Process Behavior Charts

Performance expectations are the Key Performance Indicators (KPIs) on which you ll focus your design and optimization efforts, and which you will track over time with your Design Scorecards (Technique 39) and/or Process Behavior Charts (Technique 52). 

While the purpose of Design Scorecards is to prevent problems, defects, and errors through superior design, they also enable better problem detection after a new solution (design) is implemented. If you are in detect-and-fix mode, any number of process-optimization techniques may help, such as Process Behavior Charts (Technique 52), Cause Effect Matrix (Technique 54), Mistake Proofing (Technique 49), and Design of Experiments (Technique 50). 

With an optimized innovation ready for the market, it s time to improve and transition the project to its owners for ongoing operation. Use the Process Behavior Charts and Control Plan techniques during and after this transition. Also use the Cause Effect Diagram and Cause Effect Matrix to diagnose, solve, or at least mitigate any implementation problems encountered. 

Use Process Behavior Charts to monitor the performance of your new innovation as it goes into production or commercialization after its design. By doing this, you create the visibility that is necessary to ensure your new innovation has successfully made the transition from the drawing board into the real world. 

We only cover the most basic chart types here, so it would be necessary to have help from a process expert or statistician if your needs are more sophisticated. As well, you may need or benefit from Process Behavior Chart software if not performing your own calculations. 

Process Behavior Charts are often called control charts, but this convention implies that a control function is performed on the contrary, Process Behavior Charts only perform a monitoring function. The control function is performed by virtue of a good Control Plan (Technique 55). 

The general sequence for constructing Process Behavior Charts is the same for all types of data, but there is some variation depending upon whether attribute data (data you can count) or variable data (data on a scale) are involved. We ll show you the steps and some details for each type of Process Behavior Chart. 

Many statistical programs (such as Minitab, SigmaXL, and IMP) automatically calculate control limits for the various types of Process Behavior Charts. If you re curious or want to perform your own calculations, see the resources listed at the end of this technique. 

Here you are looking for rule violations, which are Process Behavior Chart readings that indicate the process is out of control and needs help. 

One of the most common types of Process Behavior Charts for variable data is the Xbar/R chart, or average and range chart. While the overall procedure for constructing this chart is the same as for the C-chart, a few additional calculations are necessary due to the nature of variable data. 

The SkiBlades defect rate was mostly steady, with one exception on day 13. A team of process experts then discovered a correlation between fluctuations in the temperature of the resin curing oven and the defective pulls that broke away from the boots. At the team s recommendation, the temperature controller was replaced with a more modem unit, and the oven temperature was recorded and monitored using a Process Behavior Chart. 

Two additional Process Behavior Chart rules come into play when dealing with variable data. For these rules, the area between the process mean and the control limits is divided into thirds, that is, la, 2a, and 3a zones, as shown in Exhibit 52.3. (a stands for standard deviation, the variance or spread of a given data set). 

For more information and Process Behavior Chart calculations, see ... 

The National Institute of Standards and Technology (NIST) has a good online resource for Process Behavior Charts called the Engineering Statistics Handbook. The portion on Process Behavior Charts can be found at ... 

Either way, measurement data needs to be recorded consistently and accurately to help you compare data points over time. If your process is complicated, you ll want to employ Process Behavior Charts (Technique 52) to quickly and visually track when the process goes out of control. 

---