1 Introduction to Statistical Quality Control The material contained in the introduction has been taken from the following

g sources: Farnum, N.R. (1994), Modern Statistical Quality Control and Improvement (Chapter 2), Duxbury, New York. Montgomery, D.C. (2009), Introduction to Statistical Quality Control (Chapter 1), Wiley, New York. Snee, R.D. (1999), Why Should Statisticians Pay Attention to Six Sigma?, Quality Progress, September, pages 100-103. Wadsworth, H.M., Stephens, K.S., and Godfrey, A.B. (1986), Modern Methods for Quality Control and Improvement (Chapters 1,2), Wiley, New York. 1.1 The Meaning of Quality and Quality Improvement Quality is conformance to requirements or specications. This is the denition used by Crosby who suggests that in order to manage quality adequately, we must be able to quantify and measure it. Quality is tness for use. This is a more general denition attributed to Juran. It places emphasis on the consumer aspect of quality. It emphasizes that requirements and specications translate tness for use into measurable quantities. Quality is inversely proportional to variability. This denition of D. Montgomery (your textbook author) implies that if the variability in a products important characteristics decreases, the quality of the product increases. Thus, quality improvement is the reduction of variability in processes and products. Quality is the totality of features and characteristics of a product or service that bear on its ability to satisfy given needs. This denition was the 1978 consensus denition in ANSI/ASQC Standard A3. If you

replace the word given with stated or implied you have the ISO denition. It implies that quality is an attribute of a product. A product is a quality product if it meets all the requirements established for it; that is, it is a defect-free product. Quality is the loss imparted to society from the time the product is shipped. This denition of Dr. Genichi Taguchi represents a signicant change in how quality was viewed. Instead of trying to dene all the good things that a quality product should be, Taguchi focuses on a quality from a loss perspective. That is, he emphasizes minimizing the losses a product can create. That is, when a product fails to perform correctly, when it breaks down, or when some of its parts do not conform to specications, it creates undesirable costs to society (which includes the manufacturer and the customer). These costs can be measured in terms of the energy and time to x the problem, as a monetary expense, as loss of customer goodwill, or as lost market share. A companys quality goal is to minimize any possible loss to society. By focusing on the loss to society, it draws attention to all problems that may eventually result in the loss of customers. 4 The quality of a product can be evaluated in many ways referred to as dimensions of quality. Therefore, it is important to dierentiate among these dimensions: 1. Performance: Will the product do the intended job? 2. Reliability: How often does the product fail? 3. Durability: How long does the product last? 4. Serviceability: How easy is it to repair the product? 5. Aesthetics: What does the product look like? 6. Features: What does the product do? 7. Perceived quality: What is the reputation of the company or its product? 8. Conformance to standards: Is the product made exactly as the designer intended?

 Quality improvement is the reduction of variability in processes and products. One goal of this course is to learn how to implement statistical tools that can be used to improve quality. An alternative denition of quality improvement is the reduction of waste. For every product, the user or consumer denes certain properties of perceived quality referred to as quality characteristics. Quality characteristics are often related to specications. For a manufactured product, specications dene a range of acceptable measurements for the quality characteristics of the product components. The desired value for a quality characteristic is called the nominal or target value. A target value is typically bounded by an acceptable interval of values that will not impact the performance of the product if the quality characteristic value is in that range. The largest allowable value for a quality characteristics is called the upper specication limit (USL), and the smallest allowable value for a quality characteristics is called the lower specication limit (LSL). A quality characteristic may only have one specication limit (USL or LSL, but not both). A nonconforming product is a product that fails to meet one or more of its specications. A specic type of failure is called a nonconformity. A nonconforming product is not necessarily unt for use. For example, a cosmetic nonconformity (like color irregularity) will not eect the tness of the product for use. A nonconforming product is considered defective if it has one or more defects. A defect is a nonconformity that is serious enough to signicantly aect the proper usage of the product. 1.2 History and Evolution of Quality Control and Improvement The need for inspection of manufactured products has been around as long as processes and factories that generate these products have existed. Before industrialization, individual workers or artisans inspected their own work.

