Skip to content

Latest commit

 

History

History
24 lines (20 loc) · 1.58 KB

README.md

File metadata and controls

24 lines (20 loc) · 1.58 KB
Raw

Multi-Agent Voxblox

This repository contains a multi-agent version of Voxblox. It is based on the original Voxblox implementation, with the additional capacity of fusing the sensory data from multiple robots.
This repository is used in a wider framework for multi-robot path planning, available here.

If you use this code in your academic work, please cite (PDF):

@inproceedings{bartolomei2020multi,
  title={Multi-robot Coordination with Agent-Server Architecture for Autonomous Navigation in Partially Unknown Environments},
  author={Bartolomei, Luca and Karrer, Marco and Chli, Margarita},
  booktitle={2020 {IEEE/RSJ} International Conference on Intelligent Robots and Systems ({IROS})},
  year={2020}
}

Installation

To install the multi-agent version of Voxblox, follow these instructions. In addition, clone the following repository:

$ git clone git@github.com:VIS4ROB-lab/comm_msgs.git # with ssh
$ git clone https://github.com/VIS4ROB-lab/comm_msgs.git # with https

ROS Nodes

The multi-agent version of Voxblox allows to use both the standard implementation and the modified one. In particular:

  • To run the standard Voxblox nodes (single agents), use the executables: esdf_server and tsdf_server
  • The executables that support the multi-robot pose-graph backend are: esdf_pose_graph_server and tsdf_pose_graph_server