In the "AI-Waste" research project, the recycling share is to be increased by at least 10 % through innovative approaches. Methods of image recognition and machine data analysis are combined to optimize the overall waste treatment process. The project, led by the Know-Center, is being implemented in cooperation with the Institute for Machine Vision and Display at Graz University of Technology, JOANNEUM Research Forschungsgesellschaft and Komptech GmbH.
The mountains of waste or plastic are continuously growing worldwide. For waste processing, the different composition of waste - consisting mainly of plastics and composites as well as organic fractions such as paper and cardboard - is challenging because it varies greatly seasonally and regionally. Existing waste processing plants do not have a widely used or suitable technology to automatically detect the quality of intermediate steps within a plant. As a result, it can happen, for example, that the proportion of plastic bottles is well separated, while the remaining waste components such as cardboard packaging are inadequately separated.
"Our goal is to describe the type and composition of the waste in the running process, which represents a technological milestone. To achieve this, we are combining image data with machine data for the first time. The data is collected under realistic and application-oriented conditions," says Dr. Robert Ginthör from the Know-Center and explains the procedure as follows: "In order to capture the different properties of the waste in the best possible way, we use 2D and 3D methods in image processing. The image analysis software is trained using deep learning algorithms to recognize and distinguish the waste." Finally, researchers derive models from image and time-series data from the plant to optimize the plant.
Innovation for business and the environment
"Digitalization offers untapped potential, especially in our field of activity. Constant innovation is the only effective means of achieving long-term success as a company," emphasizes Dr. Christian Oberwinkler, CTO of Komptech GmbH, which is supporting the project as a technology partner in the mixed waste treatment sector.
KI-Waste is funded within the framework of the Styrian Future Fund and the Graz Climate Fund. Economic and Research Provincial Councilor, MMag.a Barbara Eibinger-Miedl: "Climate and environmental protection have a high priority in Styria. Modern processes in an innovative circular economy are an essential building block for achieving Styrian goals in this area. The KI-Waste research project will make a valuable contribution in this regard and at the same time strengthen the international visibility of Styria and Austria as a technology location."
The result of KI-Waste will be a recommendation for action on how AI can be used in process optimization for the waste and circular economy. Waste management companies will benefit from an increase in efficiency, an increased recycling rate and reduced energy consumption, which will subsequently have a positive impact on the environment.
The project was launched at the beginning of 2021 and is scheduled to run for a period of two years. The results are also expected to provide preliminary work for other industrial sectors, such as the pharmaceutical or steel industries, where image data also need to be analyzed together with time series data. It will also help to optimize image recognition in general in terms of measurement accuracy and measurement position.