Open-Source Embedded Vision System for Industry 4.0 Applications

Abstract

Vision systems are an important part of modern manufacturing processes in the Industry 4.0 era, using images from cameras to inspect and sort workpieces. However, it was found that vision systems have high costs in terms of both hardware and software, making it difficult for small and medium-sized factories to access this technology. Therefore, this research presents a low-cost, open-source embedded vision system that can communicate with industrial controllers. It is a combination of a PIXY 2 camera module, a ready-made camera module that uses artificial intelligence to detect colors and images, with an open-source embedded Arduino. Arduino will receive data from the PIXY 2 camera module and convert it into a communication format according to the Modbus protocol. In the experiment, the developed open embedded vision system will be connected to a PLC controller and a touch screen display to test the efficiency of detecting and sorting workpieces with different colors under different distances from the camera module and lighting conditions. From the experimental results, it was found that the longest distance from the camera module to the workpiece that can detect and sort the color of the workpiece correctly, both when the light on the camera module is off and on, was 80 millimeters. As for the connection to a PLC controller and a touch screen display, it was found that data can be transmitted and received correctly every time. It can also detect and sort the color of the workpiece immediately without any delay time.