 Industrialization made goods available to the masses, but introduced a basic change in commerce by separating the workers who made products from the customers who received them. This depersonalization introduces the need for producers to develop new means to monitor and control the quality of their products. 5 Inspectors were now employed to inspect the work of others. Defective products would presumably be found by the inspector and removed or repaired before shipment. Thus, the early focus of quality control was on inspection to identify faulty products and prevent them from being shipped to customers. As factories became larger, this burden became too much for the inspector and, consequently, quality control tools and techniques were developed to assist the inspector. Because it was recognized early that variation in measurements is a statistical phenomenon, many of the quality control tools and techniques are statistically based. Although the area of experimental design is a major statistical quality method, it will not be included in the following discussion. Earliest Developments With the industrial revolution, there were suddenly thousands of unskilled workers involved in high-speed manufacturing operations. Managers usually did not understand the manufacturing process and had little or no experience running large operation. In 1875, Frederick Taylor introduces management principles referred to as Scientic Management. He advocated breaking each process into smaller, more easily completed steps and then have procedures written which detailed how each of these steps should be carried out. Managers should manage the system and workers should do as they are told. The focus was on productivity. Productivity improved but quality remained a problem. In the early 1900s, the utilization of the assembly line was pioneered by Henry Ford in the

automobile industry. Ford developed automated assembly concepts, self-checking, and inprocess inspection. Much of the earliest work in quality control was centered in the Bell Telephone System. The need for some methods that could generate condence in the quality of the systems instruments was recognized early in the development of this company. Thus, the building of a national communications network created a need for rigorous inspection for assuring the quality of a product being supplied to other companies. Instead of having each company that receives the products do redundant inspections, the growing AT&T company (in 1907) reorganized and directed the Western Electric Co. to provide inspection and testing for the various Bell Operating Companies under the direction of the Engineering Department of AT&T. This inspection and testing was extended by 1908 to manufactured products, installed products, and purchased materials. Although very few statistical techniques existed in 1908, sampling was used extensively and eorts were made to judge conformity to requirements and to assess durability and reliability. By the 1920s, quality assurance was applied to all phases of quality: design, manufacture, and installation. Introduction of the panel switching machine (then a very complex electromechanical device) led to the necessity for a large number of inspectors. The burden was on making the product right with the mantra of Do it right the rst time. To the current day, many industries have adopted this philosophy to increase productivity and to lower costs. In 1925, the members of the Inspection Department were transferred to the newly formed Bell Telephone Laboratories. This department contained Walter Shewhart, Harold Dodge, George Edwards, and others. These men were to become the founders of modern quality control 6methods. As their work proceeded to develop the art and science of inspection engineering, they coined the name Quality Assurance. The department evolved with Edwards as Director of Quality Assurance, Shewhart responsible for theory, and Dodge in charge of methods.

 From 1925 to 1941, the development of quality control and quality assurance methodology was remarkable. Statistical methods were developed that introduced the concept of monitoring a production process to determine if it was stable over time and capable of producing quality products. It shifted the focus from inspection of a nal product to studying the process that is used to make the product. In JASA (1925), the Shewhart paper The Application of Statistics as an Aid in Maintaining Quality of a Manufactured Product introduced the control chart. Within 25 years, the control chart was a basic manufacturing quality control tool used around the world. Also in 1925, Dodge introduced the basic concepts of sampling inspection by attributes and the concepts of consumers risks and producers risks. These concepts led to the risks of Type I and Type II errors in the testing of statistical hypotheses. Developments in the World War II Era At the beginning of American involvement in WWII, there was an immediate need for large quantities of war materials everything from guns to radar equipment to shoes. American industry rapidly expanded at the same time as the armed forces expanded. As a result, many people who had never before worked in manufacturing were employed in those new or expanding factories. Thus, the quality of the goods being manufactured suered from the lack of skilled personnel. Training programs which emphasized the use of control charts and acceptance sampling plans were then established by companies and by government organizations. For example, the War Department engaged Bell Telephone Laboratories to go to manufacturing plants and teach employees simple quality control methods. Other instructors were selected from local universities. In 1940, the War Department published a guide for using control charts to analyze process data. Between 1940-1943, Bell Laboratories developed the forerunners of the military standard sampling plans for the U.S. Army.

