Getting started¶

Component-based software architectures in robotics¶

Robots are highly complex systems that embed numerous sensors and actuators, in the service of a variety of algorithms performing heterogeneous tasks. They often have to deal with severe requirements (timing constraints, limited energy, memory and processing resources, etc.) and must show a robust conception as uncertainty about their environment is high, and unexpected events can have critical consequences.

To facilitate the development of robotic software, component-based architectures, where components are independent processes, have become the de facto standard in robotics. Each software component is dedicated to a given task, from low-level control to high-level processing. Components communicate with each other with the help of a software piece called the middleware.

For instance, consider a robot embedding a camera and performing object detection in the images. One component could be in charge of acquiring the images using the camera’s driver, and would output them. Another component could input the images and run an algorithm on them to detect objects, producing the detection result as output for any other component in need of this information. Routing data from the output of the first component to the input of the second one is ensured by the middleware, this is called data flow.

Components offer services to the user to modify their behaviour and adapt to different situations. To follow with the example above, the component acquiring images could provide a service to select which camera to use, another one to configure parameters such as the image size and the number of frames per second, a third one to explicitly request the start of the acquisition, etc. The component detecting objects could have a service to change some parameters in the detection algorithm, another one to choose which image stream to take as input (because there could be several components streaming images from different cameras), etc. Making services available to the user is again handled by the middleware, this is called control flow.

Component-based software architectures offer great benefits in robotics, in particular [1]:

• Modularity

  • As many operations handled by a robot require to have their own thread of execution (e.g. data acquisition for sensors, motion control for actuators), having them in separate programs eases their concurrent execution.
  • The architecture can be adapted to the needs of the robot: adding a new hardware piece such as a sensor will result in running a new software component to drive it.
  • The system can be distributed over a network, as the middleware seamlessly ensures communication between components running on different host machines.

• Re-usability

  • Common components can be used across robots without having to recode them from scratch.
  • Components can be packaged and easily shared in the robotics community, where open source software prevails.
  • Re-usable components reduce development cost and time, while improving software quality and sustainability.

ROS, a software platform for robotics¶

The previous section identifies the middleware as the software piece ensuring data flow between functional components, and allowing their control.

ROS is a widely known software platform in robotics, providing not only a middleware, but also implementing a wide range of commonly-used functionalities into software components (such as localisation, mapping, path-planning, obstacle avoidance, etc.), with a build system and a packaging system for easy compilation and installation. ROS benefits from a large community of users and developers, and runs on many robots today. This makes ROS a common choice as a robotic software platform, as it is for Two!Ears.

ROS embraces the principles of component-based software architectures, allowing distributed computation, software reuse and rapid testing [2]. If you will be a user of a robotic platform running ROS, the core tutorials can help you to get familiar with the ROS environment. The main ROS terminology, introduced in the tutorials, is recalled here:

Nodes

Software components using ROS middleware are called ROS nodes.

Topics and messages

Data flows are called topics. A node that outputs data publishes on a topic. A node that inputs data subscribes to a topic. The data elements flowing on topics are called messages. Each message is made of various data fields forming part of a data structure called message type. As a given topic only carries one message type, the term topic type is equally used.

Services and actions

Nodes can provide services to control them. Some special services that can take a long time to execute are called actions.

GenoM3, a tool to develop robotic components¶

The process of developing robotic components can significantly be improved by the mean of a tool called GenoM3. As a result of two decades of research on real time architectures for autonomous systems [4] [5], GenoM3 brings valuable attributes to robotic components:

• Middleware independence

  Components developed with GenoM3 are middleware independent, i.e. they are not tied to a specific middleware and can be compiled for different middlewares without changing their source code.

  This is achieved through the notion of templates: a GenoM3 template is a set of instructions which, when applied to the component’s source files, automatically generates the code related to middleware communication. A clear separation of concerns between the algorithmic core and the middleware is thus conducted, helping towards the good design of robotic components.

