SafetyNet-AI

• 

  Landing Page

• 

  Home Page

• 

  Dashboard

• 

  Live Monitor

• 

  CCTv footage

• 

  Worker Profile

Inspiration

Workplace accidents cost billions and ruin lives. Human supervisors cannot watch every camera, 24/7, everywhere. We built SafetyNet AI to be an eye that never blinks.

What it does

It's Context-Aware: Unlike old cameras that just detect "movement," this understands jobs. It knows a welder needs different gear than a driver. No Special Hardware: It works on standard webcams and laptops. Interactive: You can actually talk to it ("Hey AI, what risks do you see?"). Gamification: The "Safety Score" encourages workers to be safer to keep their score at 100.

How we built it

The Brain (Google Gemini 1.5 Flash Lite): Simple explanation: This is the "intelligence." We don't just use computer vision; we use a Large Multimodal Model. It understands context. It knows the difference between a "welder" (who needs a mask) and a "manager" (who needs a helmet). The Interface (React.js & Tailwind CSS): Simple explanation: A modern, responsive web dashboard that looks like enterprise software. It handles the 3D graphics and smooth animations. The Eyes (HTML5 Video & Canvas): Simple explanation: We capture live video frames from the webcam, convert them to images, and send them to the AI instantly. The Voice (Web Speech API): Simple explanation: We use the browser’s native ability to listen (Speech-to-Text) and speak (Text-to-Speech), allowing workers to talk to the system hands-free.

Built With

• 1.5
• canvas
• flash
• gemini
• google
• html5
• react
• tailwind