 One of the best-known individuals working with the War Department was Dr. W. Edwards Deming. He worked with the War Department on developing statistical techniques and taught many quality control courses. Deming later became one of the leading international quality consultants. Another development during this period was the establishment of two groups of statisticians and engineers to do research in the use of statistics in quality control. One of these was the Aberdeen Proving Ground in Maryland. This group was formed by Walter Shewhart and included Harold Dodge. Their work resulted in standards for statistical control of quality. The second group was the Statistical Research Group at Columbia University. Results from this group included the development of sequential analysis and the use of multivariate analysis in quality control with the introduction of the T 2 statistic by Harold Hotelling. In 1944, the rst journal devoted to scientic advances in statistical quality control, Industrial Quality Control, was published. (In 1969, it is replaced by Quality Progress and the Journal of Quality Technology). 7 At the end of the WWII, groups of faculty and graduate students in the training centers formed local quality control societies. In 1946, many of these societies were combined to form the American Society for Quality Control (ASQC). By the end of WWII, much of the manufacturing capability in Europe and Asia had been destroyed. American manufacturing capabilities, which had grown as part of the war eort, were ready to be converted to peacetime production. There was an insatiable demand worldwide for manufactured products. Thus, immediately following WWII, the guiding rule was that if you could make it you could sell it. U.S. manufacturers could sell whatever they produced, so the emphasis was now on

quantity of production and not on product quality. Many of the wartime quality lessons were forgotten or ignored, and customers did not seem to care because jobs were plentiful, wages were good, and if something broke, you could always replace it. That is, the U.S. developed into a throw-away society: one bought a product and when it failed to perform satisfactorily, threw it away, and bought another. Postwar Japan faced an entirely dierent situation. Food and shelter was scarce, and the factories were in ruins. Japan seemed unable to produce manufactured products at a level of quality acceptable to the rest of the world. In short, Made in Japan meant junk. Japan assessed and corrected the causes of their failure. The Japanese Union of Scientists and Engineers took the initiative to deal with this problem that threatened Japans economic existence. In 1950, they invited Deming to explain his ideas for quality improvement to leaders of Japanese industry. Dr. Deming had been preaching quality management in the U.S., but no one was listening because everyone was too busy making money. The Japanese listened to Deming and others (e.g., Juran) and soon mastered the inspection and process control concepts and turned their industries around. In 1951, in his honor, the top Japanese named the individual and plant quality awards the Deming Awards. The Japanese went on to develop their own quality improvement techniques (e.g. Taguchi and Ishikawa). By the 1970s, they had achieved world leadership in quality. Modern Developments One of the earliest techniques developed was the control chart. Its use in Shewharts original form is still widespread. When employed this way, it is commonly called a Shewhart chart. Other types of control charts have been developed in recent years. These include the cumulative sum (CUSUM) chart and the exponentially weighted moving average (EWMA) chart which use past history in the decision-making procedure. As a result, both charts have the ability of detecting small parameter changes quicker than the Shewhart chart. CUSUM and EWMA charts were introduced, respectively, by British statistician E.S. Page in 1954 and S.

Roberts in 1959. Many acceptance sampling standards and procedures were originally developed as military standards by the War Department (United Stated Department of Defense). Although many are still commonly known by their military names, in the U.S. most standards are currently published by the American National Standards Institute (ANSI) and are called voluntary standards. ANSI is responsible for overseeing the quality control standards in the U.S.. 8 Since WWII, the subject of reliability has developed rapidly. Much of the early work in this eld was in the aerospace and electronics industries. From the reliability eld, concepts of life testing of products has evolved. This will be discussed later in the course. Leaders in Total Quality Control: One of the earlier ideas (1951) was called Total Quality Control and was suggested by A.V. Feigenbaum of the General Electric Co.. Feigenbaum stressed that responsibility for quality control involved all departments and not just the quality control department. An outline of the entire scope of managing quality, often called Total Quality Management (TQM), was presented by J.M. Juran. His pioneering work in postwar Japan has been credited with forming the basis for much of Japans quality successes. Juran stresses the necessity for leadership by top management, extensive training programs to make basic quality control techniques available to every worker in the company, and a dedication to quality at every step: market research, product design, vendor relations, manufacturing, delivery, and service. Philip Crosbys eorts focused on quality improvement and reducing the cost of quality. Kaoru Ishikawa made signicant contributions to Japanese quality. He stated that to achieve company-wide quality control the stress is to be placed on the participation of all in quality control work from the president downward to foremen and operators. It is interesting to note that many of the pioneers of the quality movement (including Shewhart, Deming, Juran, Crosby, and Ishikawa) were trained as physical scientists or engineers.