  ROS appears among the middlewares supported by GenoM3. When using ROS templates to compile a GenoM3 component, the resulting program is a genuine ROS node. Only the development process differs from what could be done by writing a ROS node without the GenoM3 tool.

• Model-driven design

  GenoM3 emphasises the clear definition of robotic components by adopting a model-driven approach. A GenoM3 component is first defined by a description file, called the dotgen file, with the .gen extension. This file gathers in a single place all the definitions related to the component’s interface, needed by a user to interact with it.

  Each GenoM3 component has its own dotgen file, mainly including the definition of its services and its data flows by the mean of ports (either from the component to the outside, or the other way round). Each service is defined by a name and a list of input and output parameters with their related data type. Each port is defined by a name, a direction (either in or out), and a data type for the data elements it uses. The dotgen file often include in-line documentation to help the understanding of the component’s features (for instance, the role of a given service parameter).

  On the basis of the model specified in the dotgen file, GenoM3 automatically generates real time code as well as skeletons of functions run by the services. So, the developer just has to fill these functions, called codels (for code element) with its algorithms. The corresponding algorithmic core is written in separate C or C++ source files or libraries.

• Powerful framework

  GenoM3 facilitates the development of essential features for robotic components, such as:

  • the definition of finite state automata with an optional clock,
  • clean interruption mechanisms,
  • efficient error handling.

If you decide to be a user of GenoM3 components, the tutorial GenoM3 through an example will help you to learn how to read dotgen files. For a developer of GenoM3 components, the documentation defines the whole grammar for writing dotgen files, and how dotgen specifications are mapped into C or C++ codels.

Install ROS¶

The ROS distribution you can install will depend on your Ubuntu version:

• If you have Ubuntu 14.04 LTS or 13.10, follow the ROS Indigo installation instructions.
• Otherwise, follow the ROS Hydro installation instructions.

Install the GenoM3 tools through robotpkg¶

This section will guide you through the installation of the GenoM3 tools. GenoM3 is open-source software (available at https://git.openrobots.org/) and can be compiled from source, but the common installation uses robotpkg, a compilation framework and packaging system for robotics software (more information at http://robotpkg.openrobots.org/).

cd
git clone git://git.openrobots.org/robots/robotpkg

cd ~/robotpkg
git clone git://git.openrobots.org/robots/robotpkg/robotpkg-wip wip

cd ~/robotpkg/bootstrap
./bootstrap --prefix $HOME/openrobots

echo >> ~/.bashrc
echo '# ROBOTPKG' >> ~/.bashrc
echo 'export ROBOTPKG_BASE=$HOME/openrobots' >> ~/.bashrc
echo 'export PATH=$PATH:$ROBOTPKG_BASE/bin:$ROBOTPKG_BASE/sbin' \
     >> ~/.bashrc
echo 'export PKG_CONFIG_PATH=$PKG_CONFIG_PATH:\
     $ROBOTPKG_BASE/lib/pkgconfig' >> ~/.bashrc
echo 'export ROS_PACKAGE_PATH=$ROS_PACKAGE_PATH:\
     $ROBOTPKG_BASE/src/ros-nodes:$ROBOTPKG_BASE/share' >> ~/.bashrc
echo 'export PYTHONPATH=$PYTHONPATH:\
     $ROBOTPKG_BASE/lib/python2.7/site-packages:\
     $ROBOTPKG_BASE/lib/python3.2/site-packages' >> ~/.bashrc
echo '# ROBOTPKG END' >> ~/.bashrc
source ~/.bashrc

Install a GenoM3 component from the sources¶

cd path/to/component/folder
genom3 skeleton -i component.gen
./bootstrap.sh
mkdir build && cd build
../configure --prefix=$ROBOTPKG_BASE --with-templates=ros/server,ros/client/c
make install

# assuming that you are in the software repository
cd RoboticPlateform/bass-genom3
genom3 skeleton -i bass.gen
./bootstrap.sh
mkdir build && cd build
../configure --prefix=$ROBOTPKG_BASE --with-templates=ros/server,ros/client/c
make install