I only realized the seriousness of the situation when OOC alarms blared non-stop that day
I remember it was a Tuesday afternoon. The OOC (Out of Control) alarms on the production line machines blared incessantly, and red lights flashed all over the station, more lively than Chinese New Year. I rushed over to check and found that the measured value of a critical parameter had skyrocketed. What was strange was that it was clearly within the control limits during the previous shift handover. This was bad; if this batch of products really had an issue, tracing back would mean scrapping an unknown number of wafers. A group of us gathered around the machine, our scalps tingling, thinking: "Damn! Which part of the process went wrong?"
What was the problem? Simply put, 'not following SOPs'
To be honest, many people find ISO 9001 rigid and bureaucratic. But when the OOC alarms kept blaring that day, I truly understood that it is essentially a 'kung fu manual' to prevent screw-ups. It's not there to restrict you, but to protect you, providing a basis and method for your work.
Think about it, why would OOC occur? Perhaps the equipment wasn't calibrated on time, a new operator went online without proper training, or we engineers changed parameters without recording them. These issues are actually closely related to the core requirements of ISO 9001 clauses. Simply put, it requires you to establish a 'Quality Management System' to ensure that what you produce consistently meets customer requirements from start to finish.
How to actually implement it? Stop relying on gut feeling!
ISO 9001:2015 contains many clauses, but for us engineers, there are a few key points you absolutely must grasp:
- Documented Information (7.5): Simply put, your SOPs, WIs, and specification documents must be clearly written, updated, and everyone must follow them. That day when OOC occurred, we later found that a certain parameter's set value had been secretly changed but not recorded in any document, causing the next shift to continue using the wrong setting. You see, poor document control is the beginning of screw-ups.
- Resource Management (7.1): This includes personnel, infrastructure, environment, and monitoring equipment. Personnel training must be in place. For example, new hires in our factory must pass certification before operating machines; if they don't score above 85% on the SOP exam, sorry, they can only observe from the side. Equipment also needs regular maintenance and calibration. For instance, our measuring instruments are sent out for calibration annually to ensure the Cpk value does not fall below 1.33. Otherwise, your measured data is simply inaccurate, and a low DPMO would just be self-deception.
- Operational Planning and Control (8.1): This clause is the most important. From order receipt, design, production, to shipment, each stage must be planned regarding how to do it and how to inspect it. For example, in our production process, each station has SPC (Statistical Process Control) charts to monitor parameters in real-time. If seven consecutive points fall on the same side of the average, even if still within control limits, we will intervene and investigate, as this indicates an abnormal trend in the process.
- Nonconformity and Corrective Action (10.2): What do you do when something goes wrong? You cannot pretend not to see it. You must record, analyze the cause, take corrective actions, and confirm that the improvements are effective. Regarding the OOC incident that day, we put all related wafers on hold, conducted a thorough trace and analysis to identify the root cause, and then revised the machine parameter change process, making the controls more rigorous.
The most common pitfall: Too many processes, nobody wants to use them
Speaking of ISO, the biggest pitfall I've ever encountered was writing processes that were 'too perfect' in the past, so perfect that no one wanted to use them. Every step required approval, every detail needed a form, and eventually, everyone bypassed the processes to save trouble. The result? When a screw-up occurred, it was impossible to trace which part of the process went wrong.
Later, I learned my lesson: processes must be streamlined, practical, and executable. The key is not to write down everything, but to identify the 'critical control points' and ensure these areas do not fail. It's like flying on a plane; the pilot won't tell you how every button is pressed, but he will ensure all necessary checks are completed before takeoff.
One Thing You Can Do Today
Go back and check the SOPs you have on hand: are they actually being used?