Waste Sorting

Challenge

In waste management, it is crucial to sort out specific green bags containing bio waste from other types of waste in a waste loading area. This task is usually performed manually by workers who need to identify and separate the green bags containing bio waste from other bags. Failure to do so can lead to environmental issues, health hazards, and increased costs of waste management.

Solution

To overcome this challenge, we installed a 3D color camera above the waste loading area to capture images of the waste bags. We used AI to train a model that can accurately recognize the specific green bags that contain bio waste, while also differentiating them from other green bags that should not be sorted as bio waste. The AI-powered robot can then sort out the bio waste bags from the rest of the waste, streamlining the entire process and reducing the need for manual labor.

Dashboard

Collect statistics in real-time, providing important insights such as the number of picks, the type of waste, and other valuable data. The dashboard displays this information in real-time, allowing for easy monitoring and analysis of the waste sorting process.

Results

• Accurate bio-waste bag identification — the AI model distinguishes target green bags from visually similar non-bio bags, a differentiation manual workers frequently get wrong under time pressure
• Improved bio-waste diversion rates — correctly sorted bio-waste reaches composting instead of general waste, reducing downstream treatment costs and environmental impact for Halden Municipality
• Real-time sorting dashboard provides pick counts and waste-type data per shift for operational planning