Issue link: https://iconnect007.uberflip.com/i/597087
26 The PCB Magazine • November 2015 Companies tend to be extremely protective when it comes to sharing their dirty laundry with customers, with the "unenlightened" be- ing the most resistant. Statistically-challenged companies have argued everything from a flat- out "We don't believe in SPC," to "Our custom- er return rate is very low, so why do we need SPC?" The answer is really quite simple: In ad- dition to the performance-related issues that reach a customer, like late delivery and defec- tive product, the cost of inefficiencies in any process are directly passed along to the custom- er in some form. Scrap (or overrun to cover an- ticipated scrap), rework and inefficient process- ing are factored into the pricing model, so the lower the efficiency, the higher the total cost to the customer. And as we all know, "It's always about the dollars." Early in my career, a wise old mentor told me, "Steve, never argue about what can be mea- sured." Being an engineer by trade and Ger- man by lineage, he knew a little about precision craftsmanship. This advice has stuck with me, and in the quest for continuous improvement has translated into "How can we get better if we don't know where we are now?" followed by "How can we know where we are now without metrics?" An effective SPC program should be a key metric of continuous improvement and a valuable management information system. no Fear The great statistical myth is that SPC is an- other big scary concept that can only be suc- cessful in companies that have dedicated qual- ity engineers devoted to this complicated disci- pline. The biggest reason for the perpetuation of this myth is a lack of understanding of basic statistical principles, which results in the ten- dency to focus on the math and not the ben- efits. I will break down the basic principles into simple, everyday concepts that will allow you to see the SPC "big picture" and focus on the continuous process improvement aspect. The math is the easy part of SPC due to the abun- dance of inexpensive statistical software avail- able to do the number-crunching heavy lifting. Instead, I would challenge you to concentrate on the theory, benefits and practical application of statistical tools as a path to SPC without tears. From a fundamental standpoint, SPC is sim- ply "making sense of experience," and the skill- set needed is one that has already been honed over the course of your normal day-to-day ac- tivities. For example, this summer my wife and I were driving past a corn field when she ob- served, "There must have been a lot of rain here this summer because the corn is higher than normal for this time of year." What my wife had just done was make a subconscious statistical comparison of a current data set (this summer's corn crop) to a historical data set (past summer's corn crops), and formed an educated guess as to the cause for the difference (excessive rain). From a conceptual viewpoint, statistics are no more difficult than this. Identify Critical Processes to Avoid Useless Data A common mistake many companies make is to try to statistically monitor and control every single variable in every one of their pro- cesses. Show me a company that proudly boasts that they have 127 control charts in their SPC program and I will show you a company that just doesn't get it. Customers do not want to see wallpaper; that many processes cannot pos- sibly be controlled and managed properly, nor should they need to be. Use the technical re- sources at your suppliers to help determine the specific critical aspects of the process that FeATure DATA AnALyTICS THroUGH STATISTICAL TeCHnIQUeS " The answer is really quite simple: in addition to the performance-related issues that reach a customer, like late delivery and defective product, the cost of inefficiencies in any process are directly passed along to the customer in some form. "