PIERO Mobile Robot — Model-Based Control & Autonomous Navigation

A modular MATLAB/Simulink control stack for a differential-drive robot, covering the path from motor identification and low-level control to odometry, autonomous path following, and reactive obstacle avoidance. The same control architecture can switch between an identified plant model and an Arduino Mega 2560 hardware target.

This project was developed during a 2023 Mobile Robotics Laboratory course and later extended with a scanning ultrasonic sonar.

What this project demonstrates

Area	Implementation
Embedded robotics	Arduino I/O, PWM motor actuation, encoder acquisition, and fixed-step execution
System identification	Discrete motor transfer functions derived from measured encoder responses
Feedback control	Per-wheel closed-loop speed control and robot orientation control
Robot modelling	Direct/inverse differential-drive kinematics and odometry
Autonomous navigation	Pure Pursuit waypoint tracking and a custom MATLAB Function alternative
Reactive behaviour	Stateflow-based obstacle avoidance using infrared distance sensors
Model-Based Design	Reusable subsystem references and a hardware/simulation variant architecture
Experimental extension	Servo-driven ultrasonic scanning, obstacle localisation, and serial polar visualisation

System architecture

flowchart LR
    WP[Waypoints] --> PP[Pure Pursuit]
    IR[Infrared sensors] --> OA[Stateflow obstacle avoidance]
    SONAR[Optional scanning sonar] -.-> OA
    PP --> ARB[Command arbitration]
    OA --> ARB
    ARB --> MCI[Inverse kinematics]
    MCI --> CTRL[Wheel-speed control]
    CTRL --> VAR{Plant variant}
    VAR -->|Simulation| TF[Identified motor models]
    VAR -->|Hardware| HW[Arduino + motors]
    HW --> ENC[Wheel encoders]
    TF --> MCD[Direct kinematics]
    ENC --> MCD
    MCD --> ODO[Odometry]
    ODO --> PP

Loading

The separation between navigation, kinematics, control, plant, and sensing makes it possible to test algorithms in simulation before deploying the same model structure to the robot. See the model and architecture guide for the complete inventory.

Results

The final model combines waypoint tracking, odometry, speed control, infrared sensing, a Stateflow avoidance supervisor, and a simulation/hardware plant variant.

The optional extension mounts an ultrasonic sensor on a servo and visualises angle–distance measurements as a polar scan:

• Watch the sonar scan demo
• Watch the integrated sonar navigation demo

Getting started

Software

The models were authored in MATLAB/Simulink R2023b. Use R2023b or a newer compatible release with:

• MATLAB and Simulink
• Robotics System Toolbox for the Pure Pursuit block
• Simulink Support Package for Arduino Hardware for deployment and Arduino blocks
• Stateflow for the behaviour supervisors

System Identification Toolbox is useful for repeating the motor-identification workflow, but the identified motor models are already included.

Open the project

1. Clone or download the repository.
2. Open MATLAB in the repository root.
3. Open AngelPieroProject.prj.
4. Start with Modelos/PieroFinalProject.slx for the integrated baseline.
5. Select the simulated or hardware plant variant in the model before running it.

From the MATLAB Command Window:

openProject("AngelPieroProject.prj");
open_system("Modelos/PieroFinalProject.slx");

To check that every committed model can be loaded:

addpath("scripts");
results = validateProject;
assert(all(results.Loaded));

Hardware deployment additionally requires the PIERO platform (or equivalent differential-drive hardware), an Arduino Mega 2560, motors with encoders, motor drivers, and compatible distance sensors.

Repository guide

Path	Contents
Modelos/	Top-level experiments and integrated navigation models
Subsistemas/	Reusable control, plant, sensing, kinematics, and odometry components
Data/	Encoder S-function source, wrapper, TLC file, and Windows MEX binary
Funciones/	MATLAB sonar visualisation utility
Tareas/	Eight illustrated engineering notes in Spanish
docs/	Architecture guide and the original weekly development journal
scripts/	Non-destructive model-load validation

Project notes

• Model and architecture guide
• Engineering tasks / prácticas (Spanish)
• Optional sonar project (Spanish)
• Original development journal (Spanish)

Scope and authorship

This repository is Ángel Muñoz Rocha's personal project record from a collaborative university laboratory. It includes his development notes, model integration work, Arduino/Simulink investigation, sensor work, and the individual sonar extension. The journal preserves the original team context and contributions.