MA

Industrial Watch

An AI-powered monitoring system designed to enhance workplace efficiency and ensure product quality in industrial environments.

React Native
Python
YOLOv8
MediaPipe
FaceNet
TensorFlow
Computer Vision
October 2023
10 Months

Project Gallery

Industrial Watch - Image 1

Features

  • Trained separate YOLOv8 models for bottles, centrifugal discs, and fabric defects with 90%+ detection accuracy.
  • Multi-threaded computer vision pipeline combining YOLOv8 cigarette/mobile detection with MediaPipe posture monitoring at 5 FPS.
  • Consecutive-frame validation to reduce false alerts.
  • Defect inspection with confidence-thresholded detection, batch yield calculation, and multi-angle inspection.
  • Face-recognition attendance tracking using FaceNet with a weighted productivity scoring system.

Interested in working together?

Get in Touch
MA

© 2026 Muhammad Anees