Teaching Industry 4.0: An Open-Source Framework for Cooperative Robotics in the Learning Factory

Modern factories are changing. They are becoming places where machines talk to each other, make their own decisions, and move around without human drivers. This shift, often called the fourth industrial revolution, relies on three main ideas working together. First, there are "cyber-physical systems," which simply means physical machines like robots that are connected to digital brains. Second, the "Internet of Things" allows these machines to send data about their status and surroundings over the internet. Third, artificial intelligence helps the machines see and understand what they are looking at. For engineers of the future, learning how to build and manage these talking, thinking machines is no longer optional; it is essential. However, teaching these complex skills in a classroom is difficult. Real factory robots are expensive, and the software that makes them work is often too complicated for students to tinker with. Schools need a way to show students how these different parts fit together without breaking the bank.

A team of researchers at the Monterrey Institute of Technology and Higher Education in Mexico has built a solution to this problem. They created a working model of a smart factory cell that students can build, break, and fix. This project is not just a simulation on a computer screen; it is a physical setup where real robots move real objects. The goal was to prove that a low-cost, open-source system could handle a complex task: picking up items, identifying their color, and sorting them into different bins, all while keeping a digital record of every move. The researchers wanted to see if this kind of hands-on project could actually teach students how to integrate robotics, internet connectivity, and artificial intelligence.

The system they built relies on two main robots working in a team. One is a mobile robot, an Omron LD-60 Autonomous Intelligent Vehicle (AIV) that drives itself around the floor. The other is a stationary Omron TM5-900 collaborative robot that sits at a table. To make them work together, the researchers connected them through a network that acts like a nervous system, allowing the robots to send messages back and forth instantly. When the mobile robot arrives at the table, it stops and waits. The robotic arm then reaches out to grab a block. Before the arm moves the block, a camera takes a picture of it. This is where the artificial intelligence comes in. Instead of using a simple, pre-programmed rule to guess the color, the system sends the picture to a powerful cloud-based AI service. The AI looks at the image and tells the system exactly what color the block is.

Once the color is known, the system makes a decision. If the block is green, the mobile robot is told to drive to the green waypoint. If it is blue, it goes to the blue waypoint. The robotic arm then picks up the block and places it onto the mobile robot, which drives to the correct location to drop it off. Throughout this entire process, every single action is logged into a cloud database. This means that anyone with internet access can watch a live dashboard showing exactly where the robots are, what they are doing, and if any errors occurred. The researchers designed the parts of the robot, like the gripper that holds the blocks, using 3D printers, keeping the cost of the hardware very low. They made all the computer code and design plans available for free, so other schools can copy the setup exactly.

The team tested this system rigorously. They ran the sorting process one hundred times in a row to see how often it worked without failing. The robots successfully completed the task ninety-four percent of the time. When they looked specifically at how well the artificial intelligence identified the colors, it was correct ninety-four point two percent of the time. The system was also fast; the time it took for the robots to send messages to each other was measured at thirty-two milliseconds, which is a fraction of a second. The researchers also measured how much the students learned. Before the project, the students rated their understanding of these technologies on a scale. After building and running the system, their self-reported confidence and knowledge jumped significantly.

The project did reveal some limitations that are important for future learning. While the system worked well, the researchers noted that relying on the internet for the AI vision creates a risk. If the internet connection fails, the robots cannot "see" the blocks. They also found that the time it took for the AI to process an image was about one point eight seconds, which is fast for a classroom but might be too slow for a high-speed factory. These issues were not seen as failures, but as valuable lessons. They provide a concrete example for students to discuss how to make systems more reliable, perhaps by adding backup systems or processing data locally on the robot instead of in the cloud.

This work demonstrates that it is possible to build a sophisticated, cooperative robotics system that is affordable enough for a university classroom. By combining a self-driving vehicle, a robotic arm, cloud-based intelligence, and a digital tracking system, the researchers created a complete learning tool. The system achieved a sorting reliability of ninety-eight percent and a vision accuracy of ninety-four point two percent, proving that low-cost components can perform complex industrial tasks. More importantly, it showed that when students build these systems themselves, they gain a much deeper understanding of how the future of manufacturing will actually work. The project offers a blueprint for schools to teach the next generation of engineers how to manage the smart, connected factories of tomorrow.