DiKueRec

In Germany, over 3 million refrigerating appliances reach the end of their technical lifetime every year and require recycling in dedicated electronic waste treatment facilities. Plastic, iron, copper and many other materials can be recovered in the process. The fluids contained in such appliances (e.g. chlorfluorcarbons CFC and hydrochlorfluorcarbons HCFC) have a considerable global warming and ozone depleting potential and need to be absorbed with great caution during recycling. Diversity of refrigerators and contained fluids make complex processing necessary. Economical adaptation of refrigerating appliances recycling is becoming increasingly difficult in the situation of steadily growing model variance. Novel technologies are urgently needed to improve the efficiency of refrigerating appliances recycling.

To guarantee that strict processing quality requirements are fulfilled in the future, recycling plants need to be upgraded with sensors and digital twins. An intelligent combination of data and algorithms will contribute to efficient control of recycling plants. The vision of the DiKueRec project is to guarantee safe and efficient treatment of current and future refrigerator models. This needs to be done in consideration of economical aspects to secure jobs in the recycling industry.

The research project is designed to run for 36 months. At the beginning of the research work, historical data on the composition of input material flows will be collected on-site. The status quo of recycling plants will be described using value stream mapping methods in order to discover potentials for upgrading the plants with sensors and intelligent control. Based on the collected data, digital twins of refrigerating appliances, processing machinery and material flows will be created.

In the following work packages project partners will implement the improved data collection and build up intelligent control as a demonstrative solution. The aim of the project is to structure the incoming data to create digital twins and to analyze the data using Machine Learning (ML). Predictions of ML models are to be utilized to control process parameters of the involved treatment processes. Alongside with the value-adding data collection and utilization, an intuitive graphical user interface (GUI) to display plant status and historical data has to be implemented. Solutions for data access for the stakeholders will be studied. Simultaneously, associated refrigerator manufacturers will be involved in analyzing how they can integrate durable identifiers for recycling in their products. Finally, solutions in the described work packages will be validated and optimized based on the feedback and know-how of recycling plant staff in consideration of economical and ecological aspects.

This research and development project (fund number: 02WDG016D) is funded by the German Federal Ministry of Education and Research (BMBF) within the Action Plan ”Natürlich.Digital.Nachhaltig” and managed by the Project Management Karlsruhe (PTKA).

Federal Ministry od Education and Research (BMBF)