLab/Swarm-SLAM.git
cd Swarm-SLAM
mkdir src
vcs import src < cslam.repos

Packages summary:

• cslam: contains the Swarm-SLAM nodes;
• cslam_interfaces: contains the custom ROS 2 messages;
• cslam_experiments: contains examples of launch files and configurations for different setups;
• cslam_visualization: contains an online (optional) visualization tool to run on your base station to monitor the mapping progress.

Follow the start-up instructions to install, build and run Swarm-SLAM.

How to cite our paper:

@ARTICLE{lajoieSwarmSLAM,
  author={Lajoie, Pierre-Yves and Beltrame, Giovanni},
  journal={IEEE Robotics and Automation Letters}, 
  title={Swarm-SLAM: Sparse Decentralized Collaborative Simultaneous Localization and Mapping Framework for Multi-Robot Systems}, 
  year={2024},
  volume={9},
  number={1},
  pages={475-482},
  doi={10.1109/LRA.2023.3333742}
}