Quality control is a methodology employed in manufacturing to prevent defects in manufactured products. Abbreviated as QC, the method has been implemented in a number of ways each of which has its own name and following. Quality control is typically associated with statistical approaches. Quality management has strong philosophical aspects based on the insight that quality is as much the result of management approaches as it is of specific activities. The modern quality movement is a fusion of American know-how originally developed at Bell Laboratories and Japanese enterprise and implementation. The several waves of quality control methods that have swept U.S. manufacturing since the 1950s are almost unthinkable except against the backdrop of Japanese industry achieving a world class reputation and thus producing stimulus. The movement is very closely associated with an American mathematician and physicist, W. Edwards Deming, although Deming was one of two prominent individuals who helped the Japanese forge their approaches to manufacturing; the other was Joseph M. Juran, a Rumanian immigrant to the United States. America's embrace of QC followed its successful application in Japan. Some type of statistical quality control is practiced in connection with most demanding manufacturing processes, but the more "qualitative" (no pun intended) aspects of QC have never been wholeheartedly embraced.
Modern quality control originated with Walter A. Shewhart, then working at Bell Telephone Laboratories. Shewhart devised a control chart named after him in 1923 and in 1931 published his method in Economic Control of Quality of Manufactured Product. Shewhart's method saw its introduction at Western Electric Company's Hawthorn plant in 1926. Joseph Juran was one of the people trained in the technique. In 1928 he wrote a pamphlet entitled Statistical Methods Applied to Manufacturing Problems which was later incorporated into the AT&T Statistical Quality Control Handbook which is still in print. In 1951 Juran published his very influential Quality Control Handbook.
W. Edwards Deming went to Japan to assist in the preparation of the 1951 Japanese Census. Being an expert on statistical methods, the Japanese Union of Scientists and Engineers (JUSE), having heard of Shewhart's techniques, invited Deming to lecture on statistical quality control. Deming gave a series of lectures in 1950 aimed both at describing SQC and at motivating his audience of executives. He pointed out the linkage between quality, productivity, and potential gains in market share. He found an enthusiastic audience. JUSE also invited Juran to lecture in 1954 with similar success, but by that time Deming had achieved wide prominence in Japan. With the great success enjoyed by SQC in Japan, and through his own abilities as a teacher and promoter of quality control and related management approaches, Deming became the iconic figure in the field, the "father of quality control." JUSE established the, by now, prestigious Deming Prize for quality-related achievements by individuals and organizations.
Japanese improvements in industrial performance eventually aroused interest in the United States in the early 1970s, led by Lockheed Corporation. Quality control then took on a life of its own in this country.
QUALITY CONTROL FUNDAMENTALS
Before the advent of statistical quality control, control was exercised by inspecting the output of manufacturing processes and removing defective items. The modern technique established an upstream method for detecting deviations from specified quality—early detection—used to trigger analysis of causes and then changes to manufacturing procedures.
SQC requires that the producer first identify several characteristics of a product to be measured, typically its dimensions, fit with other parts, smoothness, reflectivity, etc. Carefully conducted test runs are made first; every part is measured and its measurements are recorded. Upper and lower boundaries are set for every measurement from one or repeated test runs, with the idea that any part that falls within these boundaries conforms to the product's quality standard. The center line between the boundaries is then used as a base-line for measurement. Once this quality standard is set, production can begin.
The quality control activity during production consists of taking samples from the run continuously, taking measurements on the samples immediately, and then plotting them rapidly on a Shewhart Chart. During production, measurements typically fall close to the center line, some above it, some below it, some on the center line. A certain amount of divergence is natural and cannot be avoided. So long as the plotted points are within the accepted boundaries, the product conforms to the quality standard. But SQC demands that if the plotted points begin to show a trend away from the center, rather than clustering randomly around it—or, worse yet, begin to fall outside the boundaries in either direction—then production must stop. The incoming raw material, the production machinery, and other inputs, such as lubricants, must next be examined to discover why results are trending in the wrong direction or fall outside the acceptable range.
SQC thus provides early warning that quality is deteriorating. When the method is applied strictly, production cannot resume until problems are detected and fixed—as shown by brief test runs. Needless to say, money is saved by preventing wasteful production of parts later, products that fail to fit, or parts that result in product failure in use. In aircraft and autos, such failures can mean injury and death and massive lawsuits. Corrective actions taken early improve the process as a whole. In due time they lead to better equipment designs.
The technique also lends itself to the gradual ratcheting up of quality. This is accomplished by setting "acceptable boundaries" more narrowly and then modifying the production process until the new quality goal is met. This, of course, may require substantial changes to the process or the raw materials used. In Deming's conceptualization of the process, quality is thus "designed in" rather than "inspected out." The concept of "continuous improvement" arose in such efforts to raise quality. Its downstream consequences are lower cost in production and in warranty service, advantages in pricing, and higher customer satisfaction leading to brand loyalty and market share.
