IoT Food-Waste Monitor
Automated Food-Waste Monitoring Platform - IoT & Computer Vision
An IoT and computer-vision platform that automatically identifies and weighs food waste in commercial kitchens, turning it into analytics on a web dashboard. An over-the-bin device pairs a camera with a scale to log every wasted item with zero staff input, helping hospitality teams cut food waste, costs, and carbon footprint.
The platform combines edge devices running a computer-vision model with a Django API and a React analytics dashboard - covering everything from item recognition at the bin to actionable reporting in the browser.
Edge Device & Computer Vision.
An over-the-bin edge device automatically identifies and weighs each food-waste item via camera and scale, requiring no staff input. A computer-vision model trained in TensorFlow and Keras recognizes food types from camera images to classify what is being thrown away, with OpenCV handling image processing at the edge.
Real-Time Monitoring & Analytics
Weight and sensor data are streamed and stored in InfluxDB, visualised in real time via Grafana. A web dashboard with charts and reports turns waste data into actionable insights on cost reduction and carbon footprint, giving hospitality teams a clear picture of where waste is occurring and what it costs.
Device Fleet Management
A fleet of edge devices across mixed hardware - Intel NUC and Raspberry Pi 3 - is managed and updated over the air via BalenaOS, supporting multiple camera types and ensuring consistent software versions across all deployed units.
Services Provided
Web
- Python
- Django REST Framework
- PostgreSQL
- React
- MUI
- Recharts
- Mixpanel
- Azure
Device / Edge
- Python
- pyserial
- imageio
- BalenaOS
- InfluxDB
- Grafana
- Azure Storage
- Intel NUC
- Raspberry Pi 3
Computer Vision
- TensorFlow
- Keras
- OpenCV