That Day the CPK Report Came Out, and the Room Went Silent for Three Seconds
I recall several years ago, a machine on our production line had perfectly fine test data the day before. Yet, the next morning, the day shift engineer, upon checking the daily report, discovered that the CPK of a critical parameter had dropped to 1.08. Honestly, this number was as good as a failing grade. The crucial point was that the machine started having issues at 2 AM but wasn't discovered until past 8 AM. During those six hours, the machine kept running, and we silently produced a batch of products that were likely destined for scrap. In the meeting room, everyone exchanged glances, unsure of what to say.
Where Did the Problem Lie?
This is one of the most common issues we encounter in a semiconductor factory: an abnormality occurs, but no one knows about it! Or rather, the person who knows is not the one who should address it immediately. By the time the data and reports come out, it's too late. At this point, the concept of Visual Management becomes extremely important. Simply put, it's about making abnormalities 'visually obvious'. Imagine if the CPK of that machine dropped below 1.33, a nearby light would change from green to yellow; if it dropped below 1.10, it would immediately turn red, or even trigger an alarm. Then, the day shift engineer, upon entering, wouldn't even need to check a report to know where the problem was.
Therefore, the key is that Visual Management presents all the information you need to know in the most intuitive and fastest way, allowing you to instantly assess the situation as good or bad, and whether intervention is required.
How to Implement It?
The simplest and most common method is using a dashboard. You can place a small screen next to the machine, or even a laminated chart, displaying the real-time status of critical parameters.
- Set Clear Standards: Using the CPK example just discussed, we can define:
* Yellow Light: 1.33 < CPK ≤ 1.67 (Caution, potential intervention needed)
* Red Light: CPK ≤ 1.33 (Abnormal, immediate action required)
You can even include DPMO, for example: DPMO > 6210 triggers a yellow light.
- Visual Representation: Don't rely solely on numbers; they are difficult to interpret at a glance. Use colors, icons, and trend charts. If a parameter's value suddenly spikes or drops, use arrows or color changes to provide alerts.
- Place in Conspicuous Locations: Information isn't just for show; it's to ensure that all relevant personnel, from operators to production line managers, can see it immediately. Posting it directly next to the machine or at the entrance of the production line is far more effective than reports stored on a computer and updated only hourly.
In other words, transform the data from your computer into 'traffic lights' on the production floor.
Common Pitfalls
We've fallen into quite a few traps before. The most common one is 'information overload'. Everyone thought Visual Management was great, so they put every conceivable piece of data onto the display, resulting in a cluttered screen that looked like a stock trading interface. Operators couldn't understand it at all, or simply bothered not to look.
Frankly, Visual Management isn't about displaying everything; it's about 'showing only what's most important.' You must filter out key indicators that truly represent the health of the production line. Too much information can dilute the effectiveness of alerts. Another common pitfall is 'doing it only once.' Someone puts up a chart and thinks everything is sorted, but the data isn't updated, the standards aren't adjusted, and eventually, it just becomes wallpaper. Visual Management is dynamic; it requires regular review to maintain the timeliness and effectiveness of information.
One Thing You Can Do Today
Identify the parameter on your production line that most frequently causes issues, and sketch out its 'traffic light' standards on paper.