Statistical methods provided the tools needed to perform their jobs as scientists or engineers. In the late 1970s and early 1980s, a number of major changes began to take place in American industry. The automotive sector was in a serious slump and thousands of workers lost their jobs. The ripple eect into other parts of the economy triggered a worldwide recession. It was in this same era that the Japanese quality eorts of 25 years, began to pay o. By the mid-1980s, the balance of trade was the worst it has ever been in the history of the U.S.. Many U.S. companies then developed quality programs and training courses. Unfortunately, these occurred at the end of the manufacturing process after the product was manufactured. The R&D (Research and Development) groups were excluded from QC activities. American industries were treating the symptoms but ignoring the root causes for poor quality. Eventually, many U.S. industries began to adopt Japanese strategies (which were often based on earlier U.S. strategies) to improve product quality. ISO (International Organization for Standardization) was founded in 1946 to establish a series of international standards for products and production processes. Currently, over 90 countries are members. In 1987, ISO published the ISO 9000 series of generic standards for quality management and quality assurance. These standards apply to companies of every size and to all industrial sectors goods, services, and information. Various extensions of the original ISO 9000 standards have been introduced since 1987. ISO 9000 standards provide guidance for suppliers who want to implement eective quality systems. They can also be used by customers to evaluate the adequacy of a suppliers quality system. To avoid the need for each customer to check each supplier, a system of registration was established. ISO audits a suppliers quality system and certies 9those suppliers that are in compliance with ISO standards. A customer can ask a supplier to register as a precondition to placing an order with that supplier. Registration gives

the customer an assurance that the supplier will provide quality products. In 1988, the Malcolm Baldridge National Quality Award was established by the U.S. Congress. More recently, the focus of quality has expanded beyond the production industries to the service industries (which deal with the performance of labor for the benet of another). Thus, greater emphasis is placed on human factors and their impact on service quality. Examples: food services, health care, cleaning and building services, transportation, education, nance, and government. In 1997, many companies begin to adopt Motorola six-sigma approach to quality. After 2000, the ISO 9000:2000 standards are issued. 1.3 Statistical Methods for Quality Control and Improvement These will be discussed throughout this course. 1.4 Management Aspects of Quality Control Quality Planning Quality planning is the development of strategic activities designed to improve the quality of a product. The planning will include both statistical and management activities. Quality Assurance Another term related to an overall approach to quality control is quality assurance. It is designed to include checks and controls on the quality control procedures in use. For a specic product or service, this may involve verication audits and the evolution of the quality factors that aect the specication, design, production, procurement, installation, test and inspection, sales, marketing, and use of the product or service. Quality assurance makes sure that quality control is doing what it should be doing. Such an eort should be organized within a factory, company, or corporation and motivated by management. Quality Assurance is a system of activities whose purpose is to provide an assurance that the overall quality control is in fact being done eectively.

 Quality Assurance: All those planned or systematic actions necessary to provide condence that a product or service will satisfy given needs. This is the ANSI/ASQC Standard A3 denition (1978). Quality Control Quality control is the regulatory process through which we measure actual quality performance. This is the denition given by Juran (1974). It puts quality into an action mode that makes use of standards and measurements. 10 Quality control: The operational techniques and the activities which sustain a quality of product or service that will satisfy given needs; also the use of such techniques and activities. This is the ANSI/ASQC Standard A3 denition (1978). The Statistical Control of Quality is application of statistical principles and techniques in all stages of design, production, maintenance and service, directed toward the economic satisfaction of demand. This denition was given by Deming (1971) emphasizes the statistical applications, and it also makes explicit the economic objectives of quality control. Statistical process control (SPC) can be considered as statistical quality control applied to a process or to a product resulting from a process. The DuPont Co. denition of SPC is: SPC is the totality of all process activities directed at improving process consistency through detecting changes in measured characteristics, identifying causes of changes, and preventing recurrence of those causes. Total quality is an eective system for integrating the quality-development [planning], qualitymaintenance [control], and quality-improvement eorts of the various groups in an organization so as to enable production and service at the most economical levels which allow for full customer satisfaction. This denition of total quality control was given by Feigenbaum (1983) and stresses

managerial and systemwide aspects. Benets of Quality Control in Industry Improving the quality of products. Increasing the productivity of manufacturing processes. Reducing manufacturing and corporate costs. Determining and improving the marketability of products. Reducing consumer prices of products. Improving and/or assuring on-time deliveries and availability. Assisting in the overall management of an enterprise. Design Quality vs Manufactured Quality Design quality refers to the quality of those product characteristics that will appeal to potential customers. It takes into account what it will cost to produce the product and what the customers are willing to pay for the product. Design quality is an important aspect of quality to be considered in planning the design, the manufacture of a product, and the marketing of a product. It can be considered as the potential for achieving manufactured quality. After a product has been designed and the manufacturing process has begun, the process does not always produce each unit in conformity with the design. For example, defects arise in materials, parts, subassemblies, assemblies, and in the nal product. When these defects are not related to the design, but are related to an inadequate, poorly planned, poorly controlled manufacturing process, we are concerned with manufactured quality. 11 Any defective or nonconforming materials, parts, assemblies, and nished products that are discarded or reworked during the manufacturing process result in increased cost and customer dissatisfaction. The lost time and eort in producing the defective or nonconforming product, the loss of business resulting from delays in delivery, and other associated costs attributable to a poorly manufactured product are the consequence of manufactured quality (also known

