mir_robot

Disclaimer

This package is intended to be a migration package of mir_robot using ros1 to ros humble. It is currently still under heavy development - to be used with caution! The following packages have been migrated to humble:

• mir_actions
• mir_description
• mir_driver
• mir_gazebo
• mir_msgs
• mir_navigation
• sdc21x0

This repo contains a ROS driver and ROS configuration files (URDF description, Gazebo launch files, move_base config, bringup launch files, message and action descriptions) for the MiR robots. This is a community project created by us (DFKI, the German Research Center for Artificial Intelligence) to use the MiR Robots with ROS. We are not affiliated with Mobile Industrial Robots. If you find a bug or missing feature in this software, please report it on the issue tracker.

Supported MiR robots and software versions

Package overview

• mir_actions: Action definitions for the MiR robot
• mir_description: URDF description of the MiR robot
• mir_dwb_critics: Plugins for the dwb_local_planner used in Gazebo
• mir_driver: A reverse ROS bridge for the MiR robot
• mir_gazebo: Simulation specific launch and configuration files for the MiR robot
• mir_msgs: Message definitions for the MiR robot
• mir_navigation: move_base launch and configuration files

Installation

Preliminaries

ROS2

If you haven't already installed ROS2 on your PC, you need to add the ROS2 apt repository. This step is necessary for either binary or source install.

Source install

For a source install, run the commands below instead of the command from the "binary install" section.

# create a ros2 workspace
mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/

# clone mir_robot into the ros2 workspace
git clone -b humble-devel https://github.com/relffok/mir_robot src/mir_robot

# use vcs to fetch linked repos
vcs import < src/mir_robot/ros2.repos src --recursive

# use rosdep to install all dependencies (including ROS itself)
sudo apt update
sudo apt install -y python3-rosdep
rosdep update --rosdistro=humble
rosdep install --from-paths src --ignore-src -r -y --rosdistro humble

# build all packages in the workspace
source /opt/ros/humble/setup.bash
cd ~/ros2_ws
colcon build

You must source the workspace in each terminal you want to work in:

source ~/ros2_ws/install/setup.bash

Alternatively, you can add the following line to the end of your ~/.bashrc, and then close and reopen all terminals:

echo "source ~/ros2_ws/install/setup.bash" >> ~/.bashrc

Gazebo demo (Navigation with existing map)

### gazebo
ros2 launch mir_gazebo mir_gazebo_launch.py world:=maze rviz_config_file:=$(ros2 pkg prefix mir_navigation)/share/mir_navigation/rviz/mir_nav.rviz


### localization (existing map)
ros2 launch mir_navigation amcl.py use_sim_time:=true map:=$(ros2 pkg prefix mir_gazebo)/share/mir_gazebo/maps/maze.yaml

### navigation
ros2 launch mir_navigation navigation.py use_sim_time:=true

• HEADLESS MODE: To run the simulation without the gazebo graphical interface (performance improvement) add to the gazebo launch command:

gui:=false

If necessary, initialize the position with "2D Pose Estimate". You should see sensor inputs and costmap by now. Now, you can use the "2D Goal Pose" tool in RViz to set a navigation goal for move_base

Gazebo demo (mapping)

### gazebo:
ros2 launch mir_gazebo mir_gazebo_launch.py world:=maze

### mapping (slam_toolbox)
ros2 launch mir_navigation mapping.py use_sim_time:=true slam_params_file:=$(ros2 pkg prefix mir_navigation)/share/mir_navigation/config/mir_mapping_async_sim.yaml

### navigation (optional)
ros2 launch mir_navigation navigation.py use_sim_time:=true

Running the driver on the real robot

Start up the robot

• switch on MiR base

Connect to the MiR web interface

• connect to MiR_R??? wifi (password "mirex4you"), for example from your Android phone/tablet
• disable other network connections (mobile data / LAN / etc.)
• open mir.com (192.168.12.20) in Chrome (!)
• log in (admin/mir4you)

Synchronize system time

The internal robot PC's is not synchronized (for example via NTP), so it tends to get out of sync quickly (about 1 second per day!). This causes TF transforms timing out etc. and can be seen using tf_monitor (the "Max Delay" is about 3.3 seconds in the following example, but should be less than 0.1 seconds):

Determine delay using ROS1

Since MiR is running a roscore (ROS1), the following tf_monitor can be excecuted after sourcing ROS1 (i.e. noetic) first:

source /opt/ros/noetic/setup.bash
export ROS_MASTER_URI=http://192.168.12.20:11311

$ rosrun tf tf_monitor
Frames:
Frame: /back_laser_link published by unknown_publisher Average Delay: 3.22686 Max Delay: 3.34766
Frame: /base_footprint published by unknown_publisher Average Delay: 3.34273 Max Delay: 3.38062
Frame: /base_link published by unknown_publisher Average Delay: 3.22751 Max Delay: 3.34844
[...]

