Traversing Mars: Cooperative Informative Path Planning to Efficiently Navigate Unknown Scenes

This repository will contain the code for our informative path planning (IPP) framework allowing a mobile scout (e.g. an MAV) to explore the cost-optimal path of a follower robot (e.g. a rover) to a goal location in unknown environments. (Models: Courtesy NASA/JPL-Caltech.)

This work was supported by ETH Research Grant AvalMapper ETH-10 20-1 and the Swiss National Science Foundation (SNSF) grant No. 214489.

Credits

If you find this useful for your research, please consider citing our paper:

• F. M. Rockenbauer, J. Lim, M. G. Müller, R. Siegwart and L. Schmid, "Traversing Mars: Cooperative Informative Path Planning to Efficiently Navigate Unknown Scenes" in IEEE Robotics and Automation Letters, vol. 10, no. 2, pp. 1776-1783, Feb. 2025, doi: 10.1109/LRA.2024.3513036 [ IEEE ] [ ArXiv ]

  @ARTICLE{10783051,
    author={Rockenbauer, Friedrich M. and Lim, Jaeyoung and Müller, Marcus G. and Siegwart, Roland and Schmid, Lukas},
    journal={IEEE Robotics and Automation Letters}, 
    title={Traversing Mars: Cooperative Informative Path Planning to Efficiently Navigate Unknown Scenes}, 
    year={2025},
    volume={10},
    number={2},
    pages={1776-1783},
    keywords={Robots;Costs;Space exploration;Path planning;Navigation;Planning;Mobile robots;Green products;Wheels;Uncertainty;View planning for SLAM;path planning for multiple mobile robots or agents;space robotics and automation},
    doi={10.1109/LRA.2024.3513036}
    }

Installation

There are two ways to install the planner. The easier way using a docker container is described below. If you want to install the planner into your local ros workspace, follow the istructions here.

1. Setup docker as described here

2. Clone the repository:

git clone https://github.com/ethz-asl/scouting-ipp.git
git clone [email protected]:ethz-asl/scouting-ipp.git

3. Build the docker image:

cd scouting-ipp/docker/ 
docker build -t scouting_ipp:1.0 -f scouting_ipp.Dockerfile .

4. Set up the data folder:

cd ../..
cp scouting-ipp/docker/data <your_data_directory>

Note: The data directory will become multiple GBs very quickly.

5. Run a docker container and mount the data directory into it:

docker run -it --mount src=<your_data_directory>,target=/home/root/data,type=bind scouting_ipp:1.0 /bin/bash

6. Run the planner either using the examle.launch file:

roslaunch ros_planner_planexp example.launch 

or run an experiment:

roscd ros_planner_planexp
source scripts/run_experiment.sh

Animation

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

License

MIT