Chapter 13: Six Sigma Management and Lean Tools Chapter Outline What is Six Sigma? Organizing Lean-Six Sigma DMAIC Overview Define Phase Measure Phase Analyze Phase Improve Phase Control Phase Taguchi Design of Experiments Background of the Taguchi Method The Taguchi Process Design for Six Sigma Lean-Six Sigma from a Contingency Perspective Overview In 1995, General Electric began implementation of Six Sigma with a goal of being a Six Sigma company by 2000. Six Sigma is a methodology that puts the entire quality atmosphere within a firm into focus. Six Sigma involves all of the topics that have been presented so far under a single organization. On page 337, the text provides a definition of Six Sigma: First, Six Sigma represents a well-thought out packaging of quality tools and philosophies in an honest effort to provide rigor and repeatability to quality improvement efforts. Second, Six Sigma is much more cost-reduction-oriented than traditional continuous improvement. This pragmatic, overall approach is what makes Six Sigma successful. Many companies have reported outstanding results with lean-Six Sigma. There are also many failures. Keys to lean-Six Sigma success are skilled management, leadership, and long-term commitment. Discussion Questions 1. Where do you think that Lean-Six Sigma can be used effectively? Actually, the concept works well within any type of organization. The primary challenge is to better define objectives with appropriate measurable criteria. The following is a list of U.S. companies that use Six Sigma: 3M, A.B. Dick Company, Abbott Labs, Adolph Coors, Advanced Micro Devices, Aerospace Corp, Airborne, Alcoa, Allen Bradley, Allied Signal, Ampex, Apple Computers, Applied Magnetics, ASQC, Atmel, Baxter Pharmaseal, Beatrice Foods, Bell Helicopter, Boeing, Bombardier, Borden, Bristol Meyers - Squibb, Bryn Mawr Hospital, Campbell Soup, Cellular 1, Chevron, Citicorp, City of Austin, TX, City of Dallas, TX, Clorox, Cooper Ind, Dannon, Defense Mapping Agency, Delnosa ( Delco Electronics in Mexico), Digital Equipment Corp, DTM Corp, Eastman Kodak, Electronic Systems Center, Empak, Florida Dept. of Corrections, Ford Motor Company, GEC Marconi, General Dynamics, General Electric, Hazeltine Corp, Hewlett Packard, Holly Sugar, Honeywell, Intel, Junior Achievement, Kaiser Aluminum, Kraft General Foods, Larson & Darby, Inc, Laser Magnetic Storage, Lear Astronics, Lenox China, Littton Data Systems, Lockheed Martin, Loral, Los Alamos National labs, Martin Marietta, McDonnell Douglas, Merix, Microsoft, Morton Int'l, Motorola, NASA, Nat'l Institute of Corrections, Nat'l Institute of Standards, National Semiconductor, Natural Gas Pipeline Company of America, Northrop Corp, PACE, Parkview Hospital, Pentagon, Pharmacia, PRC, Inc, Qualified Specialists, Ramtron Corp, Rockwell Int'l, Rohm & Haas, Seagate, Society of Plastics Engineers, Solar Optical, Sony, Star Quality, Storage Tek, Symbios Logic, Synthes, Technicomp, Tessco, Texaco, Texas Commerce Bank, Texas Dept. of Transportation, Texas Instruments, Titleist, Trane, TRW, Ultratech Stepper, United States Air Force, United States Army, United technologies, UPS, USAA, Verbatim, Walbro Automotive, Walker parking, Woodward Governor, Xerox It’s obvious that Six Sigma has a broad ranging attraction. 2. How will risk assessments vary from industry to industry? Figure 13.6 from page 345 presents the Six Sigma Project Worksheet. As you look at the questions, it becomes apparent that various industries will accentuate different metrics. 3. What industries would be the best candidates for the Lean-Six Sigma approach? Why? The list of companies that use Six Sigma listed in question 2 makes it obvious that many companies in many industries have used Six Sigma. The list includes hardware companies such as A.B. Dick, Apple Computers, and Textronics, aerospace companies such as Bell Helicopter, Boeing, Lockheed Martin and Northrop, and pharmaceutical companies such as Abbott Labs and Bristol Meyers. The industry is unimportant. The focus is on companies who have a product to sell. That product can be either manufactured or a service 4. What is different between Lean-Six Sigma and traditional quality improvement? Lean-Six Sigma is a documented methodology for insuring that the continuing quality program is effective. Lean-Six Sigma is focused on cost reduction. If properly applied, Six Sigma should pay for the expense of its implementation. Traditional quality improvement tends to be an add-on to the existing systems. The results of this add-on approach are not predictable. By proceduralizing the effort with a thoroughly documented approach, the results are more predictable and reliable. 