RELATED ISSUES, PRACTICES, AND MEASURES
Statistical quality control, as described above, is the fundamental description of quality control in the modern context. It is centered on measuring deviations from a norm and then taking actions to eliminate such deviations. But quality control, almost from the outset, came to be surrounded by what might be called a "cultural" radiation—namely management approaches, philosophies, and practices aimed at creating the right environment for a quality-driven industrial process. These radiations in part came from Japanese management culture, very different from U.S. practice, the ideas of Deming—which both influenced and reflected Japanese practice—and their elaborations by others.
Deming, for instance, in his 1982 book Out of the Crisis, formulated "14 points for management" which generally urge greater collaboration of effort, a longer-range view, commitment to improvement, constancy of purpose, and humane treatment and involvement of people. Some of Deming's points were revolutionary (e.g., to cultivate a single supplier for a resource, to eliminate management by objectives, numerical goals, and annual reviews of employees) while others have been adopted, albeit not always in the spirit in which Deming proposed them (e.g., team approaches, continuous improvement). Deming also condemned managing for the short term, management mobility (job hopping), reliance on technology rather than real solutions, and running operations by available numbers rather than a feel for the whole.
Quality Circles, covered in detail elsewhere in this volume, played a role in the 1980s. These circles were envisioned as collaborative efforts by teams of workers engaged in discovering problems and the solutions to them. The cultivation of team-approaches popular in the mid-2000s for every type of activity owes much to the precedent set by quality circles.
Just-in-time (JIT) procurement, which is much facilitated by the selection of a single, well-qualified supplier, arose from Japanese practice aimed at taking cost out of inventory control while maintaining very-high quality. Closely related to JIT is the practice of "supplier partnering" in which suppliers and their customers use the same quality standards.
Total Quality Management (TQM), also covered elsewhere in this book, was also a method promoted by Deming. TQM is a philosophy of management in which the operational elements are continuous quality improvement, quality circles, and strong management backing.
ISO 9000 is yet another result of the "quality movement." It represents a series of standards developed by the International Standards Organization which defined, for different industrial operations, managerial and operational standards that, if followed, will produce high quality. Companies can obtain ISO 9000 certification showing that they follow the standards ISO has laid out. This certification, then, can be used in advertising to inform the public that the company meets the ISO standard. The measure, of course, is somewhat indirect in that it guarantees certain practices, not necessarily what results from their use.
Lean Manufacturing is a practice pioneered by Toyota and widely imitated. It is called "lean" because it attempts to achieve results with minimum input of labor, space, cost, and time. The method relies heavily on JIT, cross-trained and highly motivated employees, and equipment arrangements that both save space and also cluster related tooling in close proximity. Layouts and arrangements are organized so that changes between production cycles can be accomplished swiftly. Lean manufacturing thus is well-suited to predictable production orders making a single item. Quality management is intense so that one-pass production is possible.
Six Sigma is actually a quality goal in the achievement to which a variety of QC approaches may be applied. It was initially named by Motorola and "six sigma" was achieved there. But the concept "developed legs" when Jack Welsh, the larger-than-life CEO of General Electric adopted it at GE. So what is Six Sigma? It means production in which 1 million pieces made will have virtually no defects. "Sigma" is the Greek letter used to designate the standard deviation from a norm as measured in statistics. If every part of a 1 million production run is defective, the sigma measurement will be infinite. If 10 percent of parts are defective, sigma will be 2.8; with 1 percent defect, the sigma increases to 3.8. Thus the higher the sigma, the greater the perfection. A sigma of 6.0 is reached when defects are down to 3.4 items out of a million, representing 0.00034 percent—in effect about as close to perfection as one can practically get. While Six Sigma is much discussed, it is not so much a method as the target of a QC effort or program.
The U.S. is not alone in generating ever-new buzzwords to be used in quality control. A new phrase from Japan is poka-yoke, meaning "avoid error." The coinage of a Toyota engineer, the concept refers to "designing in" methods of preventing mistakes. An example is a guard put on a drill press to stop the press before it drills too deep. The practice of poka-yoke lies in the identification of opportunities for "mistake-proofing" and then actions to make them happen. This approach is one variant for defect correction after the statistical charts show that something is amiss.
STATUS AND FUTURE
Quality control in the modern sense has come a long way since its initial formulation in the 1920s and its widespread adaptation in the 1950s, first in Japan, then across the world. It has become a routine part of many manufacturing processes, and, as signaled by Six Sigma, the goals are becoming ever-more ambitious. The more managerial and people-oriented aspects of the quality movement have a spottier history and have produced a series of initiatives that have come, have gone, and have resurfaced in various guises. Ever-new labels indicate attempts to capture something evidently difficult to hold for long: "quality circles," "teaming," the "learning organization," "knowledge management," "empowerment," and the like. The social phenomenon appears to indicate that it is ultimately easier to produce defect-free gadgets than perfect humans. The difficult we'll do immediately; the impossible will take a little time.
see also ISO 9000; Quality Circles; Total Quality Management (TQM)
Basu, Ron, and J. Nevan Wright. Quality Beyond Six Sigma. Elsevier, 2003.
Brownhill, Mark. "Beyond Poka-Yoke." Fabricating & Metalworking. February 2005.
