Retro-SLAM

Install   •   Paper   •   Contact Us

Retro-SLAM: Improving Map Consistency in Graph-Based LiDAR SLAM Through Information-Aware Odometry and Retroactive Loop Closure

---

Install

pip install retro_slam

Running the system

Next, follow the instructions on how to run the system by typing:

retro_slam_pipeline --help

This should print the following help message:

Config

You can generate a default config.yaml by typing:

retro_slam_dump_config

which will generate a retro_slam.yaml file. Now, you can modify the parameters and pass the file to the --config option when running the retro_slam_pipeline.

Install Python API (developer mode)

For development purposes:

python3 -m pip install --upgrade pip
git clone https://github.com/PRBonn/retro-slam.git
cd retro-slam
make editable

ROS2 support

cd ~/ros2_ws/src/ && git clone https://github.com/PRBonn/retro_slam && cd ~/ros2_ws/ && colcon build --packages-select retro_slam_ros

For more detailed instructions on the ROS wrapper, please visit this README

Citation

If you use this library for any academic work, please cite our original paper:

@article{gupta2026arxiv,
  author   = {S. Gupta and N. Trekel and L.Wiesmann and C. Stachniss},
  title    = {{Improving Map Consistency in Graph-Based LiDAR SLAM Through Information-Aware Odometry and Retroactive Loop Closure}},
  journal  = {arXiv preprint},
  year     = 2026,
  volume   = {arXiv:2607.13516},
  url      = {https://arxiv.org/pdf/2607.13516},
}

Note

1. We use a fork of KISS-ICP for LiDAR odometry and the uncertainty estimation proposed in the Retro-SLAM manuscript is implemented in this fork as well.
2. We use the MapClosures repository for local-map-level loop closure detection in our work.

Acknowledgements

This project builds on top of KISS-SLAM.

Contributing

We envision this as a community-driven project. We love to see how the project is growing, thanks to the contributions from the community. We would love to see your face in the list below; open a Pull Request!

Contact Us

For questions or feedback:

• GitHub Issues: https://github.com/PRBonn/retro_slam/issues