5. Can you think of an example where the Taguchi quality loss function (QLF) would work in real life? Discuss how it might work. QLF defines a situation where a problem has a “potential loss to society” (page 359). This function defines all of the losses due to poor quality. Taguchi says that, in addition, these losses are usually associated with lack of quality. There is also a potential for environmental damage. In the mountains of Colorado, closed gold and silver mines are a frequent sight. These old mines show the effects of “mine tailings,” or the waste products from the mining operation. These mine tailings usually contain dangerous chemicals that were used in the operation. The effect of the damage that they caused is still obvious in many cases. Governmental requirements and social pressure have forced mining operations to “clean up their act.” This situation has led to a more efficient operation as well. 6. How does the Taguchi concept of ideal quality compare to other definitions of quality discussed in Chapter 1? Traditionally, manufacturing focuses on the concept of tolerance ranges. In other words, a manufactured product can have a dimension that fits within a specified range to be successful. Taguchi’s concept of robust design focuses on target values. The target value states that one should aim for the best result instead of one “that works.” Figure 13-19 describes the Taguchi process. Note its resemblance to the traditional scientific method. 7. How could the Taguchi method be used to design a course in quality management? Identify all the variables, measures, and objectives. If the discussion involves using the course content as the topic for the Taguchi method, the topics for analysis and discussion might include: Control variables: Mix of topics taught Time spent on each topic Number of exams given Number of assignments given Amount of time spent lecturing versus doing activity-based learning in class, etc. Noise variables: Intelligence of the students Ambient temperature in the classroom How well the school football does in a particular year Measures/objectives: Test scores: higher is better Amount of stress on the students: nominal is best Starting salaries: the higher the better This analysis of the three main topics – control variables, noise variables, and measures/objectives – is the focus of the Taguchi method. Taguchi Method for Designing a Quality Management Course 1. Objective: To design a course in quality management that maximizes learning effectiveness while minimizing variability in student performance. 2. Key Variables: • Factors (Controllable Variables): • Teaching Method: Lectures, case studies, group projects. • Course Materials: Textbooks, online resources, video tutorials. • Assessment Types: Exams, quizzes, practical assignments. • Class Format: In-person, online, hybrid. • Student Interaction: Group discussions, peer reviews, individual assignments. • Noise Factors (Uncontrollable Variables): • Student Background Knowledge: Prior experience with quality management. • Learning Environment: Physical/technical distractions. • Student Engagement: Motivation, participation levels. 3. Measures (Responses): • Learning Outcomes: Exam scores, project grades. • Student Satisfaction: Course evaluations, feedback surveys. • Retention of Knowledge: Performance on post-course assessments. • Engagement Metrics: Attendance, participation in discussions, assignment submission rates. 4. Objectives: • Maximize: Learning outcomes and student satisfaction. • Minimize: Variability in student performance across assessments, course completion time, and engagement gaps. 5. Experimental Design: Use Taguchi’s Orthogonal Arrays to test different combinations of factors (e.g., lecture vs. case studies, in-person vs. online) while controlling for noise factors. Analyze results to identify the optimal course design that yields the best overall learning and engagement with minimal variation. This approach ensures a robust course design that adapts to varying student needs while maintaining high learning quality. 