Welcome to this Accellerato Program.

The goal of this program is very simple: bring you straight to the practical part of modern robotics engineering.

This is not a long theoretical course where we only explain concepts in isolation. Here, we focus on the complete pipeline: from the concept of a robotic application, to simulation, to software architecture, and finally to real deployment on industrial hardware.

By the end of this course, you should have a clear understanding of how to build a realistic robotics application using ROS2, MoveIt, Gazebo, vision systems, edge computing, and a real robot.

The course is organized around three main blocks.

First, we start with the ROS2 migration and industrial simulation pipeline.

In this module, we take a robotics application and we structure it properly in ROS2. We work with robot description packages, MoveIt2 configuration, ROS2 Control, Gazebo simulation, depth cameras, point clouds, inverse kinematics, and pick-and-place logic.

The objective is to understand how to simulate a robot application in a way that is useful for real industrial development.

Then, we move into the second block, where we explore how to interface robotics applications with LLMs and AI tools.

The goal here is to understand how language models can become part of a robotics workflow: not as magic, but as software components that can help us reason, generate commands, interact with the system, and support more intelligent robot applications.

Finally, we move to the real deployment pipeline.

Here we use a real industrial setup based on:

FR3WML robot
Jetson Orin Nano
Intel RealSense D455
ROS2
MoveIt
YOLO-based vision
Behavior Trees
industrial software architecture

This is where the course becomes very concrete.

We look at the hardware architecture, the software architecture, the Jetson containers, the RealSense streaming, the inference pipeline, 3D and 6D pose estimation, and the transition from simulation to a real robot.

The final objective is that you understand how all these pieces communicate together.

The PC runs the main ROS2 orchestration and MoveIt stack.

The Jetson runs the camera streaming and vision inference.

The RealSense provides RGB, depth, and point cloud data.

The robot controller receives motion commands through the bridge layer.

And the application logic is structured using Behavior Trees, which is one of the best ways to keep industrial robot code modular, readable, and scalable.

Some lessons in this course may reference concepts that are explained in more detail in the complete program,

Robotics Software Engineering Library — Full Industrial Stack (learn-robotics-with-ros-s-school.teachable.com/p/full-robotics-program)

That full program also covers the foundations of ROS, ROS2 programming, publishers, subscribers, services, actions, custom interfaces, and lower-level robotics software development.

But you do not need to complete that full program before taking this accelerator.

Here, the focus is different.

This course is designed to be straight to the point.

You will see how the pieces are connected in a real application and how a robotics software engineer thinks when building a complete system.

So the expected output of this course is not just that you “watch lessons”.

The expected output is that you become able to look at an industrial robotics problem and understand how to structure the solution:

robot simulation
motion planning
camera integration
AI inference
ROS2 communication
real robot deployment
industrial application logic

This is the mindset we want to build.

You are not only learning how to move a robot from point A to point B.

You are learning how to design and deploy a complete robotics application.

So welcome to the course.