You’ve received a customer call on a Friday afternoon, “Our boards aren’t working. Didn’t your team test these? We have an urgent delivery and this quality is not assuring.” A lump forms in your throat. You know you have processes in place for quality control, but faulty boards are still making it through to shipping. Something in the process needs to change.
Quality issues and errors happen, especially with complex human-operated tasks involved with PCB and electronics manufacturing. Often, organizations may be equipped with systems like an automated optical inspection (AOI), which is typically used for inspection in the “middle” of a manufacturing process. Even though the AOI systems bring value to quality control, it is still leaving the front and back end of the process vulnerable to errors.
Quality errors not only cost potential thousands of dollars or more in claims from an unhappy customer, but they can also cause damage to your brand and reputation as a supplier. To avoid costly human errors, there are some helpful tips to apply to processes, protecting your brand and quality production from end-to-end.
Incoming quality control is the 1st line of defense in quality issues. With a PCB Assembly shop, a good place to start is by evaluating the quality of the PCBs received as well as ensuring the correct components are ready prior to assembly. With Electronics manufacturing, the first priority is ensuring you’ve received the correct parts before beginning to build the product.
Visual inspection is commonly used at this first phase, with checks for any upfront quality issues. This sounds simple in theory, however, operators manage multiple complex products, and differentiating “good” and “bad” components can be a challenge.
To address this, a good practice is to keep a reference “golden” image of a board or part and use that to ensure the correct component is in place before starting production. The latest technologies around AI can aid visual inspections by allowing users to train a system to automatically inspect received parts and boards to indicate whether the items correctly correlate with the product about to begin production.
Always inspect the first article. The first board or component being manufactured should be reviewed before the green light is given for a full production run. First Article Inspection (FAI) can be a long process in order to verify that a board is manufactured correctly against the specifications provided by the customer, however, a proper inspection at this intersection is key to avoiding costly quality errors on a full batch of production. The FAI should identify potential issues based on historical complications with the product as well as indicate any deviations by comparison with a previously manufactured sample.
The main risk for quality errors at this phase resides with human-operated visual inspection. As we know, when humans face fatigue, multitasking, and decision-making, we are prone to errors. To facilitate the FAI process, manufacturers can add advanced AI capabilities to support human decision-making, improving quality control by using AI to compare the stored “golden” reference images against the FAI, automatically isolating any discrepancies.
Although AOI systems are great for inline inspection, to help increase coverage of inspection for different types of defects, including those not detectable through AOI such as through-hole components, glue or seal defects, it is beneficial to incorporate a product sampling inspection strategy to pair with the AOI process.
Standalone visual inspection systems can help complement the AOI process as a secondary inspection, helping to identify defects such as missing or incorrect components or screws, as well as issues with reverse polarity and colour.
Sure the AOI system serves a valuable purpose, but what is the risk of error following the AOI system inspection, before final packaging and shipping? If processes include soldering, manual tasks, and on-site transportation, quality control is still vulnerable to damages, defects and line-down costs.
AOI systems can give a false sense of security. While it is possible for 90% quality control through AOI, there’s still a 10% chance that something else could go wrong — especially if there is some level of manual work involved. Humans will follow instructions as best they can, however varying skills from person to person and communication interpretations can leave room for errors in production processes. This can also be a concern with new operators requiring onboarding direction, time, and experience to inspect with the same speed and efficiency as colleagues.
To ensure best practices are in place for the final inspection phase, explore concerns like reoccurring problems, requirements for consistency, and implementing camera-based inspection and AI support when possible for optimal objectivity.
With the proper controls, checks and equipment, there is still a window for quality errors and a chance defective products will still reach the customer. Often this can happen when the product is damaged during shipping, or once arriving at the customer site. How can manufacturers improve traceability and search back to where the error may have occurred in the process?
Just as it is important to take an incoming inventory, it is important to keep a record of every product that has been shipped. Perhaps implement a serialization process whereby each product has a unique barcode and saved image of the outgoing batch.
The benefit? Next time a customer calls with a quality claim or issue, files are available indicating the state of the product before leaving the facility, giving valuable insight into quality process errors.
When evaluating areas for improvement, remember, automation and vision inspection systems at key points in manufacturing or assembly facilities can reduce risks with quality control and help to maintain brand integrity for leading suppliers. Ensure operator fatigue and related errors are avoided with AI-supported decision-making for manual inspection.
Copyright ©2022. All Rights Reserved BNP Media.
Design, CMS, Hosting & Web Development :: ePublishing