as the quality of conformance to design). Higher design quality usually means higher cost, but higher manufactured quality usually means lower cost. Because the goal should be to provide good quality at low cost, both aspects of quality need to be considered simultaneously. Quality is also an attitude of mind. At the corporate level, insucient preproduction testing or design qualication may result in the necessity to recall (e.g. automobiles, food products) and repair or replace the product. Understanding of quality concepts contributes to correct implementation and management of product quality, and the entire production enterprise benets. A manufacturer must be aware of the physical, cultural, and operational environment in which the product will be used. If quality is understood, not just as the tness for use of the end product but as the performance of the entire production enterprise, great reductions in cost are possible. The Deming Philosophy W. Edwards Deming is the most recognized name in quality management. His believed that the primary responsibility for quality lies with management. He often stated that 85% of quality-related problems can only be solved by management and not the workforce. This is the key to the Deming philosophy and the following summary of his recommendations to management (known as Demings 14 points for management). 1. Create a constancy of purpose toward improvement of products and services with the aim to become competitive, stay in business, and to provide jobs. This can only be achieved if requisite resources and eort are devoted to this end. 2. Adopt the new philosophy. That is, a company should reject old beliefs about quality that some defectives, poor workmanship, and service are to be tolerated. The company must sincerely strive to put quality as a prime objective. 3. Cease dependence on inspection to achieve quality. Eliminate the need for mass inspection by building quality into the product in the rst place. Wherever possible, seek to replace

inspection for defects with statistical process control methods for prevention of defects. 4. Vendors should be chosen on the basis of their ability to supply quality products, and not solely on their ability to supply low-cost products. Eliminate the practice of giving contracts to the lowest bidder without regard to quality. Take advantage of building long-term supplier relationships. 5. Seek continuous improvement everywhere to improve quality and productivity, and thus constantly decrease costs. Continuous improvement is to be sought in all the companys activities, from marketing, accounting, and manufacturing to payroll, engineering, and procurement. 6. Institute training on the job. Train the workforce in the tools, both statistical and nonstatistical, that are needed to improve quality. 127. Improve supervision, empower workers to make decisions about a process, and require management to make necessary process changes and exhibit leadership in quality improvement. The aim of supervision should be to help people and technology to do a better job. 8. Eliminate employees fear of asking questions about processes and procedures so that everyone may work eectively for the company. 9. Break down barriers between departments. Dierent departments should become familiar with one anothers operations and take advantage of this knowledge to improve products and processes. 10. Numerical goals, productivity targets, quality targets, and so on should be eliminated. Instead, emphasize continuous improvement. Stressing targets will not increase quality, but stressing quality will achieve such goals. Stressing targets often creates adversarial management/workforce relationships because the majority of causes of low quality and low productivity belong to the system, and thus like beyond the power of the workforce. 11. Work standards and quotas are another form of numerical target and should be eliminated. Instead, substitute improvement aids and helpful supervision. 12. Remove barriers that rob the hourly worker of his/her right to pride of workmanship. Give

workers the power to control their particular process or operation. Similarly, remove barriers that rob people in management and engineering of their right to pride of workmanship. Eliminate performance reviews and merit systems for promotions and salary. 13. Everyone should be trained in quality techniques. In particular, the quality department (if one exists) is no longer solely in charge of determining quality or using statistical methods. 14. Only top management can create the atmosphere and structure that will make these points work. Demings philosophy requires great change in management style for many companies. Under his system, managers are required to be leaders and facilitators, not merely authority gures. They must play an active role in changing systems to improve quality, and not play the passive role of creating targets that the workforce is required to meet. The text also includes Demings Seven Deadly Diseases of Management and Twelve Obstacles to Success. Jurans Organizing for Improvement Joseph Juran was a quality pioneer whose activities paralleled that of Deming. Like Deming, Juran was also invited by the Japanese. He conducted quality control seminars beginning in 1954, and had consistently been at the forefront of quality improvement. Unlike Deming who emphasizes statistics and the role of management, Jurans strengths are in the implementation and organization for change. He describes his steps for solving quality problems as a breakthrough sequence which are summarized below: 1. Establish the existence of a problem. That is, present data and arguments that prove to others, especially management, that a problem exists and is worthy of study. 2. Pare down the list of possible projects to the vital few whose solution will provide the greatest impact. Pareto charts are used at this stage. 133. Organize for change. Obtain agreement on the projects aim, obtain necessary permissions

