Course Overview

This intensive 42-hour course focuses on the implementation of Tiny Machine Learning (TinyML) algorithms within the control systems of mechatronic systems, integrating knowledge of mechanics, electronics, and advanced control. Through the Model-Based Design approach, students will learn to develop, simulate, and implement intelligent controllers using MATLAB, Simulink, and hardware such as the Arduino Nano 33 BLE Sense and Dynamixel AX12 servomotors. Combining theory and hands-on practice, this course provides participants with modern tools to integrate artificial intelligence into the design of control systems for industrial and research applications.

Learning Outcomes

By the end of the course, students will be able to:
• Design mechatronic systems with an integrated approach to mechanics, electronics, and control.
• Incorporate TinyML algorithms into the implementation of intelligent controllers.
• Model, simulate, and validate mechatronic control systems using MATLAB and Simulink.
• Program and implement TinyML-based controllers on Arduino Nano 33 BLE Sense hardware.
• Optimize the performance of TinyML-controlled systems by integrating sensors and actuators.
• Apply validation and evaluation techniques to measure the performance of intelligent controllers.

Course Content

1. Introduction to Mechatronic Systems:
Basic concepts and architecture of a mechatronic system.
Interaction between mechanics, electronics, and control.
2. TinyML for Intelligent Controllers:
Fundamentals of Tiny Machine Learning.
Applications of TinyML in system control.
3. Model-Based Design:
Fundamentals of modeling and simulation.
Using MATLAB and Simulink for creating dynamic models.
4. Intelligent Controllers with TinyML:
Designing predictive control algorithms using neural networks.
Training and optimizing data-driven control models.

5. Implementation in Hardware:
Programming the Arduino Nano 33 BLE Sense to execute TinyML controllers.
Integration with Dynamixel AX12 servomotors and sensors.
6. System Testing and Validation:
Methods for evaluating the accuracy and stability of TinyML controllers.
Optimizing energy consumption and system latency.
7. Final Project:
Design, implementation, and presentation of a mechatronic system with TinyMLbased intelligent control.

Instructional Method

The course will employ:
• Theoretical classes to explain the fundamental principles of mechatronics, control, and TinyML.
• Practical laboratories using MATLAB, Simulink, and hardware for the implementation and testing of intelligent controllers.
• Case studies of real-world applications of TinyML in mechatronic systems.

• A final group project to consolidate knowledge through the design and development of an intelligent control system.

Required Course Materials

Software:
MATLAB and Simulink with Toolboxes such as Deep Learning and Control System,
Arduino IDE.
Hardware:
Arduino Nano 33 BLE Sense.
Dynamixel AX12 servomotors.
Sensors for data collection (e.g., accelerometers, gyroscopes, temperature sensors).

Assessment

The evaluation is designed to assess the achievement of the course objectives, using a combination of practical activities, theoretical tasks, and a final project. This summative assessment ensures that students acquire and apply the key knowledge and skills from the course.

Participation in Practical Activities (20%). Ensures direct application of concepts in controlled environments using tools like MATLAB, Simulink, and the proposed hardware.
Individual Assignments (30%). Measures individual progress in mastering key topics: modeling, control, and TinyML.

Case Study Analysis (10%). Reinforces the connection between theory and real-world applications, stimulating critical analysis and problem-solving skills.
Final Group Project (40%). Integrally evaluates the knowledge acquired throughout the course, encouraging teamwork, creativity, and practical problem resolution in a mechatronic system using TinyML.

Evaluation Criteria	Description	Weight
Participation in Practical Activities	Assessment of performance in laboratory sessions using MATLAB, Simulink, and hardware (Arduino Nano 33 BLE Sense and Dynamixel AX12 servomotors).	20%
Individual Assignments	Activities focused on modeling, simulation, and development of control algorithms and TinyML using MATLAB and Simulink. Each assignment will assess a specific learning objective.	30%
Case Study Analysis	Analysis and discussion of real-world TinyML applications in mechatronic systems. The evaluation will focus on understanding and linking theoretical concepts to practical examples.	10%
Final Group Project	Design, implementation, and presentation of a mechatronic system with TinyML-based intelligent control. The project will be evaluated on the following dimensions:	40%
	Design and Modeling: Clarity and accuracy of the developed dynamic model.	10%
	- Controller Implementation: Functional integration of TinyML controllers with Arduino Nano 33 BLE Sense and Dynamixel servomotors.	10%
	- Verification and Validation: Evaluation of requirements performance, accuracy, and system stability.	10%
	- Final Presentation: Quality of the presentation and project documentation.	10%