Extrusion machines are quite big – the material traverses several hundred meters from one end to the other end of the extrusion line. During this traversal the material continuously changes its consistency and shape. At the end of the line the material is checked whether or not it matches certain quality criteria on several properties. If not, a material is classified as scrap and needs to be discarded or re-worked. In this project we aim at predicting a scrap event occuring at the end of the pipeline already at the very beginning of the pipeline. A successful prediction will be able to reduce the scrap rate for the machines as machine operators can intervene based on this information. After several integral transformations (particularly to map the condition of the material at the beginning of the pipeline to the exact same part of the material at the end of the pipeline), we build machine learning models to accurately predict scrap events on three material properties: thickness, weight and width. A model is trained for each of these properties. The model is then exported going to be pushed on to an edge device in the plant. It communicates with a GUI running on the IPC located next to the extruder line to deliver live predictions. To generate the live predictions we query live data (not available through the MDL) via the OPC-DA interface at the start of the extrusion pipeline and predict whether or not the material will result in a scrap event if left uncorrected.
Dubravko Dolic, Head of Applied Analytics & AI | Continental Tires
Studied Sociology with focus on statistical methods Dubravko Dolic made his first IT/programming experience already during his time at universities in Oldenburg and Northern Ireland. His professional career was shaped by a huge number of projects located between IT and business. Always striving to make data analyses easy available he was consultant for BI and Data analysis for more than 15 years. Since 2017 he is responsible for Data Science at the central IT for Continental Tires.
David Koll, Senior Data Scientist | Continental Tires
David is a Senior Data Scientist actively shaping the digital transformation at Continental with AI. While experienced in supply chain topics he is currently focused on the implementation of AI projects in the area of Industry 4.0. David is a Computer Scientist by education, obtaining his PhD from the University of Göttingen in Computer Science in 2015 after research visits to the Universities of Oregon (USA), Uppsala (Sweden) and Fudan (China).