Client Overview:

Objectives

1. Automate the Metal Sheet Inspection Process:
   Replace the manual inspection process with an automated, AI-driven system capable of inspecting the entire surface of metal sheets while they move at high speeds on the production line.
2. Detect Microscopic Defects:
   Implement a system that can detect defects as small as 50 microns, including scratches, patches, holes, lines, and other surface anomalies, ensuring high-quality products and reducing costly returns.
3. Customize the Inspection for Various Sheet Dimensions:
   Adapt the imaging system to accommodate different widths and lengths of metal sheets, ensuring that every type of sheet produced is fully inspected.
4. Establish a Defect-Specific Rejection Threshold:
   Design an algorithm that allows for different rejection thresholds based on the nature and size of the defect, minimizing the chances of false positives while ensuring critical defects are detected.
5. Improve Customer Satisfaction and Reduce Product Returns:
   Ensure that only defect-free sheets are shipped to customers, reducing the likelihood of product returns and improving overall customer satisfaction.

Challenges

1. Detecting Microscopic Defects:
   One of the main challenges was to inspect the entire surface of the metal sheets while detecting defects as small as 50 microns. These defects, such as fine scratches or microscopic holes, are not visible to the naked eye, making manual inspection ineffective.
2. High-Speed Production Line:
   The metal sheets moved at a high speed along the production line, making it difficult to perform thorough inspections in real time using traditional methods. The client needed a solution capable of keeping up with the speed of production without compromising on accuracy.
3. Varying Sheet Dimensions:
   The company produced sheets of various widths and lengths, which required a customized inspection system to cover the entire surface area regardless of sheet size. Standard imaging devices would not be suitable for sheets with larger dimensions.
4. Rejection Thresholds for Different Defect Types:
   Different types of defects, such as scratches, holes, and patches, required varying levels of tolerance based on their impact on product quality. Designing an algorithm that could dynamically adjust rejection thresholds for each defect type was a significant challenge.

Solution

To address these challenges, we developed a custom solution using a Contact Image Sensor (CIS) imaging system, tailored to the client’s production requirements.

1. Custom CIS Imaging System with 12 mm Working Distance:
   We implemented a long CIS imaging system, customized to match the maximum width of the metal sheets produced by the client. The system was designed to operate with a working distance of 12 mm between the sensor and the sheet, allowing it to capture high-resolution images of even the smallest defects (as small as 50 microns). The CIS system provided superior imaging quality, capable of covering the entire width of the sheet.
2. Real-Time Image Processing with AI-Powered Live Vision Software:
   The CIS imaging system was integrated with our proprietary Live Vision software, which processed the images in real-time as the sheets moved along the production line. Our AI model, combined with advanced image processing algorithms, identified various types of defects—including scratches, patches, holes, and lines—within milliseconds of image capture.
3. Customizable System for Different Sheet Dimensions:
   The CIS imaging device was fully customizable to adapt to different sheet sizes. It could be easily adjusted based on the width of the metal sheet being produced, ensuring complete surface coverage regardless of the product dimensions.

Implementation Process

1. Discovery and Requirements Analysis:
   We began by conducting an in-depth analysis of the client’s production process, the types of defects they needed to detect, and the various dimensions of the metal sheets they manufactured. This helped us identify the necessary specifications for the imaging system and defect detection algorithms.
2. Design and Development of CIS Imaging System:
   Based on the client’s requirements, we designed and developed a customized CIS imaging system with a 12 mm working distance. The system was engineered to operate at the high speeds of the production line while maintaining the ability to detect defects as small as 50 microns.
3. Integration with Live Vision Software:
   Our Live Vision software was integrated into the CIS system, allowing real-time image capture and processing. The AI-powered software analyzed the images and flagged any defects, with results displayed on the operator’s interface for immediate feedback.
4. Development of Defect Rejection Algorithm:
   We designed a defect rejection algorithm with adjustable thresholds based on defect size and type. This enabled the system to classify defects and ensure that only serious issues, such as holes or large scratches, triggered rejection.
5. Testing and Calibration:
   The system was tested extensively on the client’s production line, with calibration adjustments made to optimize defect detection accuracy. We ran multiple tests to ensure that the system could detect defects consistently, even at high speeds.
6. Training and Support:
   We provided training to the client’s operators and technicians, ensuring that they were equipped to monitor and maintain the system. We also offered ongoing technical support to address any issues or calibration needs as the system was integrated into regular production.

Results

1. Increased Inspection Accuracy:
   The CIS system, with its high-resolution sensors and AI-powered processing, was able to detect defects as small as 50 microns, ensuring that even the most minor imperfections were caught before the sheets moved to the lamination and coating stages.
2. Faster Inspection Process:
   The automated system significantly reduced the time needed to inspect metal sheets, allowing the client to maintain their high-speed production without compromising on quality. Real-time image processing enabled the system to inspect sheets as they moved on the production line, eliminating delays.
3. Reduced Product Returns and Customer Dissatisfaction:
   With the improved defect detection accuracy, the client was able to reduce the number of defective sheets that reached their customers. This resulted in fewer product returns and increased customer satisfaction, as the sheets met the high-quality standards expected by the client’s end-users.
4. Customizable for Different Sheet Dimensions:
   The system’s ability to adapt to various sheet sizes ensured complete surface coverage for every product, regardless of width or length. This customization was essential for the client’s diverse product line, providing flexibility and scalability for future production needs.
5. Lower Scrap Rate and Revenue Loss:
   By identifying and rejecting defective sheets earlier in the production process, the system helped the client significantly reduce their scrap rate. Defective sheets were caught before further processing, preventing revenue loss due to wasted materials and reducing overall production costs.
6. Improved Production Efficiency:
   The automated system streamlined the client’s inspection process, allowing for higher throughput while maintaining stringent quality control. Operators were able to focus on other critical tasks, and the inspection process became more reliable and consistent.

The implementation of the customized CIS imaging system with AI-driven Live Vision software revolutionized the client’s metal sheet inspection process. By automating the inspection, the client was able to detect microscopic defects in real-time, reduce product returns, and improve overall production efficiency. The system’s adaptability to various sheet dimensions and its ability to handle high-speed production lines ensured that the client could maintain their high standards of quality without slowing down their manufacturing process.

This solution not only addressed the immediate challenges of manual inspection but also provided a scalable and flexible inspection platform that will support the client’s future production needs. The result was a significant reduction in defects, improved customer satisfaction, and increased profitability for the client’s business.