All Broadcasters:
Node: unknown_publisher 418.344 Hz, Average Delay: 0.827575 Max Delay: 3.35237
Node: unknown_publisher(static) 465.362 Hz, Average Delay: 0 Max Delay: 0

Determine delay using ROS2

If you don't have a ROS1 distro installed, you'll need to run the mir_driver first and execute the following once a connection is established:

ros2 run tf2_ros tf2_monitor

Fix time synchronization manually

• In the Mir dashboard (mir.com in the Mir-Wifi), go to "Service" -> "Configuration" -> "System settings" -> "Time settings" -> "Set device time on robot"

Use load from device to sync with the system time!

Fix time synchronization using ROS2:

From the package mir_restapi a node called mir_restapi_server can be run, which can execute a time sync REST API call from the driver's host machine to the Mir's host.

• Launch the node with the API key and mir hostname's IP address

  ros2 run mir_restapi mir_restapi_server --ros-args -p mir_hostname:='MIR_IP_ADDR' -p mir_restapi_auth:='YOUR_API_KEY'

• Call the time sync service from terminal by invoking

  ros2 service call /mir_sync_time std_srvs/Trigger

After time sync

Keep in mind, that the time sync causes the mir_bridge to freeze. Therefore online time syncs are not recommended.

Start move_base on the robot

• go to "Service" -> "Configuration" -> "Launch menu", start "Planner"; this starts move_base and amcl on the robot

Teleoperate the robot (optional)

• go to "Manual", press yellow button (LEDs change from yellow to blue); now the robot can be teleoperated

Relocalize robot (optional)

If the robot's localization is lost:

• go to "Service" -> "Command view" -> "Set start position" and click + drag to current position of robot in the map
• click "Adjust"

Start the ROS driver

ros2 launch mir_driver mir_launch.py

• The driver automatically launches rviz to visualize the topics and sensor messages. To disable, use rviz_enabled:=false as a launch argument.
• The driver automatically launches a seperate teleop window to manually move the robot using your keyboard. To disable use teleop_enabled:=false as a launch argument.

Mapping on MiR

Option 1: Launching the modules separately

### driver:
ros2 launch mir_driver mir_launch.py

### mapping (slam_toolbox)
ros2 launch mir_navigation mapping.py use_sim_time:=false slam_params_file:=$(ros2 pkg prefix mir_navigation)/share/mir_navigation/config/mir_mapping_async.yaml

### navigation (optional)
ros2 launch mir_navigation navigation.py use_sim_time:=false

Option 2: Use combined launch file

### combined launch file:
ros2 launch mir_navigation mir_mapping_launch.py

Navigation on MiR

Option 1: Launching the modules separately

### driver:
ros2 launch mir_driver mir_launch.py

### localization (amcl)
ros2 launch mir_navigation amcl.py use_sim_time:=false map:={path to existing map}

### navigation
ros2 launch mir_navigation navigation.py use_sim_time:=false

Option 2: Use combined launch file

### combined launch file:
ros2 launch mir_navigation mir_nav_launch.py map:={path to /name of existing map}

On Mapping

As mentioned before, you can launch the differnet modules seperately or use the combined launch commands below:

In Simulation, run:

ros2 launch mir_navigation mir_mapping_sim_launch.py

On MiR, run:

ros2 launch mir_navigation mir_mapping_launch.py

Both commands launch the simulation / driver and SLAM node.

How to map

To save the created map, use the rviz plugin "Save Map" and "Serialize Map". From time to time, segmentation faults or timeouts occur, so make sure your map is saved before shutting down the connection.

Refine existing map

The given launch commands will create a fresh new map of the environment. If you like to adapt an existing map (must be serialized!) you can deserialize it using the rviz slam_toolbox plugin.

Select map to deserialize	and continue mapping

Helpful launch arguments

• NAVIGATION: Navigation is disabled per default. If you like to teleoperate the robot using nav2, add:

navigation_enabled:=true

Mapping..	..using nav2

On Localization and Navigation

As mentioned before, you can launch the differnet modules seperately or use the combined launch commands below:

In Simulation, run:

ros2 launch mir_navigation mir_nav_sim_launch.py

On MiR, run:

ros2 launch mir_navigation mir_nav_launch.py map:={path to existing map}

Both commands launch the simulation / driver and localization (amcl) using an existing map.

Helpful launch arguments

• MAP: In Simulation, the map defaults to the maze map. On the real robot, a map should be passed via the map argument:

map:= {path_to_map_yaml}
world:={world_file} # simulation only: add respective world

Workflow

Once the simulation / driver is running and rviz is started, you need to set the initial pose on the map. This doesn't have to be accurate, just a reference, and amcl will do the refinement. To refine, move the robot around a little using the teleop window and the scan will eventually match the map.

Initialize Pose	Drifted pose	Refined pose

Using a namespace

When using a namespace for your robot, add the --ros-args -p namespace:=my_namespace to the launch files. The driver, description and gazebo packages work out of the box. However, to use the navigation stack the corresponding config.yaml files need to be adapted to match the renamed topic.

The navigation cmd_vel topic is automatically remapped by adding the namespace in the navigation.py launch file.