AI-Machine Learning for Workplace Safety

Overview

This repository illustrates the outcomes of our project, where we combined the features of YOLOv8 with the advanced features of the ZED 2i Stereo Camera to enhance workplace safety. By automating the detection of safety hazards, tracking attendance, and monitoring restricted areas, our system improves compliance and helps prevent accidents in real time.

Table of Contents

• Overview
• Introduction
• Features
  • 1. Personal Protective Equipment (PPE) Alliance
  • 2. Attendance Checking
  • 3. Collision Detection
  • 4. Monitoring Restricted Areas
• Project Showcases
• Contributing

Introduction

This repository demonstrates the outcomes (illustration) of the project, where we combined features of YOLOv8 and the ZED Stereo Camera to create an intelligent system for workplace safety. Our solution:

• Detects missing personal protective equipment (PPE).
• Monitors attendance real-time video analysis.
• Detects potential collision risks.
• Prevents unauthorized access to restricted areas.

Features

1. Personal Protective Equipment (PPE) Alliance

• Description:
  The AI system detects when essential safety gear is missing and sends an instant alert to supervisors.
• RESULT:

---

Oriented 2D Bounding Box-Single Object

Oriented 2D Bounding Box-Multi Object 1

Oriented 2D Bounding Box-Multi Object 2

---

Detect Gloves

---

2. Attendance Checking

• Description:
  This feature uses video analysis to track and record worker attendance in real time, providing visual insights via 2D-3D bounding boxes visualization, path tracking, and skeletal mapping.

• RESULT:

  ---

3D Single Object Detection

3D Multi Object Detection

---

2D-Path Tracking

2D Bounding Box Skeletal Mapping

---

Glove Tracking 2D Bounding Boxes

Glove Tracking 3D Bounding Boxes

---

3. Collision Detection

• Description:
  The system uses AI-powered cameras to continuously monitor potential collision risks and sends instant alerts to prevent accidents.
• RESULT:

Collision 2D Bounding Boxes Detection

---

4. Monitoring Restricted Areas

• Description:
  The system detects unauthorized entries into restricted areas and sends immediate alerts to ensure safety and security.
• RESULT:

Restricted Areas Monitoring

---

Project Showcases

My team and I have showcased our project at two competitions. Please refer to the attached slides and posters for further insight into our work.

• Slides:
  • ASU and Dow Vietnam STEM Program: Engineering Projects (eProjects): SLIDES_
  • HCMUT-Bach Khoa Innovation:
    • Round 1: SLIDES ROUND_1
    • Round 2: SLIDES ROUND_2
• Poster:
  • ASU and Dow Vietnam STEM Program: Engineering Projects (eProjects): POSTER

Contributing

Contributions are welcome! If you would like to improve this project:

• Fork the repository.
• Create a new branch for your feature or bug fix.
• Submit a pull request with a detailed explanation of your changes.

Author