8. Why are behavioral processes such as brainstorming important for the Taguchi method? A properly conducted brainstorming session or Delphi study obtains the expertise and experience of the people who have been involved in the process. This is extremely important. This knowledge is a corporate asset that has been gathered at no small expense over many years. Shortcutting the process by bypassing the behavioral process may save money in the short run, but in the long run, this shortcut can be very expensive. For additional discussion and outside reading, one might look into the concept of inclusion. If the homegrown subject matter experts (SMEs) are not included in the preliminary stages of a project, management risks alienating them. However, including the SMEs by seeking their council will provide a sense of inclusion, which will aid in assuring their cooperation in the effort. 9. How would you cost-benefit a Taguchi experiment? What might be some of the quantifiable parameters you would use in evaluating the worth of a Taguchi experiment? The QLF is specifically identified as the cost of deviation from a target value. Examining historic data can identify this cost. The Taguchi method is based upon specific metrics or quantifiable values. Through examining these values, one can compare the differences between historic values and anticipated expenses. A traditional cost benefit analysis can then be run to look at the cost of implementing a Taguchi experiment, the traditional costs, and the implementation cost. For example, a company recently reported that an $80 million product launch was in danger due to poor performance of the new production process. In this case, the Taguchi experiment was easy to justify. The QLF can be used to evaluate possible losses to society. 10. The chapter cites different services implementations of the Taguchi method. Do you think the Taguchi method is useful for services? Why or why not? Why do you think the technique has not been widely adopted in services? On page 357, the point is made: The Taguchi method is a standardized approach for determining the best combination of inputs to produce a product or service. This is accomplished through design of experiments (DOE) for determining parameters. The Taguchi method parses the procedures and examines their effectiveness. This is equally effective in both a manufacturing procedure and a service procedure. There are metrics that can be analyzed in either situation. In the case of a service such as a CPA’s audit, one can look at the time spent, the level of items identified for further examination, and the results of the audit. These are all quantifiable and measurable. Case 13-1: The Neiman-Marcus Cookie Using this recipe and these production procedures, develop a Taguchi experiment to find the optimal process for making chocolate chip cookies. Be careful in identifying control factors, noise factors, objectives, and designing the experiment the whole experiment. We will establish a simple Taguchi experiment. Each response will be different, so this example is instructive of a simple answer. Control factors: Variation in the quantities of each of the following ingredients: Blended oatmeal Brown sugar Sugar Butter Soda Hershey bar, grated Baking powder Vanilla Flour Chocolate chips Salt Eggs Chopped nuts Variation in baking time, temperature, types of flour or other ingredients, etc. Noise factors: These are things like uncontrollable variation in incoming ingredients, ambient temperature, humidity, etc. Objectives: If there is a subjective evaluation of taste, then the objective is the more, the better. If there is a desired number of chocolate chips in each cookie, then nominal-is-best is the objective. If you desire a certain level of crispiness, then the objective would be nominal-is-best for the crispiness measure. Let’s set up a simple experiment. Suppose we choose three variables with two levels each as follows: Variable 1: Amount of oatmeal chips Levels: 1 – 4.5 cups; 2 – 5 cups Variable 2: Amount of sugar Levels: 1 – 1.75 cups; 2 – 2 cups Variable 3: Amount of soda Levels: 1 – 1.5 tsp; 2 – 2 tsp Since three variables with 2 levels were chosen, an L4 array is chosen. Here is the array: Therefore, 4 iterations of the experiment will be run with the settings as shown in the above orthogonal array. This means that for the