DeB-AT – Detection and separation of portable batteries from mixed waste using sensor technology and artificial intelligence

Starting point / motivation

Lithium-ion batteries and other battery types have become an integral part of our everyday lives, and the quantities in various products will continue to increase in the future. Energy storage devices, which are incorrectly collected in waste management and then sent to treatment plants, are causing more and more fires. These cause negative impacts on various areas of sustainability: ecological (fire emissions, further environmental damage), economic (high investment requirement for new construction, loss of insurance eligibility) and social (safety of the workforce, burden on the neighboring environment) issues are associated with them.

Contents and goals

The DeB-AT research project aims to detect and reject waste batteries in mixed waste material streams to be treated using sensor and AI-based technologies. The innovative project consists of the following work packages (1) sampling and classification, (2) object detection and segmentation, (3) AI image analysis, and (4) ejection of batteries. The innovation content is reflected in the novelty of the technological combination of sensor technology, image recognition and AI in application to the contaminent of batteries as well as the use of gravity acceleration in existing plant concepts and addresses a circular economy challenge for which there is currently no marketable solution applicable in existing plants.

Methodological approach

By combining different image processing and sensory technologies as well as artificial intelligence, the previously difficult-to-identify contaminant batteries becomes detectable in different material streams.

Expected results

The result of the DeB-AT project aims at a detection rate of more than 95% for both portable batteries and the group of lithium-ion batteries contained therein. As a foundation stone for a marketable application, a first laboratory & pilot plant demonstrator (TRL 4) is to be created in this research project. On the one hand, this will enable a new technology to improve plant safety in waste management and, on the other hand, potentials in resource efficiency through the newly separated material flow.

Project management

• Thomas Nigl

Institute/Company

• Montanuniversität Leoben – Chair of Waste Processing Technology and Waste Management

Partners of the project consortium

• Müllex-Umwelt-Säuberungs-GmbH
• SAMsoric GmbH

Thomas Nigl
Montanuniversität Leoben – Chair of Waste Processing Technology and Waste Management
Franz-Josef-Straße 18
8700 Leoben

thomas.nigl@unileoben.ac.at
+43 3842 402 5124
www.avaw-unileoben.at