Project Name: HSE Shieldify

Team Name: Mavericks

Problem Statement

Detection & Alerting an unsafe HSE situation

Project Flow:

"The functionality of 'HSE Shieldify' follows a streamlined flow, ensuring both safety compliance and ease of use:

1. Employee Verification:

• An employee logs into the kiosk system and positions themselves in front of a camera.
• The system verifies if the employee is wearing the required personal protective equipment (PPE).
• If the verification is successful, the kiosk confirms with a 'verified' tick, allowing the employee to proceed. If the employee is not wearing the required PPE, they can retry the process.

2. Alert System:

• After several unsuccessful attempts, the system automatically alerts the manager via email or SMS, notifying them of the missing PPE. Simultaneously, logs are generated for further review.
• The kiosk system sends frames of images to the ML model, which analyzes them and returns a JSON response indicating the presence or absence of PPE components (e.g., Helmet: true, Gloves: false).

3. Dashboard & Monitoring:

• The dashboard displays statistics and logs generated by the model's analysis, alongside employee information.
• The manager can customize PPE detection requirements based on the specific needs of different departments through the dashboard website.
• Additionally, the manager can integrate CCTV camera feeds into the system, allowing them to monitor live footage. They also have the option to test video frames against the model to ensure PPE compliance.

Video Demo

Demo_HSE_SHIELDIFY.mp4

https://drive.google.com/drive/folders/1hBLiQv6c3lCqJ-lIM4fiXk66RWFqD9zQ?usp=sharing

Tech Stack

• Frontend:

  • Powered by React
  • Integrated with Chakra UI

• Database:

  • Managed using MongoDB

• Backend:

  • Developed in Python to communicates with the ML model & the frontend website

• Model:

  • ML model consists of YOLOv8 along with ultralytics and trained on Roboflow website

Directory Structure

FCRIT_College_Group6/
├── server/               # Backend server code
|   └── app/
│       ├── controllers/      # API Controllers
│       ├── routes/           # API Routes
│       ├── models/           # Database Models
│       ├── helper/           # Helper Files
│       └── main.py           # Server & Model entry point 
├── kiosk_frontend/       # Kiosk application 
│   ├── src/              
│   ├── public/
│   └── package.json      
├── dashboard_frontend/   # Dashboard application 
│   ├── src/              
│   ├── public/          
│   └── package.json      
└── README.md             # Project documentation

How to Setup Locally

Setup ML Model & Server

💻 Install

#direct to server folder
cd server

# install dependencies
pip install -r requirements.txt

📸 Execute

uvicorn app.main:app

Setup Kiosk

💻 Install & Execute 📸

#direct to kiosk_frontend
cd kiosk_frontend

# install node_modules for kiosk_frontend
npm install

npm run start

Username: emp002
Password: emp002

Setup Dashboard

💻 Install & Execute 📸

#direct to dashboard_frontend
cd dashboard_frontend

# install node_modules for dashboard_frontend
npm install

npm run start

Email: [email protected]
Password: emp001

Setup App

cd to App

Start server

python main.py

Forward 3000 port using ngrok http 3000

cd to MyApp

Update ngrok link in Dashboard Screen

npx expo start

Team Members

Name	Social Media Link
Savio Dias	Linktree
Jayesh Chaudhari	Linkedin
Tejashree Bhangale	Linkedin
Darren Dsouza	Linkedin