The Challenge

  • Identifying defect parts in plastic profile manufacturing is a painstaking but necessary procedure for the client. Defects occur in various forms and due to various causes because of the complexity of the manufacturing process.

The Approach

  • A machine learning – based quality prediction system
  • Sensor-based process monitoring in pilot line
  • Pilot line with iOT gateways and cloud connection
  • Integration of crucial data from MES that was previously unused
  • Initial business case (2 weeks)
  • Prototype (2 months)
  • Implementation (6 months)
  • Subsequent roll-out to other production lines/sites

Result & Added Value

  • ML models to predict failure probabilities before they occur based on production and measurement data
  • Data-driven optimization allowing iterative improvement
  • Est. 1m € EBITDA gains from implementation
Matthias Welge
Head of Investor Support
We are here to help

Connect with us

Need support in developing your technology solutions? Let’s have a conversation about your projects and explore how we can create value together.