Machine Learning Part Classification System

By Cormac Farrelly

“Automation does not need to be our enemy. I think machines can make life easier for men, if men do not let the machines dominate them.” -John F. Kennedy

Project Overview

This project is part of the Technological University of the Shannon (TUS) Industrial Automation & Robotics Engineering course. The objective is to design and implement a machine learning system that integrates seamlessly with industrial automation to classify parts.

Technological Components

The system is built around key technological components:

• Cognex Insight 7801 Camera for capturing images of parts and utilizing FTP.
• Siemens S7-1200 1215 DC/DC/DC PLC for controlling the automation process - the brain of the system.
• Raspberry Pi 4 for processing images, running the machine learning algorithm, and running the FTP server.
• Servo Motor and Sensors for handling the movement of parts on the conveyor belt.

Challenges and Solutions

Throughout the project, several challenges were encountered, including:

• Developing a reliable machine learning algorithm for real-time classification.
• Integrating hardware components like the PLC and Raspberry Pi.
• Establishing seamless communication between the camera, Raspberry Pi, and PLC via FTP and Snap7.

Each challenge was addressed with a systematic approach, resulting in a robust and functional system.

Thesis

The thesis documents the entire journey of the project, from initial concept to final implementation. It provides an in-depth analysis of the challenges, solutions, and the overall impact of integrating machine learning into industrial automation.

Project Poster

Conclusion and Reflection

This project marks a significant milestone in my journey as an Industrial Automation & Robotics Engineering student. It highlights the potential of machine learning in automation and sets the stage for future innovations in the industry.