1. Before you begin configuring products in SPC, we recommend that you have a plan in place for data collection. The aspects you should consider and plan before you get to this point in your product configuration process are listed. Here are some items you should consider:
a. Identify which process (es) and characteristics must be monitored in order to determine control over the overall process. What you decide to monitor depends on what you want to accomplish (e.g. validation, revalidation, cost savings, failure investigation, process control, etc.). Some other considerations may be customer needs and current/potential problem areas.
b. Determine the frequency of subgroup sampling for each characteristic on which data is collected. A line producing 20 parts per hour would not, for example, be sampled as frequently as a line producing 20,000 parts per hour. It is recommended that sampling frequency be scheduled frequently enough to ensure that the potential to detect defects is high.
c. The next important point for consideration is subgroup size (how many events or measurements are to be used to analyze the performance of a process?). One guideline is that frequent sampling may use small subgroup size, whereas infrequent sampling may require large sample sizes. Attribute data types generally require more samples per subgroup size than do variable data types. In general, an initial study of a process could typically consist of 4 to 5 consecutively produced pieces representing a single process stream.
d. Determine the number of subgroups to collect. Collect enough subgroups to assure that the major sources of variation have had an opportunity to appear. For example, if you process is making five bad parts in every 1,000 and you are sampling 20 in every 1,000, it will take you sometime before you find any out-of-spec parts.
e. Calculate control limits. If you are collecting data for the first time on a characteristic, it is unlikely you will be able to calculate appropriate control limits. If that is the case, you can configure your control limits to be the same as the customer specifications. Then, as a general rule, you should collect between 25-30 subgroups so that you have enough statistical data to calculate appropriate control limits.
2. At this point, you have created a product and listed the processes to follow in order to manufacture the product, as well as where to run those processes (i.e. locations). The next step is to specify the characteristics you want operators to look at when collecting data. In Suite SPC, characteristics are always part of a group and each group is associated with a specific process. The group can contain a list of variable data you wish to monitor during data collection, or a similar list of attribute data. A group cannot contain both variable and attribute characteristics; however, it is possible to have more than one group associated with a process. In other words, you could have one group for variable data and one for attribute data. Note that characteristics are the same type, either variable or attribute, and are collected at the same time, frequency, and process should be grouped together. To add a new group, right click on the name of a process and select Add Characteristic Group. Complete the New Group dialog that appears. Now, you can enter specific “rules” that will apply to every characteristic you put in this group. Fill out the General tab shown and then click Apply and Save. To add a characteristic, right click on the group and select Add Characteristic. Type in the name of the new characteristic and click Ok. Then you will need to add a unit of measure. The decimal format controls how many places to the right of the decimal point can be entered at the point of data entry. Then, select one from the drop down list and go to the Limits tab. The limits tab controls how statistical data is formulated and how the data is plotted on charts. Fill out the Limits tab shown and click Apply and Save when completed. Now that the limits are configured, select which test you want applied to the data once it is collected. Go to the Statistics Tab. In this tab, you can select which set of tests to apply to the data as each set is accepted during data collection. The statistical tests control how statistics are compiled on characteristic data. These test also interpret the distribution pattern and indicate such things as instability, out of control, in control, etc. A statistical technique is a collection of all information related to the behavior of the control chart and calculation of statistics. They are used for displaying charts and calculating statistics. A statistical technique includes subgroup size, distribution type, chart type, control limits, and recalculation events and time frame. These statistical techniques are stored in a pool so that you can configure them once and use them in many different characteristic configurations. SPC provides a few common groups of statistical settings in a small group of default statistical techniques. Selection of a set of tests in the Technique Name field automatically completes the remaining fields in this tab. Select which tests to apply. Click Ok, Apply and Save. Go to the Statistical tests tab and it will list a summary of the tests chosen to apply to the data. Now onto the Data Acquisition tab and fill out the fields. Click Ok, Apply and Save. That’s the basics on how to configure a part for data collection in SPC.
Also see the video How Do I? - Part Configuration for more information.
Copyright 2019 by Waypoint Global LLC