January 10, 2013
Is your Big Picture Cloudy?
We all love our computers, don’t we? We love how quickly they can perform complex calculations, how easily they can be used to learn and communicate, and how much time they can save us in our jobs.
But computers can also be, well, kind of finicky, can’t they? They interpret things in a very precise manner. And if you’re not careful, that can make your job much harder.
Let’s consider a simple example. Your company manufactures widgets at two sites, with two production lines at each site, and measures outside diameter as the key quality characteristic. It’s your job to monitor overall performance across both sites and report to senior management. However, here’s how your site quality managers have set up their data collection naming:
Site 1 (Arizona)
- Parts: Widget A, Widget B, Widget C
- Lines: Line 1, Line 2, Line 3
- Test: OD
Site 2 (Virginia)
- Parts: Widget_A, Widget_B, Widget_C
- Lines: VA_L1, VA_L2, VA_L3
- Test: Outside Diameter
Can you see how difficult it’s going to be to answer your senior management when they ask “how’s our overall quality?” As far as your computer is concerned, “Widget A” and “Widget_A” are completely different items. The same goes with “OD” and “Outside Diameter”. Can you imagine if you had twenty sites? It would quickly become a nightmare.
This is where spending some time developing (and then enforcing) standard naming conventions can help immensely from standpoints of analysis, system complexity, and system maintenance. Getting folks to agree on such things can sometimes involve some pain, since obviously everyone thinks their way is best. But we’ve seen many examples of the reward being well worth the effort.
Here are just a few examples of parameters that can benefit from standardization:
- Part names – Obviously, the same parts should have exactly the same name wherever they’re produced
- Line/machine names – In addition to a standard format, it can be useful to include a site designator in the line/machine name
- Test names – In addition to using identical test names, it is best if a naming convention is defined so as sites need to add new tests they are consistent with existing tests. One example of a test naming convention is Type_Test_Measurment Device_Units. A test for net weight in this format would look like: Wt_Net Weight_Scale_g.
- Employee names – Decide if employees will be entered as First Name Last Name or Last Name, First Name and stick to it.
All of the items listed above seem like common sense, but they can be the difference between a system that is difficult to navigate and a system that is intuitive. I have seen both types of systems firsthand, and the perceived usefulness can be judged by these small criteria. Establishing standards isn’t particularly difficult, but enforcing them can be tedious at times. Stick with your standards and the results will be worth it!