The TRACES project supports the Pohlsche Heide composting plant in the detection of plastic and other materials
For organic waste to become good compost, it must be free of foreign materials such as plastic, glass, or metal. However, many piles of organic waste do contain such materials. This poses a major problem for composting facilities. A new regulation in the organic waste ordinance, in effect since May, stipulates stricter limits: if more than three percent of the waste contains foreign materials, the facility is permitted to reject it.
A new project is being tested at the Pohlsche Heide landfill in Hille. Called TRACES, it was developed by a team of scientists from Bielefeld University of Applied Sciences (HSBI). The idea: Artificial intelligence (AI) uses images of organic waste to identify foreign materials and their quantities.
Two master's students, Emilia Bensch and Marius Sangel, have already analyzed hundreds of images for this purpose. They mark plastic bags, paper, or other contaminants in the photos. This allows the AI to learn to recognize these materials on its own. The goal is to use around 5,000 images to train the AI by 2026.
The employees of the composting plant are also helping out. In workshops, they evaluate the images and share their experiences. This helps to further improve the AI. The application is intended to serve as a support tool – like a second pair of eyes. It will help those involved decide whether a pile of waste contains too many foreign materials.
In the future, AI could even help with waste collection. This would prevent many foreign materials from ending up in the composting plant in the first place. That would be a major step forward for the environment and the circular economy.
Source: www.its-owl.de
