More safety through sensor-based waste sorting

According to a study by the BDE Federal Association of the German Waste, Water and Recycling Industry, there are more than 10,000 fires in waste sorting plants in Germany every year. Lithium-ion batteries and rechargeable batteries were the cause in around 80% of cases: installed in smartphones, electric toothbrushes or singing greeting cards, they often end up in plastic waste together with packaging. They can be damaged and catch fire, particularly during the recycling process in sorting plants. The damage is estimated at around 1 billion euros per year.
Insulate critical batteries at an early stage

The DangerSort project aims to reduce the risk of fire in landfill sites: "We are developing a sensor-based sorting system that uses X-ray technology and artificial intelligence to detect risky lithium-ion batteries and rechargeable batteries and separate them from the rest of the waste stream at an early stage," explains Johannes Leisner, Head of the Sorting and Laboratory Systems Group at the X-ray Technology Development Center at Fraunhofer IIS. So far, there have been no preventative measures against battery-related fires, only downstream solutions such as improved extinguishing systems. In addition, sensor-based technology makes it easier to recycle batteries and rechargeable batteries and thus close their product cycle.

In the X-ray sorting system at Fraunhofer IIS, a high-speed conveyor belt traveling at up to 3 m/s transports the waste stream. Above the conveyor belt is an X-ray source that works like a baggage scanner at the airport and scans the material flow. Thanks to the X-rays, it also identifies rechargeable batteries and batteries that are installed or concealed by other waste. An X-ray detector mounted under the conveyor belt takes pictures at the speed of the conveyor belt. This produces a continuous series of X-ray images.

This sequence is then evaluated: "We use an AI system for this that is particularly fast in image processing and is normally used in autonomous driving," says Johannes Leisner. "We have adapted and retrained it so that it can also analyze X-ray images and specifically identify electrical appliances with lithium-ion batteries and rechargeable batteries."

Sorting is initiated on the basis of the collected data: Special compressed air valves separate the critical electrical appliances from the rest of the waste stream, with a series of approximately 5 mm air nozzles removing them from the belt and conveying them into a separate chamber. The right timing between the evaluation of the image and the activation of the nozzles is crucial here. "It's difficult to capture and isolate the different battery sizes during the separation process - everything from ten-kilo e-bike batteries to button batteries," says Leisner.

The sorting system is currently still in test operation at Fraunhofer IIS; the system is to be delivered to the waste disposal company Lobbe at the beginning of June and tested in practice for the first time. The project, which is funded by the German Federal Ministry of Education and Research, will run until the end of August 2025. The development of the prototype system is part of the Plastic Packaging AI Hub: A total of 51 partners from industry, science and society are working closely together in the KIOptiPack and K3I-Cycling laboratories. The aim is to advance the application of AI methods for a resource-efficient circular economy in the field of plastic packaging in Germany.