Skip to content

Latest commit

 

History

History
123 lines (97 loc) · 6.18 KB

README.md

File metadata and controls

123 lines (97 loc) · 6.18 KB

SkiROS2: Skill-based robot control platform for ROS

Overview

SkiROS2 is a platform to create complex robot behaviors by composing skills - modular software blocks - into behavior trees.

Robots coordinated with SkiROS can be used in partially structured environments, where the robot has a good initial understanding of the environment, but it is also expected to find discrepancies, fail using initial plans and react accordingly.

SkiROS offers the following features:

  • A framework to organize the robot behaviors within modular skill libraries
  • Scalable skill model with pre-, hold- and post-conditions
  • A reactive execution engine based on Behavior trees
  • A world model as a semantic database to manage environmental knowledge
  • Reasoning capabilities and automatic inference of skill parameters
  • An integration point for PDDL task planning
  • Automatic generation of planning domains based on the skills and entities in the world model
  • ROS, RViz and tf integration
  • Python APIs for skill handling, the world model and task planning

A figure of the SkiROS2 architecture

Getting Started

Videos

We have video introductions to the platform with varying lengths. Feel free to choose depending on your time budget. The longer ones always include the content of the short ones.

1min Pitch 5min Short Introduction 20min ROSCon 2023
Screenshot of the 1 minute introduction Screenshot of the 5 minute introduction Screenshot of the ROSCon 2023 talk
1min URL 5min URL ROSCon Talk URL

Installation

Compatibility

SkiROS is compatible with Ubuntu 18.04/ROS Melodic and Ubuntu 20.04/ROS Noetic, Python 2 and 3. We are also working on a ROS 2 port for Humble.

Installation Instructions

To use SkiROS you must have ROS installed on your machine. You also need pip to install python dependencies.

Clone this repository into your catkin workspace src directory:

mkdir -p catkin_ws/src && cd catkin_ws/src
git clone https://github.com/RVMI/skiros2
# Clone the base skill set into the skiros2 directory in your catkin workspace.
git clone https://github.com/RVMI/skiros2_std_lib
# Optionally, you can clone the skiros2_examples repositories here as well:
git clone https://github.com/RVMI/skiros2_examples

Install dependencies defined in each package.xml using rosdep and the python dependencies:

rosdep install --from-paths . --ignore-src --rosdistro=$ROS_DISTRO -y
# Install Python dependencies
pip install -r requirements.txt --user

Build the workspace with catkin

cd ~/catkin_ws
catkin_make # or 'catkin build'
source ./devel/setup.bash

Launch SkiROS2

roslaunch skiros2 skiros.launch
# Or try one of the examples like
roslaunch skiros2_examples turtlesim_example.launch

Task Planning

Optionally, if you want to use the task planning skill provided in the standard library, install the fast downward planner with the following script:

cd skiros2/scripts
./install_fd_task_planner.sh

When asked for the install folder, you can insert a location of your preference or just leave the default.

Creating own skills

The skiros2_template_lib provides the necessary structure to create a new library from scratch. It also has an example launch file to start the system.

Build the Doxygen documentation

To generate the doxygen documentation, first install Doxygen and doxypypy for better compatibility with python doc-strings:

sudo apt-get install -y doxygen
pip install doxypypy --user

Then, in the top level directory, generate the documentation and view it with a browser, e.g. Firefox:

doxygen Doxyfile
firefox skiros2_doc/html/index.html

Citation

The SkiROS2 paper is available on IEEE Xplore here (arXiv preprint). If you are using it in your work, we would be pleased if you would cite it:

@inproceedings{mayr2023skiros2,
  title={SkiROS2: A Skill-Based Robot Control Platform for ROS},
  author={Mayr, Matthias and Rovida, Francesco and Krueger, Volker},
  booktitle={2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages={6273--6280},
  year={2023},
  organization={IEEE}
}

Acknowledgements

This platform has been developed in the RVMI lab at Aalborg University.

During its creation it has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 723658, Scalable4.0.
This work was partially supported by the Wallenberg AI, Autonomous Systems and Software Program (WASP) funded by Knut and Alice Wallenberg Foundation