and authority, obtain necessary diagnostic skills, establish time horizons, and plan for implementing the solution. 4. Collect and analyze the data. Statistical tools are used at this stage. 5. Ascertain the eect (on people as well as on processes) of proposed changes and try to limit resistance to these changes. 6. Make the changes. 7. Put controls in place to maintain the improvements (referred to as foolproong by Juran). Jurans breakthrough sequence is more of a problem-solving technique than it is a list of management objectives. From a management viewpoint, Juran (like Deming) subscribes to the view that the majority of quality problems can only be xed by management and not by the workforce. Other Philosophies There have been other quality management approaches tried by companies. These include Total Quality Control and Total Quality Management (TQM). These are discussed briey in the text. Currently, the most popular approach is the six-sigma approach. The Six-Sigma Approach The goal of the six-sigma approach to business improvement is to nd and eliminate causes of mistakes or defects in business processes. To achieve this goal, the focus is placed on outputs that are of critical importance to customers resulting (hopefully) in enhanced process performance, improved customer satisfaction, and increased savings and revenues. Six sigma is strategic approach designed to work across all processes, products, and industries. The principles not only apply to manufacturing and electronics (where it originated at Allied Signal, Motorola, and General Electric), but to any type of business (e.g., banking and nancial services, chemicals, pharmaceuticals, utilities, health care, etc.). While many six sigma tools are not new, the approach and deployment of the tools are unique (especially to those accustomed to using traditional statistical methods). Six sigma utilizes a standard deviation (sigma) which is a measure of the capability of the

process to achieve defect-free performance where a defect is dened as anything that results in customer dissatisfaction. The ability to produce products and services with only 3.4 defects per million is described as six sigma and is considered as world-class performance for many processes. The six sigma approach combines statistical methodology with quality processes to create a new methodology. This knowledge is applied to process improvement projects with the goal of increasing bottom-line numbers and customer satisfaction. The following are three aspects of the six-sigma method that are not emphasized in other approaches: 1. Integrating the human element of improvement (e.g., bottom line, management leadership, sense of urgency, customer focus, project teams, culture change) and the process elements of improvement (e.g., analysis of process variation, quantitative measures, statistical methods, process management). 142. Focusing on the bottom line. (No six sigma project is approved unless the bottom line impact has been identied.) 3. Linking improvement tools into an overall approach. (The ve-phase improvement process design, measure, analyze, improve, control (DMAIC) is the template.) This is reviewed in Chapter 2 of the text. From a statistical point of view, consider the properties of the normal distribution in terms of ppm (parts per million) defective. The following table summarizes the expected number of defective items when LSL and USL are set at target k: Specication Limits Percent Inside Specs ppm Defective 1 sigma 68.27 317300 2 sigma 95.45 45500 3 sigma 99.73 2700

4 sigma 99.9937 63 5 sigma 99.999943 0.57 6 sigma 99.9999998 0.002 At rst look, this table seems reasonable for quality performance. However, this corresponds to a single quality characteristic. What happens if a product has multiple quality characteristics and each one has to meet specications. If the characteristics are independent of each other, and if any one of them is outside of specications, then the ppm defective will be much higher than the table values. For example, if a product has n independent components, then the probability all n components meet specications is :9973 n . For n = 100, this would equal .763 which is unacceptably high. To address this problem, Motorola decided to adopt six-sigma quality. That is, they adopted 6 limits which is 2 parts per billion defective. In the case with n = 100 components, the probability that a unit is nondefective is :999998 or :2 ppm. 2 The DMAIC Process The objective of the design (D) step is to identify opportunities and to verify/validate the potential for improvement. The objective of the measure (M) step is to evaluate and understand the current state of the process. This includes collecting data on measures of quality. The objective of the analyze (A) step is to use the data from the measure step to determine cause-and-eect relationships in the process and to understand the dierent sources of process variability. The objective of the improve (I) step is to use the results of the measure and analyze steps

to determine specic changes to implement to improve the process. The objective of the control (C) step is to control the improvements made until future improvements can be made. A detailed discussion of each of these ve steps in given in Chapter 2 of the text. 15