Irwin, Stephen. "ISO 9000—A plus for airports: consultant Steve Irwin offers reasons why airports can benefit from a new standard." Airport Business. March 2006.
Juran, Joseph M. Architect of Quality. McGraw-Hill, 2004.
"The Life and Contributions of Joseph M. Juran." Carlson School of Management, University of Minnesota. Available from http://part-timemba.csom.umn.edu/Page1275.aspx. Retrieved on 12 May 2006.
Montgomery, Douglas C. Introduction to Statistical Quality Control. John Wiley & Sons, 2004.
"Real-World Six Sigma." Industrial Engineer. September 2005.
"Teachings." The W. Edwards Deming Institute. Available from http://www.deming.org/theman/teachings02.html. Retrieved on 12 May 2005.
What It Means
Quality control is a system by which products or services are inspected and evaluated to determine whether they meet expected levels of overall quality. Manufacturers monitor the quality of what they produce in order to ensure customer satisfaction, which subsequently helps a company reach its sales targets. Quality control standards are also used to determine whether or not a product is safe for consumers to use.
In manufacturing, the quality control process involves either analyzing a small, random sampling of a specific product or analyzing the process by which the product is made. In some cases inspectors conclude that certain improvements need to be made to a product before it achieves an acceptable level of quality or safety. All products, from automobiles to toys, are subject to quality control standards.
Although quality control is primarily used to evaluate quality in the manufacturing and engineering industries, quality control measures have also been implemented in other fields, notably education, health care, and the service industry. The service industry is a sector of the economy that includes restaurant service, retail sales, the distribution and delivery of goods, and other jobs that require interaction between employees and customers and therefore demand a high level of employee performance.
When Did It Begin
The concept of using a system to enforce quality levels dates to the Middle Ages, a historical era in Europe spanning from the fifth to the fifteenth century. During this period craftsmen formed guilds (associations of workers involved with the same trade), which created specific rules pertaining to issues of quality in the construction of tools, textiles, and other products. Craftsmen who did not observe the standards required by the guild (for example, by building a product of poor quality) faced fines or, in extreme cases, expulsion from the guild.
In the nineteenth century the manufacturing industry widely adopted the system of division of labor, in which individual workers are assigned responsibility for one specific aspect of the production process. With this system the responsibility for quality control shifted from the individual craftsmen to the foreman, a supervisor who oversees certain aspects of manufacturing operations. Another important development in quality control was statistical process control, a system that analyzes the process of production rather than the actual finished product. It was pioneered by Walter A. Shewhart (1891–1967), a prominent American engineer and statistician. During World War II (1939–45) statistical process control became the predominant form of quality control.
More Detailed Information
To gauge quality levels in a given product or service, modern systems of quality control rely predominantly on statistical analysis (the collection, evaluation, and interpretation of data). For this reason quality control is also known as statistical quality control. There are two principal forms of statistical quality control. The first form, usually referred to as acceptance sampling, focuses on an evaluation of the finished product. In acceptance sampling, an inspector determines the quality of a given batch of products by closely examining a small, random selection of products from that batch. If a certain percentage of the sample products fail to meet quality standards, then the entire batch of products is rejected.
For example, an inspector might be in charge of determining the quality of a new model of electric fan. Using the acceptance sampling method, the inspector chooses 10 fans out of 1,000 that have just been produced. He plugs in each of the fans to ensure that the fan’s electrical system is working (in other words, that the fan will turn on). If the fan has speed settings, he will test the settings to ensure that each is functioning at its proper speed. The inspector will also test the product’s safety. For example, he might measure the temperature on the surface of the fan to see whether or not it remains at a temperature that is safe for consumers.
The second principal form of quality control is known as statistical process control. Whereas acceptance sampling determines the quality of a batch of products by inspecting a select portion of those products, statistical process control focuses on the process by which the product is made. By using statistical analysis, or the analysis of specific product data, to determine the patterns by which certain problems occur, statistical process control can identify which aspects of the production process might need to be improved in order to ensure that the finished product meets quality control standards. With this method the fan inspector makes note of specific problems in a batch of products. If certain problems recur, the inspector submits his or her data to an analyst, who then studies the data to determine what aspect of the production process might be causing the problem. For example, if inspectors find a recurring problem with the stability of the fan blades, an analyst might determine that there is a problem with the stage of production that involves attaching the blades.
After World War II there emerged a new focus in quality control as more and more companies began to evaluate the performance of management (people such as employers, managers, and directors). These new methods evaluated the performance of the individual manager, measuring such criteria as job experience, knowledge, and skill. In addition, this new quality control model also examined the personality of the individual manager, paying particular attention to qualities of motivation, self-confidence, and the ability to work cooperatively with coworkers. This new focus on management performance is an integral part of a strategy called Total Quality Management, which is intended to monitor quality in all aspects of a company’s operations.
qual·i·ty con·trol • n. a system of maintaining standards in manufactured products by testing a sample of the output against the specification. DERIVATIVES: qual·i·ty con·trol·ler n.