Process capability studies determine whether a process is unstable, investigate any sources of instability, determine their causes, and take action to resolve such sources of instability. After all sources of instability have been resolved in a process, the natural behavior of the process is called its process capability. Process capability compares the output of a process (called “Voice of the Process”) with the customer’s specification limits for the outputs (called “Voice of the Customer”). A process must be stable (have an established process capability) before it can be improved. Consequently, a process capability study must be successfully completed before a process improvement study can have any chance for success.
Process improvement studies follow the Deming cycle of Plan, Do, Study, Act. First, managers construct a plan to decrease the difference between customer needs (Voice of the Customer) and process performance (Voice of the Process). Recall, that a plan is an intention to move from an existing method or flowchart to a revised method or flowchart by incorporating one or more change concepts. Second, they test the revised flowchart’s (Plan) viability using a planned experiment (Do). Third, they collect data and study the results of the planned experiment to determine if the plan (revised flowchart) will decrease the difference between customer needs and process performance (Study). Fourth, if the data collected about the revised flowchart show if the plan will achieve its objective(s).
Finally, the revised flowchart is standardized through “best practices” and training (Act); and the managers responsible for the plan return to the Plan phase of the Deming cycle to find further revisions to the flowchart that will further reduce the difference between customer needs and process performance. If the data collected about the plan show that the plan will not achieve its objective(s), the managers responsible for the plan return to the Plan phase of the Deming cycle to find a different revision to the flowchart that will reduce the difference between customer needs and process performance. Hence, the Deming cycle follows a never-ending path of process and quality improvement.
This chapter is divided into four sections: specifications, process capability studies, process improvement studies, and quality improvement stories. The quality improvement story is an effective format for quality management practitioners to present process capability and process improvement studies to management.
11.2 Specifications (Voice of the Customer) and Created Dimensions
Specifications fall into two broad categories: performance specifications and technical specifications.
11.2.1 Performance Specifications
Performance specifications address a customer’s needs or wants. An example of a performance specification can be seen in restaurants rated by the Red Michelin Guide. The customers of these restaurants set their performance specifications as “a perfect dining experience.” Perfection is measured in terms of the synergistic experience created by the interaction of food, service, ambience and price. The Red Michelin Guide rates restaurants on a one to three star scale. Only the best restaurants in the world receive Michelin stars. A restaurant receives one Michelin star for consistently serving very good food in a good setting, but it is not considered worthy of a special traveling effort.
A restaurant receives two Michelin stars for consistently serving excellent food, including specialties and wines of choice in a great setting. The restaurant is worth a detour from one’s existing travel itinerary. A restaurant receives three Michelin stars for serving excellent food and great wine, with impeccable and elegant service and ambience. The restaurant is one of the best restaurants in the world and is worth a special trip. All starred restaurants have a high average level of quality with very little variation around the average. A three star Michelin chef is an artist; it is as if Picasso was painting for your pleasure. Three star Michelin restaurants provide performance specifications. They guarantee satisfaction at the point of delivery. Nothing short of perfection is acceptable.
11.2.2 Technical Specifications
Technical specifications describe the desired values of quality characteristics at delivery. There are three types of technical specifications: individual unit specifications; acceptable quality level (AQL) specifications; and distribution specifications.
Individual Unit Specifications. Individual unit specifications state a boundary (upper or lower specification limit), or boundaries (both upper and lower specification limits), that apply to individual units of a product or service. An individual unit of product or service is considered to conform to a specification if it is on or inside the boundary or boundaries; this is the goal post view of quality. Individual unit specifications are made up of two parts, which together form a third part. The first part of an individual unit specification is the nominal value.
This is the desired value for process performance mandated by the customer’s needs. Ideally, if all quality characteristics were at nominal, products and services would perform as expected over their life cycle. The second part of an individual unit specification is a tolerance. A tolerance is an allowable departure from a nominal value established by design engineers that is deemed non-harmful to the functioning of the product or service over its life cycle. Tolerances are added and/or subtracted from nominal values. The third part of an individual unit specification is a specification limit, or the boundaries created by adding and/or subtracting tolerances from a nominal value. It is possible to have two-sided specification limits:
USL = Nominal + Tolerance
LSL = Nominal – Tolerance
where USL is the upper specification limit and LSL is the lower specification limit; or one-sided specification limits (i.e., either USL or LSL only). A nominal value and specification limits form the Voice of the Customer.
An example of an individual unit specification and its three parts can be seen in the specification for the “case hardness depth” of a camshaft. A camshaft is considered to be conforming with respect to case hardness depth if each individual unit is between 7.0 mm ± 3.5 mm (or LSL = 3.5 to USL = 10.5 mm). The nominal value in that specification is 7.0 mm; the two-sided tolerance is 3.5 mm; the lower specification limit is 3.5 mm (7.0 mm – 3.5 mm); and the upper specification limit is 10.5 mm (7.0 mm + 3.5 mm).
From our earlier discussion of the philosophy of continuous reduction of variation (i.e., the Taguchi Loss Function), we saw that the goal of modern management should not be 100 percent conformance to specifications (Zero Defects), but the never-ending reduction of process variation within specification limits so that all products/services are as close to nominal as possible, absent capital investment. Specified tolerances become increasingly irrelevant as process variation is reduced so that the process’s output is well within specification limits.
Acceptable Quality Level (AQL) Specifications. Acceptable quality level (AQL) specifications state a requirement that must be met by most individual units of product or service, but allow a certain proportion of the units to exceed the requirements. For example, cam shafts shall be acceptable if no more than 3 percent of the units exceed the specification limits of 3.5 and 10.5 mm. This type of specification limit is frequently referred to as an Acceptable Quality Level. AQL specifications are much like individual unit specifications, except they have a unique negative feature: they formally support the production of a certain percentage of defective product or service.
Distribution Specifications. Distribution specifications define an acceptable distribution for each product or service quality characteristic. In an analytic study, a distribution is defined in terms of its mean, standard deviation, and shape. However, from the Empirical Rule discussed in Chapter 5, it is not necessary to make any assumptions about the shape of the distribution. That is, virtually all data from a stable process will fall between the mean plus or minus three standard deviations.
As an example of a distribution specification, the case hardness depth of a camshaft shall be stable with an average depth of 7.0 mm and a standard deviation not to exceed 1.167 mm. In other words, individual units shall be distributed around the average with a dispersion not to exceed 3.50 mm on either side of the average since for a stable process, virtually all of the output will be within three standard deviations on either side of the mean [7.0 mm ± 3(1.167 mm) = 7.0 mm ± 3.50 mm = 3.50 to 10.50 mm]. The mean and standard deviation are simply directional goals for management when using distribution specifications. Management must use statistical methods to move the process average toward the nominal value of 7.0 mm and to decrease the process standard deviation as far below 1.167 mm as possible. Distribution requirements are stated in the language of the process and promote the never-ending improvement of a process.
Distinguishing between Performance Specifications and Technical Specifications. Performance specifications are not commonly used in business; instead, technical specifications are used. Unfortunately, this can cause major problems because technical specifications may not produce the performance desired by a customer. As an example, consider a hospital that serves medium (versus rare or well-done) steak to patients who select steak for dinner [see Camp, 1986]. The performance desired is patient satisfaction within nutritional guidelines. But performance specifications are not used. Instead, a technical specification of five ounces of steak is substituted; it is assumed they are equivalent.
University/College: University of Arkansas System
Type of paper: Thesis/Dissertation Chapter
Date: 10 January 2017
We will write a custom essay sample on Process Capability
for only $16.38 $12.9/page