Multi-Agent Programming Contest Workspace

Technische Universität Berlin - DAI-Labor - http://www.dai-labor.de/

Contest homepage: https://multiagentcontest.org/2017/

This is the main workspace for the TUB participation in the Multi-Agent Contest (MAPC) 2017. It applies the ROS Hybrid Behaviour Planner Framework (RHBP) on top of the ROS (Robot Operating System) framework.

Setup and Install

Dependencies

Graphhopper requires

sudo apt install maven default-jdk

tub_contest requires

sudo apt install python-pip python-pandas

RHBP requires ROS. Follow instructions for ROS Kinetic here: http://wiki.ros.org/kinetic/Installation/Ubuntu

sudo apt install ros-kinetic-desktop python-pip
pip install lindypy

Expected directory Structure

• your workspace
  • mac_workspace
  • massim
    • massim-2017-*.*

Clone and build

1. git clone --recursive [email protected]:mac17/mac_workspace.git
2. cd mac_workspace
3. catkin_make
4. Download the latest MASSIM release https://github.com/agentcontest/massim/releases
5. Adjust the script "script/script/start_massim.sh" to use the correct release.
6. If you want to use the RHBP rqt plugin you have to start rqt once with rqt --force-discover

Execution and further Documentation

Please refer to the tub_contest package for further instruction on how to launch our solution.

Packages and Workspace Structure

• script useful scipts, you might want to add this directory to your PATH variable.
• third-party third-party module dependencies that are referenced as git submodules
• src/rhbp git submodule of the used RHBP framework.
• src/mac_ros_bridge ROS package that includes a proxy ROS node that works as a bridge between the massim simulation server and ROS. It converts all simulation perception and creates all required topics from the configuration.
• src/mac_rhbp_example example MAPC agent implementation using the mac_ros_bridge, RHBP and ROS.
• src/tu_contest The agent implementations for the actual contest.