Exporting to ROS 2

This guide explains how to convert Simulink controllers into ROS 2 packages for deployment on physical robots.

---

Overview

The export process converts your validated Simulink controller into a ROS 2 node that can run on real robot hardware. This enables:

• Hardware deployment: Run the same control algorithms on physical robots
• ROS 2 integration: Interface with other ROS 2 nodes (navigation, perception, planning)
• Real-time execution: Generated C++ code runs efficiently on embedded systems

---

Prerequisites

Requirement	Description
ROS Toolbox	MATLAB toolbox for ROS 2 integration
MATLAB Coder	Code generation from MATLAB functions
Embedded Coder	Optimized code generation (optional)
ROS 2	Humble or later installed on target system

---

Step 1: Prepare Simulink Model

Configure your Simulink model for code generation:

Model Configuration

% Open model configuration
set_param('model_name', 'SystemTargetFile', 'ert.tlc');
set_param('model_name', 'GenerateReport', 'on');
set_param('model_name', 'LaunchReport', 'on');

Required Settings

Parameter	Value	Location
System target file	ert.tlc	Code Generation → System target file
Language	C++	Code Generation → Language
Generate code only	On	Code Generation → Build process

---

Step 2: Create ROS 2 Interface

Add ROS 2 Publish/Subscribe Blocks

1. Open Simulink Library Browser
2. Navigate to ROS Toolbox → ROS 2
3. Add Publish and Subscribe blocks

Configure Publishers

% Example: Publish motor commands
% Topic: /cmd_vel
% Message Type: geometry_msgs/Twist

Configure Subscribers

% Example: Subscribe to sensor data
% Topic: /imu/data
% Message Type: sensor_msgs/Imu

---

Step 3: Define Message Types

Standard Message Types

Data	ROS 2 Message Type	Fields
IMU	sensor_msgs/Imu	orientation, angular_velocity, linear_acceleration
GPS	sensor_msgs/NavSatFix	latitude, longitude, altitude
Odometry	nav_msgs/Odometry	pose, twist
Velocity	geometry_msgs/Twist	linear (x,y,z), angular (x,y,z)
Pose	geometry_msgs/PoseStamped	position, orientation

Custom Message Example

% Create custom message definition
% File: msg/RobotState.msg
float64 x
float64 y
float64 theta
float64 v
float64 omega

---

Step 4: Sensor Integration

IMU Subscriber Block

% Configure IMU subscriber
% Topic: /imu/data
% Message Type: sensor_msgs/Imu
% Sample Time: 0.01 (100 Hz)

% Extract data in MATLAB Function block
function [roll, pitch, yaw, gyro] = parse_imu(imu_msg)
    % Quaternion to Euler
    quat = [imu_msg.Orientation.W, imu_msg.Orientation.X, ...
            imu_msg.Orientation.Y, imu_msg.Orientation.Z];
    [roll, pitch, yaw] = quat2eul(quat);

    % Angular velocity
    gyro = [imu_msg.AngularVelocity.X;
            imu_msg.AngularVelocity.Y;
            imu_msg.AngularVelocity.Z];
end

---

Step 5: Actuator Commands

Publish Motor Commands

% Configure Twist publisher
% Topic: /cmd_vel
% Message Type: geometry_msgs/Twist

% Create Twist message in MATLAB Function block
function twist_msg = create_twist(v, omega)
    twist_msg = rosmessage('geometry_msgs/Twist');
    twist_msg.Linear.X = v;
    twist_msg.Linear.Y = 0;
    twist_msg.Linear.Z = 0;
    twist_msg.Angular.X = 0;
    twist_msg.Angular.Y = 0;
    twist_msg.Angular.Z = omega;
end

---

Step 6: Generate ROS 2 Package

Using Simulink Coder

1. Open Apps → Simulink Coder
2. Click Build to generate code
3. Navigate to generated folder

Generated Package Structure

ros2_package/
├── CMakeLists.txt
├── package.xml
├── include/
│   └── model_name/
│       └── model_name.h
├── src/
│   ├── model_name.cpp
│   └── model_name_node.cpp
└── launch/
    └── model_name_launch.py

---

Step 7: Build and Run

Build Package

# Navigate to ROS 2 workspace
cd ~/ros2_ws/src

# Copy generated package
cp -r /path/to/generated/package .

# Build
cd ~/ros2_ws
colcon build --packages-select model_name

# Source workspace
source install/setup.bash

Run Node

# Launch the controller node
ros2 launch model_name model_name_launch.py

---

Complete Export Workflow

flowchart TB
    subgraph Simulink["Simulink Environment"]
        MODEL[Simulink Model]
        CONFIG[Configure for<br/>Code Generation]
        ROS2_BLK[Add ROS 2<br/>Blocks]
    end

    subgraph CodeGen["Code Generation"]
        GEN[Generate C++<br/>Code]
        PKG[Create ROS 2<br/>Package]
    end

    subgraph ROS2["ROS 2 Environment"]
        BUILD[colcon build]
        RUN[ros2 run]
        ROBOT[Physical<br/>Robot]
    end

    MODEL --> CONFIG
    CONFIG --> ROS2_BLK
    ROS2_BLK --> GEN
    GEN --> PKG
    PKG --> BUILD
    BUILD --> RUN
    RUN --> ROBOT

    style Simulink fill:#e3f2fd
    style CodeGen fill:#fff8e1
    style ROS2 fill:#e8f5e9

---

Troubleshooting

Issue	Solution
Build fails	Check CMakeLists.txt dependencies
Node crashes	Verify message types match
No data received	Check topic names and QoS settings
Timing issues	Adjust node execution rate

---

References

• Webots Documentation
• Simulink Documentation
• ROS 2 Documentation
• MATLAB ROS 2 Examples
• Simulink Coder for ROS 2

---

Next Steps

After successfully deploying your controller, consider:

• Adding parameter tuning via ROS 2 parameters
• Implementing diagnostics and logging
• Setting up launch files for multi-node systems
• Creating RViz visualizations for debugging