ANNarchy - ROS

ANNarchy (Artificial Neural Networks architect, www.github.com/ANNarchy/ANNarchy) is a parallel and hybrid simulator for distributed rate-coded or spiking neural networks. This package implements an interface to the ROS middleware. The ANNarchy-ROS package is released under the GNU GPL v3 or later.

Authors

• Helge Ülo Dinkelbach (helge-uelo.dinkelbach@informatik.tu-chemnitz.de)

Installation

Pre-requisites

If not already installed please install the rospy package (https://wiki.ros.org/ROS/Installation, currently we use ROS 1 Noetic).

For conda-users, one can use RoboStack for ROS Noetic - https://robostack.github.io/GettingStarted.html

If not already installed please install the ANNarchy package (see https://annarchy.github.io/Installation.html for more details).

ANNarchy-ROS package

As the ANNarchy-ROS package is not available in PyPi yet (it's planned for the near future), you need to install it from sources via:

pip3 install .

in the source directory.

The turtlesim - demo (closed_loop)

As a first working prototype, we use the turtlesim environment already shipped with ROS. You need to run three separate terminal windows. Run the commands in the following order:

1. roscore : starts the ROS run-time environment.
2. rosrun turtlesim turtlesim_node which starts the turtlesim environment.
3. python3 simple_demo.py located in examples which runs the ANNarchy script.

This sample demonstrates the direct interaction between ANNarchy and the turtlesim environment. The position of the turtle is send to ANNarchy, being processed, a new action is computed, and then send to the turtlesim environment.

The turtlesim - demo (external_cmd)

As a first working prototype, we use the turtlesim environment already shipped with ROS. You need to run three separate terminal windows. Run the commands in the following order:

1. roscore : starts the ROS run-time environment.
2. rosrun turtlesim turtlesim_node which starts the turtlesim environment.
3. python3 simple_demo.py located in examples which runs the ANNarchy script.
4. python3 sim_control.py located in examples which runs a simple client that repeatedly sends a simulate request to the ANNarchy controller.

This sample demonstrates the direct interaction between ANNarchy and the turtlesim environment. The position of the turtle is send to ANNarchy, being processed, a new action is computed, and then send to the turtlesim environment. Contrary to the above sample, in this setup the new action is only computed being triggered by a command from an external source (i.e., sim_control).