The DIGISORT project is therefore developing a measurement technology capable of recording the relevant properties of the shredded material on-line and in-line. This recording is not done as an integral parameter, but individually for each fragment which means an enormous depth of information. The data is obtained by combining image data with spectroscopic or hyperspectral information. In the application, the data provides precise information about when which particles, with which shape and composition are processed. The corresponding large data sets must be processed and structured in order to develop engineering correlations. This is done using machine learning tools.
In a first step, the process data is exemplified in a conventional process control strategy. In the future, the data quantity and quality will also enable adaptive dynamic model-based control. The control of the separation step is intended to increase selectivity and thus product quality and yield. The separation of electrode foils serves as a relevant example of application. Those foils are investigated with model aluminium and copper mixtures as well as real shredder products. The corresponding data sets are incorporated into the data management of the Recycling / Green Battery Cluster.
In the future, the combined measurement and control technology of DIGISORT can not only be applied to separation steps but along the entire recycling process chain for quality monitoring and data acquisition. Also, the application is not limited to recycling processes of Li-ion batteries but can be adapted to a variety of other applications.