Tracking Process Centerlines, Defects and Downtime – how are they related? One constant on the production floor is change. The objective of creating and maintaining a good process centerline application is to monitor performance of a production line to minimize variation and avoid outages like Defects and Downtime. The key requirement of a centerline approach is that it must be easy to use and be available to the right people – front line operators and technicians, who can find and fix issues quickly once they are identified. Another requirement would be to accept both manually entered data as well as through automation sources, allowing for the analysis of the data to benefit from both sources. A third item would be to adapt quickly and be easy to change. One constant on the production floor is change. With change comes issues like defects and machine downtime…items you failed to predict. A thorough investigation into the cause of these hits to your efficiency usually leads to new centerline variables that must be tracked. It is a never ending source of new items for you to monitor. So if a centerline application is always expanding, how do you manage it? First, add variables to a centerline project with realistic control and specification limits where possible. If you don’t have limits, they can be calculated based on historical performance. Second, use default values (target or last value) to allow entry of values that have changed. Some companies have a trust issue with this one. Third, use alarms or rule violations to highlight problems. These can be for both out of specification and out of control situations. Remember good control limits can help identify issues prior to them becoming out of spec items. Fourth, review the history of variables you are monitoring regularly to understand historically when variation in these variables occurs. With this knowledge in hand you can adjustment sampling frequency from say once an hour to once a shift or day. This frees up time you can devote to newly discovered sources of variation. Finally, don’t fall into this trap. Typically you start monitoring a variable because it has been the source of variation or the cause of defects and downtime in the past. Be very careful in making the decision to stop monitoring a variable because you detect no variation over a period of time. Remember what Arnold Schwarzenegger’s famous line in the Terminator series “I’ll be back!” Back off on the sampling frequency, but keep an eye on it.