Thymio II Educational Robot

This example demonstrates the Thymio II robot with MATLAB/Simulink.

---

Overview

Property	Value
Type	Educational Mobile Robot
Robot	Mobsya Thymio II
Difficulty	Beginner
Control Method	Differential Drive / Reactive

---

Features

• Educational Design: Easy to learn
• 7 IR Proximity Sensors: 5 front, 2 rear
• Ground Sensors: Line following
• LED Feedback: Visual indicators

---

Specifications

Property	Value
Dimensions	112 x 117 x 53 mm
Weight	270 g
Max Speed	14 cm/s
Wheel Radius	22 mm
Wheel Base	95 mm

---

Control Architecture

System Block Diagram

flowchart TB
    subgraph Sensors["Sensor Suite"]
        subgraph Proximity["Proximity Sensors"]
            PF["5 Front<br/>IR Sensors"]
            PR["2 Rear<br/>IR Sensors"]
        end
        subgraph Ground["Ground Sensors"]
            GL["Left Ground"]
            GR["Right Ground"]
        end
    end

    subgraph Behaviors["Behavior Layer"]
        LINE["Line<br/>Following"]
        AVOID["Obstacle<br/>Avoidance"]
        EXPLORE["Random<br/>Exploration"]
    end

    subgraph Arbitration["Behavior Arbitration"]
        PRIO["Priority<br/>Selection"]
    end

    subgraph Actuators["Actuator System"]
        ML["Left Motor"]
        MR["Right Motor"]
        LED["RGB LEDs"]
    end

    PF & PR --> AVOID
    GL & GR --> LINE
    LINE --> PRIO
    AVOID --> PRIO
    EXPLORE --> PRIO
    PRIO --> ML & MR & LED

    style Sensors fill:#e1f5fe
    style Behaviors fill:#e8f5e9
    style Arbitration fill:#fff3e0
    style Actuators fill:#fce4ec

Reactive Control Architecture

flowchart LR
    subgraph Stimulus["Stimuli"]
        S1["Proximity<br/>Detection"]
        S2["Ground<br/>Color"]
        S3["Button<br/>Press"]
    end

    subgraph Response["Responses"]
        R1["Turn Away"]
        R2["Follow Line"]
        R3["Change Mode"]
    end

    subgraph Output["Motor Output"]
        VL["v_left"]
        VR["v_right"]
    end

    S1 --> R1 --> VL & VR
    S2 --> R2 --> VL & VR
    S3 --> R3

    style Stimulus fill:#ffcdd2
    style Response fill:#c8e6c9
    style Output fill:#bbdefb

Sensor Layout

flowchart TB
    subgraph ThymioTop["Thymio II - Top View"]
        subgraph FrontSensors["Front Sensors"]
            F0["prox.horizontal[0]<br/>Far Left"]
            F1["prox.horizontal[1]<br/>Left"]
            F2["prox.horizontal[2]<br/>Center"]
            F3["prox.horizontal[3]<br/>Right"]
            F4["prox.horizontal[4]<br/>Far Right"]
        end
        subgraph RearSensors["Rear Sensors"]
            R5["prox.horizontal[5]<br/>Left Rear"]
            R6["prox.horizontal[6]<br/>Right Rear"]
        end
        subgraph GroundSensors["Ground Sensors"]
            G0["prox.ground[0]<br/>Left"]
            G1["prox.ground[1]<br/>Right"]
        end
    end

    style FrontSensors fill:#e1f5fe
    style RearSensors fill:#fff3e0
    style GroundSensors fill:#c8e6c9

---

State-Space Model

Differential Drive Kinematics

% State vector: x = [x, y, theta]'
% Input vector: u = [v_left, v_right]'

% Thymio parameters
r = 0.022;      % Wheel radius [m]
L = 0.095;      % Wheel base [m]
v_max = 0.14;   % Max velocity [m/s]

% Kinematic equations
% x_dot = (v_left + v_right)/2 * cos(theta)
% y_dot = (v_left + v_right)/2 * sin(theta)
% theta_dot = (v_right - v_left) / L

