What you need to follow this course and get the maximum value from it.

Objective

In this lesson, I want to clarify what you need before starting the course.

The first two modules are mainly software-based. You can follow them with only a computer.

The final module is different. There we move from simulation to real hardware deployment, using a real robot, a 3D camera, a Jetson edge device, grippers, pneumatic components, and ROS2 communication across multiple machines.

So let’s separate the requirements into two levels: software requirements and hardware requirements.

Software Requirements

For Module 1 and Module 2, you only need a good development computer.

I recommend using a PC with:

Ubuntu 22.04 installed in dual boot
ROS2 Humble installed
Gazebo Classic
MoveIt2
VS Code
Git
Docker
At least 100 GB of free disk space
Recommended: 150–200 GB free

The reason I recommend Ubuntu 22.04 is simple: ROS2 Humble is very well supported on Ubuntu 22.04, and most of the simulation and MoveIt2 tools we use in this course are based on that environment.

In the first module, we will work with simulation, ROS2 packages, robot descriptions, Gazebo, MoveIt2, controllers, inverse kinematics, and pick-and-place logic.

In the second module, we will install and use Ollama, so make sure you have enough disk space. Local LLM models can take several gigabytes each, so do not start this course with a full hard drive.

As a minimum, I suggest 100 GB free. A more comfortable setup is 150 to 200 GB free.

I provide a link to my YouTube video where I explain how to install ROS2 Humble and set up the basic ROS2 environment: https://youtu.be/NVoZlrlyVBY

Hardware Requirements for the Final Module

The final module is where things become much more industrial.

Here, we use one of the most important features of ROS2: the possibility to build a robot-agnostic application layer.

This means that the high-level logic of the application should not depend directly on one specific robot. The packages we use are designed to be adapted to different robots by modifying the bridge layer, the hardware interface, and the configuration files.

So technically, you can follow the final module with a different robot.

But to get the maximum benefit and reproduce exactly what I do, I recommend using the same hardware setup.

Recommended BOM

1x PC with Ubuntu 22.04 and ROS2 Humble
1x FR3WML robot
1x SoftGripper with two fingers
1x Suction cup end effector
1x Gripper controller
1x Jetson Orin Nano
1x Intel RealSense D455
1x Industrial Ethernet switch
At least 3x Ethernet cables
1x 220 V AC power supply setup
1x Compressed air supply
Pneumatic tubing and fittings
Basic mechanical mounting plates/brackets

Hardware Architecture

The PC runs the main ROS2 application, MoveIt2, RViz, and the orchestration logic.

The Jetson Orin Nano is used as the edge device for vision and inference.

The RealSense D455 is connected to the Jetson and provides RGB and depth data.

The robot controller is connected through Ethernet and receives motion commands through the bridge layer.

The gripper and suction cup are controlled through dedicated outputs or a gripper controller.

The industrial switch keeps all the devices on the same network.

This gives us a realistic industrial architecture, close to what you would use in a real automation cell.

Can You Use a Different Robot?

Yes.

The final module can be followed with another robot, as long as you understand that you will need to adapt some parts.

The most important parts to adapt are:

Robot description package
MoveIt2 configuration
Controller interface
Bridge layer
Hardware interface
Robot SDK commands
Tool frame and TCP configuration
Gripper or suction control
Scene parameters

The application layer is designed to stay as generic as possible. The robot-specific part should stay in the bridge and hardware interface.

This is exactly the mindset of industrial robotics software engineering: isolate the vendor-specific logic and keep the application logic reusable.

Optional Hardware Kit

To make the final module easier to reproduce, I can also provide a complete hardware kit with the same components used in the course.

The estimated cost of the complete setup is around:

9,000 – 10,000 €

This depends on availability, configuration, shipping, and exact hardware options (delivery in 1 month).

For a quotation, contact me at:

ros.master.ai@gmail.com

Important Note

This course is not designed as a hobbyist robotics course.

The goal is not just to play with a robot in simulation.

The goal is to help you think and work like a robotics software engineer who can build real industrial automation solutions.

We will go from simulation to deployment, from MoveIt2 to real robot motion, from camera data to 3D and 6D pose estimation, from simple pick-and-place to a structured robotics application.

So the best way to approach this course is to see it as a professional training program.

You are learning how to build systems that can solve real automation tasks in real industrial environments.

Key Takeaways

To follow the first two modules, you mainly need a good Ubuntu 22.04 computer with ROS2 Humble, Gazebo, MoveIt2, Docker, and enough disk space.

For the final module, you can use your own robot, but the best learning experience comes from using the same hardware architecture: PC, FR3WML robot, Jetson Orin Nano, RealSense D455, gripper, suction cup, and industrial network.

The most important concept is this: the application should be robot-agnostic, while the bridge layer and hardware interface are the parts you adapt to your specific robot.

That is the foundation for building reusable industrial robotics software.