Electronics manufacturers can maintain quality and increase time to market by using machine vision inspection methods in their factories. These systems, for example, can spot non-conforming components early in the manufacturing process. Designing with the supply chain in mind is becoming more common in the manufacturing space.
“This fundamental function improves quality by helping manufacturers deliver high-quality end products and can also identify problems early in the process to help avoid adding value to an out-of-tolerance sub-component,” said David L. Dechow, principal vision systems architect at Integro Technologies.
Using machine vision to improve electronic inspection is a viable way for manufacturers to confirm weak points. They can then make tweaks in their production processes to target ongoing issues and eliminate them.
“Expanding on these capabilities, machine vision becomes an integral part of Industry 4.0 relative to smart manufacturing and big data,” Dechow added. “Machine vision systems not only capture and filter the final products for quality but also deliver data that can be directly used in tuning a process to help make sure that less nonconforming product is produced in the first place.”
When electronics manufacturers use machine vision inspection methods, they often do so to verify that the products they produce meet specific standards. That was the case at Foxconn, a global electronics manufacturing services provider (EMS).
The traditional approach to machine vision inspection in electronics and other industries is to teach algorithms with a so-called “golden sample,” which represents a perfect product. Systems then compare all other products against that ideal to find defects.
However, Foxconn developed a different approach that does not require a golden sample. Instead, it learns by analyzing and clustering data without human assistance. When Foxconn deployed the technology on an assembly line for handheld electronics, it could detect 13 types of defects. Moreover, this approach caused a 50 percent reduction in the labor force needed to do those checks.
Better results from machine-vision algorithms
Doing machine vision inspection well typically requires combining cameras with adequate illumination. Keeping the products well lit is essential for detecting minute details. An examination of the type and color of light used to check printed circuit boards showed that a blue diffuse dome light board was better at illuminating surface traces compared to a red one of the same type.
The results a company gets with its machine vision inspection methods also depend on how it trains those algorithms. Google recently released a product to make that stage easier. It uses machine learning and claims that its system can create accurate models from up to 300 times fewer human-labeled images than general-purpose machine learning algorithms. The resultant accuracy is also up to 10 times better than what those general-purpose algorithms offer.
The faster training capabilities mean businesses can deploy their systems in weeks rather than months. That efficiency is a welcome benefit when using machine vision inspections on high-demand products. Semiconductors are excellent examples, where machine vision is used for alignment, inspection, measurement, process control and quality control. Chips are spurring the ongoing growth of the Internet of Things (IoT). In 2020, the industrial IoT market had a value of $77.3 billion, with some analysts predicting it could surpass $110 billion by 2025.
A different report focused on semiconductor-inspection methods revealed significant market growth, too. The forecast examined trends from 2017 to 2023. Analysts believe the market could achieve a 14.3 percent compound annual growth rate during the period.
Applying machine vision inspections in a busy factory setting works particularly well when people can get reliable results quickly. A company called Instrumental has developed a solution for businesses that want to use machine vision as efficiently as possible.
Some of the company’s imaging stations feature several cameras inside, making them excellent for capturing the critical details of electronic microassemblies. The process begins when a team member places a product inside an imaging station. They only need to wait four or five seconds for a red or green light to activate. It confirms whether the item passes inspection.
This system uses artificial intelligence (AI) to study the product at various production stages. Decision-makers can also learn more about potential pitfalls by examining the images captured during each phase.
Instrumental also uses a cloud-based system that lets relevant parties view the image data from anywhere. This feature supports remote workforces and overall improved flexibility.
“Instrumental focuses on allowing users to get more out of their data, which allows them to be agile and manage factories remotely,” said Tobias Harrison-Noonan, director of business development and solutions. This capability has never been more important than now, given the big shift that occurred during the Covid-19 pandemic, which prevented engineers from getting into their factories.”
Machine vision inspection platforms are not a guaranteed win for today’s electronic manufacturers. However, they could play a significant role in their success in the near and distant future.
Emily Newton is a technical writer and the Editor-in-Chief of Revolutionized. She enjoys researching and writing about how technology is changing the industrial sector.
Your email address will not be published. Required fields are marked *
Save my name, email, and website in this browser for the next time I comment.
This site uses Akismet to reduce spam. Learn how your comment data is processed.
HOME ABOUT MEDIA KIT EPS PURCHASING SURVEY CONTACT US SITE MAP PRIVACY POLICY TERMS OF SERVICE CALIFORNIA DO NOT SELL REGISTER LOGIN
Copyright © 2022 by AspenCore, Inc. All Rights Reserved.