2027/28 course

Advanced robotics with ROS 2

Third and final course of the ROS 2 Continuing Education program. It assumes Introduction and Intermediate have been mastered and pushes the same autonomous delivery robot (RobotDelivery) to its limits: an embedded brain with micro-ROS, manipulation with MoveIt 2, containerization, digital twin, cloud-connected fleet, AI perception and reinforcement learning, IoT connectivity, and reproducible evaluation and documentation of the robot's work.

Practice repository on GitHub

Syllabus

Course structure

Topic	Title	Practical content
1	An embedded brain	micro-ROS on a microcontroller (rclc, serial/UDP transport)
2	Learning to plan	Arm manipulation with MoveIt 2 and MoveIt Task Constructor
3	Traveling in containers	Containerization with Docker: multi-stage, multi-architecture, security
4	Shadow and digital twin	Synchronization, anomaly detection and fidelity validation
5	A connected fleet	Web dashboard (rosbridge), fleet coordination with Open-RMF, containerized deployment
6	Seeing and deciding on its own	AI-based classification and reinforcement learning (Gymnasium, Q-Learning)
7	Connected to the IoT world	MQTT and FIWARE (NGSI) for IoT connectivity
8	Documenting and evaluating	rosbag2, synthetic data and reproducible benchmarking

Verifiable practice

Validation and demo repository

The ros2-avanzado-uned repository contains the automatic validation package for all 8 topics and example/demo packages for each one. It is a living repository: I will keep expanding its documentation and content throughout the course.

This course depends on third-party infrastructure (micro-ROS, MoveIt Task Constructor, Docker/Kubernetes, rosbridge, TensorFlow/Gymnasium, MQTT/FIWARE); the corresponding validators fail with a clear diagnostic message if they are not installed. The packages specific to this course use the _avanzado suffix (e.g. tema1_practica_avanzado) so they can coexist in the same workspace as those from the Intermediate course, which reuses several of its exercises.