Discover how UNX Digital is transforming the manufacturing industry with a computer vision solution that detects product defects in real time. This innovative solution not only improves the efficiency and accuracy of quality control, but also protects the company’s reputation by ensuring that only the highest quality products reach the market.

Hardware/Platforms: Jetson Orin Nano 8GB + AWS Cloud Services
Application:Real-time ensuring product quality
Industry: Manufacturing
Deployment Location: Argentina

Background: Understanding the Opportunity

In the manufacturing industry, product quality is of utmost importance. A defective or poor-quality component can have a significant impact on the company’s reputation, customer satisfaction, and revenue. Traditionally, detecting defects in products has been a manual process that is time-consuming and subject to human errors.

The Challenge: Overcoming the Obstacles of Manual Inspection

The challenge was to detect poor quality material in product manufacturing, a manual process prone to errors and time-consuming, often resulting in false positives. If a defective component or part is detected at the end of the production line, there is a loss in labor, consumables, factory capacity, as well as revenue. On the other hand, if an undetected bad part reaches the final product, there will be an impact on the customer and a market reaction. This could potentially lead to irreparable damage to the organization’s reputation.

The Solution: Computer Vision for Real-Time Quality Inspection

To address this challenge, UNX Digital implemented a solution that automates the detection of defects in hardware products using deep learning. During manufacturing processes, our deep learning application is capable of detecting defects such as a crack or scratch in milliseconds with superior accuracy and performance. In addition, the solution interprets the defect area in the image with heat maps, providing a detailed and accurate view of the location and nature of the defect. This real-time detection allows for a quick response, reducing correction time and preventing the continuous production of defective products.

Business Impact: Improving Efficiency and Protecting Reputation

With the implementation of this solution, the company has been able to significantly improve the efficiency and accuracy of its quality control process. Previously, the detection of a defective product could take between 20 to 40 minutes, during which raw materials, energy, and resources were produced and wasted. With real-time detection, these times have been drastically reduced, allowing the company to reduce losses and improve customer satisfaction. In addition, the solution has enabled the company to protect its reputation by ensuring that only the highest quality products reach the market.