first experiment, the levels are run using the first levels for each variable. During the second run, the first level is used for variable 1 and the second levels for variables 2 and 3. Suggested Answers to End of Chapter Problems 1. Part of a Six-Sigma project is to identify Xs and Ys. What are the Xs you can identify for student satisfaction with a quality management course? Use the Y (dependent variable) of student satisfaction with a quality management class. Independent variables (Xs) include items such as: Preparation of the instructor Knowledge of the instructor Room setting Quality of the materials used in teaching 2. Part of a Six-Sigma project is to identify Xs and Ys. Identify Xs and Ys for an athletic director of a major university (insert the name of your university here) who is interesting in increasing attendance at football games. Y = game attendance Xs include: Win/loss record National ranking Ticket price Tastiness of the nachos 3. Part of a Six-Sigma project is to identify Xs and Ys. Identify these variables for the owner of a copy shop who is interested in reducing mistakes in orders. Y = copy shop defects Xs include: Type of job Materials Labor 4. Develop a Problem Definition (see Figure 13-10) for the project in Problem 3. Problem Statement: In 2009, the copy center lost $X from scrap and rework resulting from mistakes in orders. This has resulted in a loss of profitability for the firm. Project Goals/Objectives: Reduce COPQ by 75% in 2007. Primary Metrics: COPQ, returned jobs, and scrap Secondary Metrics: Lost future sales, lost time, labor productivity, and lost capacity Team Members: Dolly, Garth, Alan, and Reba 5. Complete the XY matrix for the following data: Here is the solution in spreadsheet form: Which inputs are the most important? Input 4 6. Complete the XY matrix for the following data: Here is the solution. Which inputs are the most important? Inputs 3 & 4 7. Find the QLF for the following information: C=300 T=.25 V=1/3 K = C/T2 = 300/.252 = 4800 L = K * V2 = 4800 (1/3)2 = $533.33/unit loss 8. Compute the QLF for the following information: C = 250 T = .40 V = 1/6 K = 250/.402 = 1562.5 L = 1562.5(1/6)2 = $43.40/unit loss 9. It costs $50 to repair a component in a VCR. Compute the QLF for losses incurred as a result of a deviation from a target setting with a nominal tolerance of 10 plus or minus .25 mm required. The mean squared deviation is ½. K = C/T2 C = 50 T = .25 V2 = .5 K = 50/.252 = 800 L = 800(.5) = $400/unit loss 10. It costs $350 to repair a refrigerator compressor. Compute the QLF for losses incurred as a result of a deviation from a target setting with a nominal tolerance of 60 amps where a 2-amp variation is acceptable. The mean squared deviation is 1/5. C = 350 T = 2 V2 = 1/5 K = 350/22 = 87.50 L = 87.50 (1/5) = $17.50/unit loss 11. For a component, the following measurements were taken: If the nominal target value is 2 + .05, compute the QLF for this component, where the repair cost is $200. C = cost per unit T = .05 V = .00095 K = C / .052 L = .00095 x C / .052 = $.38*C / unit 12. Below are answers to the worksheets in Figures 13-6 and 13-7. Using these responses, develop a risk assessment for this project. Produce a risk and return matrix to determine if this project is worth pursuing. Use the weights in Figure 13-6. Risk Value = 54% Return Value = 8/15 = .533 = 53% This project is a “question mark.” The success of this project is uncertain. 13. Here are the answers to the worksheets in Figures 13-6 and 13-7. Using these responses, develop a risk assessment for this project. Produce a risk and return matrix to determine if this project is worth pursuing. Use the weights in Figure 13-6. Risk Value = 36% Return Value = 12/15 = .80 = 80% This project is a star. It appears to have strong potential. 14. Using the following data, perform a gauge R&R analysis where there are two replications for each part number. Anova: Two-Factor With Replication Both sample and columns p-values are significant. Both parts and operators are significant. 15. Perform a gauge R&R analysis where there are three replications. What are your findings? Anova: Two-Factor With Replication The sample p-value is significant. Only parts are significant. Operator effects are insignificant. Solution Manual for Managing Quality: Integrating the Supply Chain Thomas S. Foster 9780133798258
Download
Close
Close
