ilarioazzollini / create3_control

Control algorithms for the iRobot Create3 educational robot.

Create 3 Control

How to Build and Run

Clone the repository and start a Docker container

git clone https://github.com/ilarioazzollini/create3_control
cd create3_control
docker run -it --network=host --privileged -e DISPLAY=$DISPLAY -v $PWD:/root/create3_control irobotedu/create3-galactic bash

Build the repository

mkdir -p /root/control_ws/src
cd /root/control_ws
ln -s /root/create3_control src
colcon build
source install/setup.sh

Test the repository (optional)

colcon test

and you will find the complete test output in /root/control_ws/log/latest_test/create3_control/stdout_stderr.log

Start the action server

python3 src/create3_control/create3_control/controller_server.py

Attach a new terminal to the Docker container and send a goal request to the server

ros2 action send_goal /drive_to_pose irobot_create_msgs/action/NavigateToPosition "{goal_pose:{pose:{position:{x: 1,y: 1.0,z: 0.0}, orientation:{x: 0.0,y: 0.0, z: 0.0, w: 1.0}}}}"

Use custom parameters

Controller parameters can be configured at startup using

python3 src/create3_control/create3_control/controller_server.py --ros-args -p controller_type:=polar_coordinates -p polar_coordinates.rotate_to_goal:=False

The parameters can also be modified at runtime using standard ROS 2 parameters APIs.

ros2 param set controller polar_coordinates.k_r 1.5

create3_control's Issues

Experimental results + logger + plots

• test (plus tuning) of the controller parameters on the real robot
• investigate sim2real gap accordingly
• need for a logger for the experimental data + plots

CI

• add an action for building and testing the code

Test

• simple tests for the controllers
• check (and fix) the existing code against the (already present) flake8 and pep257 tests

Update documentation

• how to use it
• algorithms explanation