AI-Based Steel Sheet Surface Defect Inspection System

Inline and AI-Based Steel Sheet Surface Inspection Using Artificial Intelligence, Deep Learning, and High-Speed Industrial Machine Vision

On steel sheet production lines, high production speeds and elevated operating temperatures impose serious limitations on manual surface inspection. Under these conditions, small defects may remain undetected until the final coil is delivered to the customer.

This can lead to increased waste, customer quality claims, and reduced customer confidence.

Fanavar Fartak Espadana’s AI-Based steel sheet surface defect inspection system uses artificial intelligence, deep learning, image processing, and high-speed industrial machine vision to provide continuous surface inspection and automatic defect detection during production.

The Challenge

Manual inspection of steel sheet surfaces is not sufficiently reliable at high production speeds and elevated operating temperatures.

Small defects may remain undetected during the inspection process and be delivered to the customer with the final coil after production is completed.

This can result in:

• Increased waste

• Customer quality issues and claims

• Reduced customer confidence

Fanavar Fartak Espadana’s Solution

The system is installed after the finishing stage and before coiling on the production line, continuously inspecting the steel sheet surface.

It automatically detects, classifies, and locates surface defects and creates a digital quality record for each coil.

The system can also generate real-time alerts when recurring defects are detected. This capability helps operators identify production process problems before larger volumes of defective products are produced.

System Benefits

Continuous Quality Control at High Production Speeds

Inline steel sheet surface inspection is performed at production speeds of up to 25 meters per second.

Detection of Small Defects

The system can detect defects with approximate dimensions of 2 × 2 mm.

Quality Traceability for Each Coil

A digital quality record is created for each coil, enabling the product’s quality status to be recorded and traced.

Reduced Production of Defective Products

Detection of recurring defects and real-time alerts help operators identify production problems more quickly.

Technologies Used

  • Artificial Intelligence (AI) and Deep Learning
  • High-Speed Industrial Machine Vision
  • Image Processing
  • Specialized Optics and Illumination
  • Defect Detection and Classification
  • Coil-Based Traceability

System Capabilities

  •  Continuous Inline inspection of steel sheet surfaces
  •  Automatic detection of surface defects
  •  Defect classification
  •  Defect localization on the steel sheet surface
  •  Creation of a digital quality record for each coil
  •  Coil-based product quality traceability
  •  Monitoring of recurring defects
  •  Generation of real-time alerts to support the identification of production process problems
  •  Continuous 24/7 operation

Key Specifications

  • Inspection at production speeds of up to 25 meters per second
  • Detection of defects with approximate dimensions of 2 × 2 mm
  • Automatic defect detection, classification, and localization
  • Continuous inspection capability, 24 hours a day, 7 days a week
  • Creation of a dedicated digital quality record for each coil

System Innovation

The system combines high-speed AI-based intelligent inspection with product traceability and preventive maintenance support.

Recording and analyzing recurring defects can help identify production process problems in addition to controlling product quality, enabling operators to respond more quickly before larger volumes of defective products are produced.

Target Customers and Market

  • Steel sheet manufacturers
  • Steel strip manufacturers
  • Rolling lines and flat steel product manufacturing lines

This system has been commercialized and deployed in an industrial environment and has the potential for use in domestic and export markets.