Crack risk assessment for caster systems

Cracks Preventer is an artificial intelligence system built using metallurgical expertise. It is designed to predict any defects that occur during the casting process along with the breakouts, that can be caused by them. Based on the prediction, it also suggests counter measures reducing the losses caused by defects and breakouts in real time.

Cracks Preventer provides solutions to encounter defect types like longitudinal facial cracks, edge cracks, corner cracks, slivers and laminations and breakouts that might occur due to any of these defects. Cracks Preventer aims to reduce these defects, achieving up to 50% improvement defect inflicted losses.

Customer challenges addressed

  • Reduce number of LFC's in production
  • Reduce LFC induced breakouts
  • Minimize downgrade costs due to LFC's
  • Root cause analysis
  • Customized suggestions for the right countermeasure for each defect

Key features

  • Reduction in losses due to defects
  • Reduction of breakouts
  • Increased productivity and yield

Highlights

  • Domain knowhow meets machine learning

    Cracks Preventer makes use of the full SMS group skill porfolio

    Machine learning products without domain expert knowledge are not the solution the steel industry prefers. Cracks Preventer was developed with machine learning technology in conjunction with metallurgy experts. It ensures better performance accuracy that takes end-to-end data into consideration.

  • Full process chain

    Generating added value from data

    Cracks Preventer focuses on the entire process chain in steel production, allowing all mill owners to tackle LFC-related losses.

  • Performance-based solution

    Cracks Preventer is a performance-based solution

    Cracks Preventer is a performance-based solution. It eliminates risk factors for customers and reduces the costs of damages incurred.

Let's get in touch!

Use our contact form for questions, inquiries or personal contact.

Insights on your service

Show all
Let's get in touch!