% Linear and angular velocity
v = (v_left + v_right) / 2;       % Linear velocity [m/s]
omega = (v_right - v_left) / L;    % Angular velocity [rad/s]

% State derivatives
x_dot = v * cos(theta);
y_dot = v * sin(theta);
theta_dot = omega;

Sensor Model

% Proximity sensor model (returns 0-4500)
% Higher values = closer object

% Sensor angles (from robot frame)
prox_angles = [-40, -20, 0, 20, 40, 150, -150] * pi/180;  % [rad]

% Sensor range
prox_range = 0.12;  % 12 cm max range

% Ground sensor model (returns 0-1023)
% Higher values = white surface
% Lower values = black surface
ground_threshold = 400;  % Black line detection

---

Behavior Design

Line Following

flowchart TB
    subgraph LineFollow["Line Following Behavior"]
        GL["Ground Left"] --> DIFF["Error<br/>e = GL - GR"]
        GR["Ground Right"] --> DIFF
        DIFF --> CTRL["Proportional<br/>Controller"]
        CTRL --> STEER["Steering<br/>Correction"]
    end

    subgraph MotorCalc["Motor Calculation"]
        BASE["Base<br/>Speed"] --> VL["v_left = base - steer"]
        STEER --> VL
        BASE --> VR["v_right = base + steer"]
        STEER --> VR
    end

    style LineFollow fill:#e8f5e9
    style MotorCalc fill:#fff3e0

Line Following Controller

function [v_left, v_right] = line_following(ground_left, ground_right, params)
    % Simple line following using ground sensors

    % Error: positive = line is to the left
    error = ground_left - ground_right;

    % Proportional control
    k_p = params.k_line;
    correction = k_p * error;

    % Base speed
    v_base = params.v_line;

    % Motor commands
    v_left = v_base - correction;
    v_right = v_base + correction;

    % Line lost detection
    if ground_left < params.threshold && ground_right < params.threshold
        % Both sensors see black - on the line
    elseif ground_left > params.threshold && ground_right > params.threshold
        % Both sensors see white - line lost, search
        v_left = v_base;
        v_right = -v_base;  % Turn in place
    end

    % Saturation
    v_max = params.v_max;
    v_left = max(-v_max, min(v_max, v_left));
    v_right = max(-v_max, min(v_max, v_right));
end

Obstacle Avoidance

function [v_left, v_right] = obstacle_avoidance(prox_sensors, params)
    % Braitenberg-style obstacle avoidance

    % Sensor weights for left motor
    w_left = [0.5, 0.25, -0.25, -0.5, -0.75, 0, 0];

    % Sensor weights for right motor
    w_right = [-0.75, -0.5, -0.25, 0.25, 0.5, 0, 0];

    % Normalize sensor values (0-1)
    prox_norm = prox_sensors / 4500;

    % Compute motor values
    v_left = params.v_base + sum(w_left .* prox_norm) * params.k_avoid;
    v_right = params.v_base + sum(w_right .* prox_norm) * params.k_avoid;

    % Saturation
    v_left = max(0, min(params.v_max, v_left));
    v_right = max(0, min(params.v_max, v_right));
end

---

Simulink Integration

Initialization Script

TIME_STEP = 100;  % Thymio uses slower update rate

% Initialize proximity sensors
prox = zeros(1, 7);
for i = 0:6
    prox(i+1) = wb_robot_get_device(['prox.horizontal.' num2str(i)]);
    wb_distance_sensor_enable(prox(i+1), TIME_STEP);
end

% Initialize ground sensors
ground = zeros(1, 2);
for i = 0:1
    ground(i+1) = wb_robot_get_device(['prox.ground.' num2str(i)]);
    wb_distance_sensor_enable(ground(i+1), TIME_STEP);
end

% Initialize motors
left_motor = wb_robot_get_device('motor.left');
right_motor = wb_robot_get_device('motor.right');
wb_motor_set_position(left_motor, inf);
wb_motor_set_position(right_motor, inf);
wb_motor_set_velocity(left_motor, 0);
wb_motor_set_velocity(right_motor, 0);

