Graph-based Exploration Planner for Subterranean Environments
- Clone the package:
git clone https://github.com/unr-arl/gbplanner_ros
- Build the package:
catkin build -DCMAKE_BUILD_TYPE=Release gbplanner_ros
However, please be aware that the planner requires several dependencies; for example:
- Mapping framework: either Voxblox or Octomap
- Planner-Control interface package: pci_mav
- MAV simulation: RotorS, LiDAR simulator
Hence, to simplify the test process with the planner, we prepare an example workspace gbplanner_ws, which includes all relevant packages. Please follow the suggested steps in the repo gbplanner_ws for further detail.
- Launch file for simulation:
roslaunch gbplanner gbplanner_sim.launch
- To trigger the planner in command line, please use the service call:
rosservice call /planner_control_interface/std_srvs/automatic_planning "{}"
- We provide several types of environment for simulation in the package planner_gazebo_sim including room-and-pillar mine (
pittsburgh_mine.world
), long-wall mine (edgar_mine.world
), and a model reconstructed from a real mine (virginia_mine.world
)
By default, the planner is compiled with Voxblox. To compile with Octomap, set the flag USE_OCTOMAP to 1:
catkin build -DCMAKE_BUILD_TYPE=Release -DUSE_OCTOMAP=1 gbplanner_ros
Also change the config to Octomap in the gbplanner_sim.launch file
<arg name="map_config_file" default="$(arg octomap_config_file)"/>
You could find a short tutorial on the plannning algorithm and the overall architecture on our website: Link
If you use this work in your research, please cite the following publication.
@inproceedings{dang2019graph,
title={Graph-based path planning for autonomous robotic exploration in subterranean environments},
author={Dang, Tung and Mascarich, Frank and Khattak, Shehryar and Papachristos, Christos and Alexis, Kostas},
booktitle={2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={3105--3112},
year={2019},
organization={IEEE}
}
You can contact us for any question: