Six Sigma - “A rigorous and disciplined methodology that utilizes data and statistical analysis to measure and improve a company’s operational performance, practices and systems. Six Sigma identifies and prevents defects in manufacturing and service-related processes. In many organizations, it simply means a measure of quality that strives for near perfection.”
(http://www.dmreview.com/resources/glossary.cfm?keywordId=S)
Introduction
Since its inception in the late 1980s Six Sigma has become one of the pre-eminent quality management models in the business world. In fact, today the term “Six Sigma” has almost become synonymous with the word quality. Unfortunately, many people who do not have first had experience with the Six Sigma quality model find the thought of utilizing Six Sigma intimidating. Often it is perceived that one must have a PhD in statistics to survive in a Six Sigma environment. This could not be further from the truth. Many skills taught in both graduate and undergraduate business modeling and business analysis classes can be employed in the Six Sigma process to gain valuable insight into processes being monitored. The intent of this paper is to explore the utility of general business modeling techniques (Cause and Effect/Root Cause analysis, Regression, Monte Carlo Simulation, Forecasting and Optimization) in the Six Sigma quality model. To do so, I will discuss the history of Six Sigma, provide a short overview of the Six Sigma model and discuss how each of the for mentioned business modeling techniques can be used to enhance the Six Sigma model.
History of Six Sigma
Six Sigma is a quality management model that can trace its roots all the way back to the 1920s and 1930s. During that time Walter Shewart, at Western Electric and Bell Labs helped to pioneer the use of statistics (and control charts) to monitor and improve production processes. His approach to process quality management came to be known as Statistical Process Control (SPC). SPC gained some popularity among engineers in the United States over the next few decades until World War II created an absolute need for quality production practices to support the war effort. During the war, one of Shewart’s students, W. Edwards Deming, came to the forefront of the SPC movement. Deming helped train SPC to engineers around the United States in support of the war effort. Unfortunately, after the war, industry in the United States largely strayed away from SPC. Deming, on the other hand, was brought to Japan after the war to help Japanese industry get back on its feet. He helped to implement SPC in Japan, but also proceeded to bring a larger quality management philosophy to the forefront as well. Deming understood that quality was more than statistics, but a paradigm that need to be implemented at the top levels of management if a company were to be successful. His rules for quality management within an organization “14 Points” became famous. Deming’s 14 points are listed below:
- Constancy of purpose – Create constancy of purpose for continual improvement of products and service to society, allocating resources to provide for long range needs rather than only short term profitability, with a plan to become competitive, to stay in business, and to provide jobs.
- The new philosophy – Adopt a new philosophy. We are in a new economic age, created in Japan. We can no longer live with commonly accepted levels of delays, mistakes, defective materials, and defective workmanship. Transformation of Western management style is necessary to halt the continued decline of business and industry.
- Cease dependence on mass inspection – Eliminate the need for mass inspection as the way of life to achieve quality by building quality into the product in the first place. Require statistical evidence of built in quality in both manufacturing and purchasing functions.
- End lowest tender contracts – End the practice of awarding business solely on the basis of price tag. Instead require meaningful measures of quality along with price. Reduce the number of suppliers for the same item by eliminating those that do not qualify with statistical and other evidence of quality. The aim is to minimize total cost, not mere initial cost, by minimizing variation. This may be achieved by moving toward a single supplier for any one item, on a long term relationship of loyalty and trust. Purchasing managers have a new job, and must learn it.
- Improve every process – Improve constantly and forever every process for planning, production, and service. Search continually for problems in order to improve every activity in the company, to improve quality and productivity, and thus constantly decrease costs. Institute innovation and constant improvement of product, service, and process. It is management’s job to work continually on the system (design, incoming materials, maintenance, improvement of machines, supervision, training, retraining).
- Institute training on the job – Institute modern methods of training on the job for all, including management, to make better use of every employee. New skills are required to keep up with changes in materials, methods, product and service design, machinery, techniques, and service.
- Institute leadership - Adopt and institute leadership aimed at helping people do a better job. The responsibility of managers and supervisors must be changed from sheer numbers to quality. Improvement of quality will automatically improve productivity. Management must ensure that immediate action is taken on reports of inherited defects, maintenance requirements, poor tools, fuzzy operational definitions, and all conditions detrimental to quality.
- Drive out fear - Encourage effective two way communication and other means to drive out fear throughout the organization so that everybody may work effectively and more productively for the company.
- Break down barriers - Break down barriers between departments and staff areas. People in different areas, such as Leasing, Maintenance, Administration, must work in teams to tackle problems that may be encountered with products or service.
- Eliminate exhortations - Eliminate the use of slogans, posters and exhortations for the work force, demanding Zero Defects and new levels of productivity, without providing methods. Such exhortations only create adversarial relationships; the bulk of the causes of low quality and low productivity belong to the system, and thus lie beyond the power of the work force.
- Eliminate arbitrary numerical targets - Eliminate work standards that prescribe quotas for the work force and numerical goals for people in management. Substitute aids and helpful leadership in order to achieve continual improvement of quality and productivity.
- Permit pride of workmanship - Remove the barriers that rob hourly workers, and people in management, of their right to pride of workmanship. This implies, among other things, abolition of the annual merit rating (appraisal of performance) and of Management by Objective. Again, the responsibility of managers, supervisors, foremen must be changed from sheer numbers to quality.
- Encourage education - Institute a vigorous program of education, and encourage self improvement for everyone. What an organization needs is not just good people; it needs people that are improving with education. Advances in competitive position will have their roots in knowledge.
- Top management commitment and action - Clearly define top management's permanent commitment to ever improving quality and productivity, and their obligation to implement all of these principles. Indeed, it is not enough that top management commit themselves for life to quality and productivity. They must know what it is that they are committed to—that is, what they must do. Create a structure in top management that will push every day on the preceding 13 Points, and take action in order to accomplish the transformation. Support is not enough: action is required!
(http://www.lii.net/deming.html)
Although Deming’s 14 Points seem obvious by today’s standards, at the time, they were ground breaking and became the overarching framework of what is known as Total Quality Management (TQM).
As we all know, the Japanese enjoyed great success after World War II, largely as a result of adopting Deming’s TQM philosophy. The United States however, largely abandoned TQM after the war and U.S. companies didn’t pick it back up again until the 1970s and 1980s. One of the companies that turned to SPC and TQM during the 1980s was Motorola. In 1886 Motorola was struggling with quality problems and losing its competition with foreign manufacturers. As a Senior Vice President of Sales commented, “Our quality stinks.” (https://mu.motorola.com). To fix their quality problems Motorola created their own quality management template, which includes philosophies and techniques from both SPC and TQM, and called it Six Sigma. By implementing their Six Sigma quality process Motorola was able to increase the quality of their products in a dramatic fashion. Just two years later, in 1988, Motorola became one of the inaugural winners of The Malcolm Baldridge Quality Award. Since that time many companies, including GE, Home Depot and Raytheon, have adopted the Six Sigma quality process, and today Six Sigma if the premier quality management model used is the business world.
What Comprises Six Sigma
The actual term Six Sigma is derived from the term sigma as it is used in statistics referring to the standard deviation of a process. The name Six Sigma refers to the idea of controlling the quality of outputs from a process so well that a defective output occurs outside of 6 standard deviations from the process mean. This level of control would translate into producing a defective output only 3.4 times per million outputs (This is based on the idea that the process mean can fluctuate by 1.5 sigma creating a realized sigma spread of 4.5). However, the Six Sigma process encompasses far more than measuring standard deviations from a process mean. As stated previously, Six Sigma is an overarching quality model that incorporates elements of SPC and TQM and utilizes them as part of a quality process roadmap. This roadmap, that is the basis of the Six Sigma model, is called DMAIC. (There is also a DMAIV roadmap, but that is outside the scope of this paper). The DMAIC roadmap for process improvement is as follows:
Define Phase – In this phase the goals and deliverables of the project are determined. A business owner (customer) is identified and resources that will be utilized are determined. Process flows are created and a project charter is created.
Measure Phase – The measure phase is primarily focused on collecting data samples and determining what metrics and statistical tools will be utilized. Specifically, the Y=f(x) relationship is developed and process capacity and sigma baseline is determined.
Analyze Phase – The analyze phase is focused on determining the sources of variation and defects. (What is the underlying problem?) A number of analytical tools are used in this phase including: Histograms, Pareto Charts, Time series analysis, Statistical analysis, etc.
Improve Phase – Making process improvements to eliminate defects is the key to the improve phase (Fixing the underlying problem) utilizing Brainstorming, Experimental Design, Simulation software etc.
Control Phase – Creating process controls to insure future compliance with quality standards determined in the define phase is the essence of the control phase. Once this is done successfully there should be little possibility of future failures of the process.
In addition, after each phase there is a “tollgate review” which is basically a phase specific check off list to make sure each phase is completed correctly. This procedure must be completed successfully before moving on the next phase.
Modeling Techniques
As proposed in the introduction of this paper, basic business modeling techniques can be utilized for monitoring processes in the Six Sigma methodology. Although no modeling process is perfect, basic business modeling techniques can often give satisfactory decision making information, or at least enough directional output to provide the identification of logical next steps. Following are several business modeling techniques and how they can be utilized in conjunction with the Six Sigma process.
Cause and Effect
Cause and Effect analysis, done in the Analyze step of the DMAIC roadmap, is an area of Six Sigma where business modeling can be of great assistance. A key concept within the Six Sigma is the ability to determine the relationships between process inputs and outputs in an effort to identify root causes for quality problems. Typically cause and effect analysis in Six Sigma is done by creating a diagram showing on the most simplistic level what inputs affect the output of a process and give a visual relationship of how they are related. However, by creating basic spreadsheet models in conjunction with the cause and effect diagram one can get a far better picture of the relationships between the inputs of a process and the outputs. The diagram below shows the business modeling cycle for cause and effect analysis. By replicating the real life process in spreadsheet model inputs can be manipulated repeatedly to determine their impact on the process output. Once “problem” input(s) is identified, it can be corrected. If for some reason the input has fundamentally changed (For example, a specific input is no longer available on the market and a replacement is being used), other inputs can be adapted in the modeling setting to determine how to best get the process back into a state of control.
Regression
Regression analysis is also done in the Analyze step of the DMAIC roadmap. Regression is another, slightly more sophisticated, manner of determining the relationships between one or more input variables and an output variable. Unlike basic spreadsheet modeling though, regression provides the following: a specific numeric relationship for each input variable with the output variable, the degree of variability that those relationships account for, and a goodness of fit measure. A regression equation can also be used to predict outcomes for the output variable. Below is a diagram of regression analysis:
Simulation
Simulation, often known as Monte Carlo Simulation can be used during the Improve phase of Six Sigma as a way of factoring uncertain data into a spreadsheet model to give a more realistic view of potential outcomes. One of the more notable examples that students usually see of Monte Carlo Simulation is when calculating the final balance of a 401K retirement account. If the model is built in a deterministic manner all inputs are held constant. However, since the behavior of a 401K account is stochastic in nature key inputs in this model, such as rate of return, are not constant. Monte Carlo Simulation allows the variability of the input “rate of return” to be largely accounted for. For example, one might use 8% as an annual rate of return when building a deterministic 401K model. However, a more accurate way of looking at it may be as a variable whose values can be described in a normal distribution with a mean of 8% and a standard deviation of 5%. Monte Carlo Simulation will create a random variable that will fit the distribution and will run the model repeatedly thousands of times. When complete, the simulation will provide the user with a report showing a distribution of final outcomes for the model, in this case a 401K account. In a deterministic model the outcome may be $1,000,000. In the simulation the outcomes may have a mean of $750,000 with a range from $100,000 to $3,000,000 distributed as the normal curve that was used on the variable “rate of return”. Ultimately, the greatest asset of the Monte Carlo Simulations is that they can give you, with some accuracy, the worst case scenario.
This type of simulation can be highly useful in Six Sigma. A business process is ultimately very similar to the 401K example described above. Some, if not many, of the inputs in a process are uncertain by nature and can not be controlled. Monte Carlo Simulations allow Six Sigma disciples to create realistic expectations of process improvements prior to actually implementing them which can be exceptionally useful.
Forecasting
Time Series Forecasting is a technique that is typically associated with the Control phase of Six Sigma where it is generally utilized to project whether the outputs of a process into the future. Since much of the data that is tracked and analyzed in Six Sigma is sequential in nature, it makes sense that forecasting can be utilized to project future process performance. Should a process begin deviate from its normal pattern, forecasting may be able to detect upcoming quality problems. There are a number of forecasting methods that can be used (moving average, exponential smoothing, etc.). It may require some degree of experience with the process in question to determine what method of forecasting is most accurate.
Optimization
“Optimization is the name for a family of tools designed to help solve managerial problems in which the decision maker must allocate scarce (or limited) resources among various activities to optimize a measurable goal.” (Quantitative Business Modeling: Meredith, Shafer, Turban).
As stated above, Optimization Modeling is the modeling managerial problems that are relatively complex, have one or more constraints, and attempt to achieve a specific purpose such as: maximizing profit, minimizing costs, etc. In typical business modeling today many optimization problems can be done in spreadsheet programs such as Microsoft Excel with its “Solver” feature. In a Six Sigma context, Optimization modeling is highly relevant in that when the objective is to monitor, control and increase the quality of a process it is often helpful, if not required to know what the potential capability of the process is.
Summary
As increased competition on business forces companies to be more competitive to survive it seems clear that quality of products and services will be one of the significant inputs in determining which organizations thrive and which ones fail on the future. Within this context it is also evident that the Six Sigma quality process will gain acceptance in both the corporate and academic worlds. Until then, the increased emphasis on business modeling techniques (Cause and Effect/Root Cause analysis, Regression, Monte Carlo Simulation, Forecasting and Optimization) is well-positioning students and managers to take the reigns of Six Sigma due to modelings abilities to give valuable insights into the business processes for which Six Sigma was designed to improve.
References
The primary reference for this paper, and course, is:
The Six Sigma Handbook, revised and expanded, A complete Guide for Green Belts, Black Belts, and Managers at All Levels by Thomas Pyzdek 2003