% Behavior parameters
params.v_max = 7.0;           % Max motor speed (Thymio units)
params.v_base = 4.0;          % Base speed
params.k_line = 0.005;        % Line following gain
params.k_avoid = 0.02;        % Obstacle avoidance gain
params.threshold = 400;       % Ground sensor threshold

Behavior-Based Control Loop

% Behavior state
current_behavior = 'explore';

while wb_robot_step(TIME_STEP) ~= -1
    % Read proximity sensors
    prox_values = zeros(1, 7);
    for i = 1:7
        prox_values(i) = wb_distance_sensor_get_value(prox(i));
    end

    % Read ground sensors
    ground_values = zeros(1, 2);
    for i = 1:2
        ground_values(i) = wb_distance_sensor_get_value(ground(i));
    end

    % Behavior selection (priority-based)
    front_obstacle = max(prox_values(1:5)) > 1000;
    line_detected = ground_values(1) < params.threshold || ...
                    ground_values(2) < params.threshold;

    % Priority: Obstacle > Line > Explore
    if front_obstacle
        [v_left, v_right] = obstacle_avoidance(prox_values, params);
        current_behavior = 'avoid';
    elseif line_detected
        [v_left, v_right] = line_following(ground_values(1), ground_values(2), params);
        current_behavior = 'line';
    else
        [v_left, v_right] = random_explore(params);
        current_behavior = 'explore';
    end

    % Apply motor commands
    wb_motor_set_velocity(left_motor, v_left);
    wb_motor_set_velocity(right_motor, v_right);
end

function [v_left, v_right] = random_explore(params)
    % Random walk behavior
    persistent timer turn_dir
    if isempty(timer)
        timer = 0;
        turn_dir = 1;
    end

    timer = timer + 1;
    if timer > 50  % Change direction periodically
        timer = 0;
        turn_dir = sign(randn);
    end

    v_left = params.v_base + turn_dir * 1;
    v_right = params.v_base - turn_dir * 1;
end

---

Educational Activities

Activity 1: Follow the Line

% Simple line follower
while wb_robot_step(TIME_STEP) ~= -1
    g_left = wb_distance_sensor_get_value(ground(1));
    g_right = wb_distance_sensor_get_value(ground(2));

    % Both on line (black)
    if g_left < 400 && g_right < 400
        v_left = 4; v_right = 4;
    % Line to the left
    elseif g_left < 400
        v_left = 2; v_right = 5;
    % Line to the right
    elseif g_right < 400
        v_left = 5; v_right = 2;
    % Line lost
    else
        v_left = 3; v_right = -3;  % Turn to search
    end

    wb_motor_set_velocity(left_motor, v_left);
    wb_motor_set_velocity(right_motor, v_right);
end

Activity 2: Wall Following

% Right-wall follower
while wb_robot_step(TIME_STEP) ~= -1
    front = wb_distance_sensor_get_value(prox(3));
    right = wb_distance_sensor_get_value(prox(5));

    if front > 2000        % Obstacle in front
        v_left = 3; v_right = -3;  % Turn left
    elseif right > 1500    % Too close to wall
        v_left = 4; v_right = 2;   % Turn slightly left
    elseif right < 500     % Too far from wall
        v_left = 2; v_right = 4;   % Turn slightly right
    else                   % Good distance
        v_left = 4; v_right = 4;   % Go straight
    end

    wb_motor_set_velocity(left_motor, v_left);
    wb_motor_set_velocity(right_motor, v_right);
end

---

Quick Start

1. Open Webots and load examples/thymio2/worlds/thymio_playground.wbt
2. Configure MATLAB controller
3. Run simulation

---

LED Colors

Color	Behavior
Green	Line Following
Red	Obstacle Detected
Blue	Exploring
Yellow	Searching

---

References

• Thymio Official
• Webots Thymio II
• Mondada, F. et al. (2017). "Thymio: A Robot for Learning and Research"