top of page

Non-Destructive Crack Detection for Enhanced Steel Sheet Quality Assurance using Computer Vision

Client Background

A steel sheet manufacturing company, with a reputation for quality products, faced challenges in maintaining rigorous quality standards due to limitations in existing physical crack detection methods. The client, committed to continuous improvement, sought a solution to enhance the crack detection process while minimizing the impact on production efficiency.


The client grappled with the limitations of traditional physical crack detection methods in steel sheet manufacturing. These methods were time-consuming, disruptive, and posed challenges in achieving comprehensive crack detection. The client aimed to enhance the quality control process by implementing a non-destructive method that would significantly improve accuracy without compromising production efficiency.

MicrosoftTeams-image (2).png

Reach out to us!

Let’s bring your ideas to life

Our Solution

In collaboration with the client, we implemented a cutting-edge solution leveraging computer vision technology to revolutionize crack detection in steel sheet manufacturing:

  • Existing Physical Method Assessment: We conducted a thorough evaluation of the client's current physical crack detection methods to identify limitations and areas for improvement. This involved scrutinizing the effectiveness of magnetic particle inspection, dye penetrant testing, and ultrasonic testing.

  • Computer Vision Integration: We introduced a non-destructive crack detection system based on computer vision technology. High-resolution cameras were strategically placed in the manufacturing process to capture detailed images of steel sheets, enabling real-time analysis for crack detection without disrupting the production line.

  • Machine Learning Algorithms: Advanced machine learning algorithms were employed to analyze the captured images. These algorithms were trained to distinguish between surface irregularities and actual cracks, ensuring accurate and reliable detection.

  • Real-Time Monitoring: The implemented solution provided a real-time monitoring system, allowing for immediate identification and classification of cracks. This enabled swift decision-making and intervention, reducing the likelihood of defective products progressing through the manufacturing process.

  • Integration with Production Workflow: To minimize disruption, our solution was seamlessly integrated into the existing production workflow. This ensured a smooth transition to the non-destructive crack detection system without compromising operational efficiency.



The implementation of our non-destructive crack detection solution using computer vision resulted in transformative outcomes for the client:

  • Enhanced Accuracy: Computer vision technology significantly improved the accuracy of crack detection, surpassing the limitations of traditional physical methods.

  • Operational Efficiency: The non-destructive approach integrated seamlessly into the production workflow, reducing downtime associated with quality control processes and optimizing overall operational efficiency.

  • Cost Savings: By eliminating the need for destructive testing and rework due to missed cracks, the client realized substantial cost savings, contributing to improved profitability.

  • Quality Assurance: The enhanced crack detection capabilities ensured that only defect-free steel sheets progressed through the manufacturing process, upholding the client's commitment to delivering high-quality products.

  • Future-Proofing: The use of computer vision and machine learning allowed for continuous improvement through ongoing training of algorithms, ensuring adaptability to evolving manufacturing requirements and crack detection standards.

bottom of page