Skip to content

Connecting Webots and Simulink

This guide explains how to establish communication between Webots and Simulink using the bridge interface.

Communication Architecture

The Webots-Simulink Bridge uses a synchronous communication model where:

  1. Webots acts as the physics simulation engine, handling robot dynamics and sensor simulation
  2. Simulink acts as the controller, processing sensor data and generating actuator commands
  3. MATLAB functions serve as the interface layer, calling Webots API functions
flowchart LR
    subgraph Simulink["Simulink Model"]
        CTRL[Controller<br/>Blocks]
        SENSOR[Sensor<br/>Input Blocks]
        ACTUATOR[Actuator<br/>Output Blocks]
    end

    subgraph MATLAB["MATLAB Interface"]
        WB_GET[wb_*_get_value<br/>Functions]
        WB_SET[wb_*_set_*<br/>Functions]
        WB_STEP[wb_robot_step<br/>Function]
    end

    subgraph Webots["Webots Simulator"]
        ROBOT[Robot<br/>Model]
        PHYSICS[Physics<br/>Engine]
        SENSORS[Sensor<br/>Simulation]
    end

    SENSOR --> WB_GET
    WB_GET --> SENSORS
    SENSORS --> PHYSICS
    PHYSICS --> ROBOT
    ROBOT --> ACTUATOR
    ACTUATOR --> WB_SET
    WB_SET --> ROBOT
    CTRL --> WB_STEP
    WB_STEP --> PHYSICS

    style Simulink fill:#e3f2fd
    style MATLAB fill:#fff8e1
    style Webots fill:#e8f5e9

Step 1: Configure Webots World

Open your Webots world file (.wbt) and set MATLAB as the external controller:

  1. Select the robot in the scene tree
  2. Set the controller field to <extern>
  3. Save the world file
Robot {
  controller "<extern>"
  ...
}

Step 2: Initialize Sensor Variables

Define sensor device variables in your MATLAB controller script:

matlab1

% Initialize Webots robot
TIME_STEP = 16;  % Simulation time step in milliseconds

% Get device handles
gps = wb_robot_get_device('gps');
imu = wb_robot_get_device('inertial unit');
gyro = wb_robot_get_device('gyro');
accelerometer = wb_robot_get_device('accelerometer');

% Enable sensors with specified sampling period
wb_gps_enable(gps, TIME_STEP);
wb_inertial_unit_enable(imu, TIME_STEP);
wb_gyro_enable(gyro, TIME_STEP);
wb_accelerometer_enable(accelerometer, TIME_STEP);

Set the Simulink solver parameters to match Webots timing:

matlab2

Solver Configuration

Parameter Value Description
Solver Type Fixed-step Required for synchronous execution
Solver ode4 (Runge-Kutta) Fourth-order integration method
Fixed Step Size 0.016 Matches Webots TIME_STEP (16 ms)
Stop Time inf Continuous simulation 
% Configure Simulink model programmatically
set_param('model_name', 'Solver', 'ode4');
set_param('model_name', 'FixedStep', '0.016');
set_param('model_name', 'StopTime', 'inf');

Build your control system using Simulink blocks:

matlab4

Sensor Input Block

Create a MATLAB Function block to read sensor values:

function [position, orientation, angular_velocity] = read_sensors()
    % Read GPS position
    position = wb_gps_get_values(gps);

    % Read IMU orientation (roll, pitch, yaw)
    orientation = wb_inertial_unit_get_roll_pitch_yaw(imu);

    % Read gyroscope angular velocity
    angular_velocity = wb_gyro_get_values(gyro);
end

Actuator Output Block

Create a MATLAB Function block to send motor commands:

function send_motor_commands(motor_velocities)
    % Set motor velocities
    wb_motor_set_velocity(motor1, motor_velocities(1));
    wb_motor_set_velocity(motor2, motor_velocities(2));
    wb_motor_set_velocity(motor3, motor_velocities(3));
    wb_motor_set_velocity(motor4, motor_velocities(4));
end

Step 5: Synchronize Simulation Steps

The wb_robot_step function synchronizes Webots and Simulink:

function step_simulation()
    % Advance simulation by one time step
    if wb_robot_step(TIME_STEP) == -1
        % Simulation ended
        return;
    end
end

This function must be called once per Simulink time step to:

  1. Send actuator commands to Webots
  2. Advance the physics simulation
  3. Update sensor readings

Step 6: Connect to Robot

connecting1

Load and initialize the Simulink model:

% Load Simulink model
model_name = 'robot_controller';
load_system(model_name);

% Open model for visualization
open_system(model_name);

% Initialize robot connection
wb_robot_init();

% Start simulation
set_param(model_name, 'SimulationCommand', 'start');

Workspace Variables

matlab3

Monitor variables in the MATLAB workspace during simulation:

Variable Type Description
TIME_STEP Integer Simulation step in milliseconds
gps Device Handle GPS sensor reference
imu Device Handle Inertial measurement unit reference
motor1..4 Device Handle Motor actuator references
position [3x1] Double Current robot position (x, y, z)
orientation [3x1] Double Current orientation (roll, pitch, yaw)

Data Flow Summary

Direction Data Type Example Variables
Webots → Simulink Sensor readings Position, velocity, orientation, distances
Simulink → Webots Actuator commands Motor velocities, torques, positions
Bidirectional Timing synchronization wb_robot_step() return value

Next Steps