SPC for Process Industries

Kent Ehley wrote and narrates this SPC series. He says this videotape set is specifically designed for those who have already received basic training in the mechanics of creating and maintaining control charts for product control but lack the skills and understanding to effectively apply the charts to do process control. That may be true but this reviewer feels these videos can also be used for beginners with just a little help from a facilitator or coach.

This series focuses on the control of chemical solutions. However, the concepts apply to any process that exhibits natural trend variation, such as a machining operation where toolwear presents a natural source for trend variation.

Ehley quickly gets the viewer's respect. He is obviously profoundly qualified to teach this subject. His speaking style is quite relaxed, even laid back . However, his voice holds an underlying intensity. A number of graphics support the narrative, and he makes frequent use of a Quincunx machine. Plus he includes an interesting variation of Deming's &quotFunnel Experiment."

A 44-page leader's guide accompanies the video set.

Review by: Mel Silberberg, Silberberg Associates, 17 Front Street, PO Box 4551, Salem, MA 01970 Tel/fax: 508-740-9132/9397

Review of a new (January 1993 release) videotape training series aimed at providing an introductory understanding of Statistical Process Control (SPC) concepts in continuous (rather than &quotbatch") chemical processes where control of concentrations, temperatures or other process variables is essential to product quality

The release of this training package represents a welcome addition to video-based &quotquality" training material in still another &quotniche" area, i.e. the chemical process industries and those companies (e.g. printed circuitry) that incorporate such processes as plating, cleaning, painting or chemical treatment as subprocesses in their operations. (A secondary market: any process setting, such as certain machining operations with natural trend variation in tool wear and, to a more limited extent, metal-forming or - stamping operations, for example, for the prediction of equipment maintenance schedules.) The video material is well planned, providing a logical sequence in going from &quotbasic concepts" (segment #1) to an examination of various &quottypes of variation" (segment #2), to methods of &quotreducing process variation" (segment #3) and, finally, in segment #4, to a comparison of the various &quotcontrol chart application" options available, each with its advantages and disadvantages. Slightly over thirty minutes of material are provided for each segment.

At the outset, I wish to report this is a &quottalking head" video. In these days of highly polished training-through-role-play, plant visitations, etc., it is difficult to accept without challenge any serious training efforts of the &quottalking head" type. Yet this training package succeeds in its mission by utilizing the services of Kent Ehley (pronounce &quotEli"), who obviously (1) knows his subject, (2) wishes to share that knowledge without overwhelming the viewer, (3) makes excellent use of both multi-color video graphics (some of it &quotanimated"), and a &quotQuincunx" (the &quotclassical" tool normally used to teach variation in statistics classes and used here to introduce trends as well.) For those new to this device, the video makes an excellent case for its acquisition as a quality training tool!

Ehley's style is &quotlaid back": he will periodically back away from going too much in depth into a topic, by a disclaimer such as &quotbut a complete analysis of this topic is beyond the scope of this video" and will occasionally make a point (1) by comparing what he's just explained to some &quotuniversal" experience (as in his discussion of sampling, by asking, &quotHow often do you sample a zero reading on your car gas gauge before stopping to get gas?"), or (2) by the frequent use of quality heuristics (e.g. &quotvariation is the enemy of quality", "'quality' is uniformly exceeding customer expectations, to customer delight", &quotquality means doing it right the first time", etc.)

Meanwhile, using the photographic film development process as his primary example, he adeptly lays the foundation for &quotprocess control by prevention", rather than after-the-fact &quotproduct control by detection", explores causes of variation over which operators may or may not have control, discusses (and contrasts) tolerances and control limits, and introduces the viewer to the &quotaiming" process for process optimization. His section on variation introduces the use of charts for both process monitoring and adjustments, with marvelous sections on following trends as solutions are depleted and on the &quotoverzealous operator". (Here he introduces one of his many practical guidelines: &quotRemember...if variation is random, frequent adjustment will increase variation; if it is not random, then frequent adjustment will reduce variation!") Here, too, can be found his six-step approach for stabilizing a process, which he modestly says can be found in any text on SPC!

In his section on reducing variation, surprisingly, Ehley makes the point that run charts, rather than control charts, are the tool of choice and, using a &quotpickling" operation in a plating company as his example, contrasts operations with automatic continuous-replenishing provisions with those in which operators make periodic adjustments to the active ingredients. Also examined here: the reduction in variation that comes with frequent, rather than occasional replenishments.

The final section, beginning with a comparison of attribute vs variable data, is devoted to an examination of the advantages vs disadvantages of three kinds of control chart approaches: individual variables (X-bar) charting; variables plus moving range charting; and individuals plus range charting. Detailed here, though introduced earlier is a valuable segment on process capability studies and their value in establishing &quotaim" levels and control limits. Also detailed is the use of charting for exclusively maintenance functions and the need for periodic recalibration of measurement instruments. Throughout the tapes can be found the cautions and guidelines that distinguish useful training from training for training's sake. The tape ends with a justification of charting as an essential aid in process stabilization, the ultimate goal being the reduction in variation to a level undetectable to the customer. Ehley calls this &quotQuality by Process Control"!

Training is most useful when those trained can immediately apply the new materials at the worksite and the leader's guide, if well done, can be the path to that application. In this instance, the guide is somewhat of a &quotmixed bag", designed almost as to be used as a text on statistics, replete with experiments, that could be offered as a course on the topic even in the absence of the video, rather than as an aid. It does, however, provide sections that do reflect, and review, the content of the video material on a tape-by-tape basis and, in the introductory segment, a series of valid recommendations with regard to how the tapes should be used, even including procedures for multivoting and nominal group technique in the search, by trainees familiar with a process, for its most significant control measures! It is, therefore, an example of &quotoverkill" considering the recommendation that the tapes be used by people familiar with statistics, but extremely valuable in any event, whether used as text (for the course suggested in the company's advertising copy) or as a training aid in process industry settings.