Six Sigma as an instrument to improve quality.
Many companies use the Six Sigma methodology as a mandatory methodology in their plants, one manager cites three reasons:
- Six Sigma becomes imperative when evaluating and improving process capability. Six Sigma is a means of reducing the complexity of product and process designs while increasing their reliability. Six Sigma is a staging to combat which is often attributed to "bad luck"; This staging is valid not only in the workshop but anywhere in the organization.
THE VISION OF SOME COMPANIES THAT DECIDE FOR SIX SIGMA:
"Our vision is to become world-class manufacturers, making world-class products." This type of manufacturer has the best opportunity to develop its business and survive, due to its higher profitability. It has a Six Follow level in all its processes and produces services and products of equal quality.
This vision can and should be quantified. Technically, Six Sigma quality equates to a quality level of less than 0.000003 defects per opportunity (3 defects per million opportunities). Unfortunately, there is no immediate, simple and easy rule to achieve such a level of quality. Six Sigma is a methodology that will help achieve this goal.
TWO ASPECTS ON WHICH THE SIX SIGMA METHODOLOGY TURNS:
- The quality manufacturer is also the manufacturer who is capable of producing at low cost. It is less expensive to make good first time than to spend money on adjustments and corrections. Quality can be quantified, and moreover quality has to be quantified. Quality diagnosis and monitoring is a long-term commitment to quality. In the short term Six Sigma is based on measurements rather than past experiences, judgments and beliefs. If you cannot measure you do not know where you are, if you do not know where you are, you are at the mercy of chance.
SIX SIGMA HAS TWO DRIVERS:
- The first is the reduction of costs caused by poor quality. Six Sigma is oriented to concrete results, benefits in the income statement, which are long-term benefits; But with the Six Sigma methodology, short-term results are emphasized as we work for long-term results. It is about reducing costs by improving quality, specifically we fight against mistakes, failures, not only those that are detected in the test bench, but including, for example, accepting too wide margins with the consequent cost increase, defects that cause repetition of work, administrative defects, unnecessary work etc. All defects are, of course, non-value-added work.We are convinced that the level of defects currently ranges between 15 and 20% of turnover, this is a good indication of the potential savings that we have in our hands.
- The second energizer is the breakdown of complacency. Six Sigma drives towards a sense of urgency and need to rise to world standard levels. Six Sigma is an essential instrument to create such awareness, with the ultimate purpose of improving profitability and competitiveness.
THE TWO CORNERSTONES OF SIX SIGMA:
- Systematic methodology.
Six Sigma is a systematic methodology to proactively reduce costs, concentrating on improving processes rather than reacting by correcting failures once they occur. Six Sigma is based on measurements rather than on past experiences, therefore it is a methodology applicable to a wide field of business activities. Conceptually, Six Sigma requires that each problem be solved from a relationship of the form:
Y = f (x, y,… z); where Y is the dependent variable; x, y,.. z: independent variables.
Six Sigma provides the teachings to find such an equation.
Any quality improvement work requires systematic actions and the help of statistics, be they long or short term.
- "Black Belts"
They are the agents of change, trained in the use of statistical methods. Some call them "infiltrated commandos", they are well trained and are assigned ambitious objectives.
These change agents know their own organization, products, and people well and are taught the Six Sigma methodology.
WHAT IS SIX SIGMA IN A FEW WORDS:
Based on the use of statistics as established by many gurus of business management. Based on the concepts of Shewhart, Deming, Juran and Taguchi and developed by Mikel Harry.
In the short term, it provides quick solutions to simple or repetitive problems; In the long term, it provides a diagnostic methodology, robust design, tolerance establishments, while providing a simple means of communication and goal setting.
It provides improvement tools: design of experiments, regression analysis, tolerances, robust design and other systematic methods to reduce variance.
No less important is the organizational structure that Six Sigma uses with the Black Belts, specially trained employees to move the organization towards the achievement of the established objectives.
The iterative process of these change agents is represented by the scheme:
Interactive Six Sigma Process
A BRIEF SUMMARY OF HOW IT IS IMPLEMENTED:
The overall objective is long-term, but in organizations we are forced to present short-term results as well, while working in the long term; Six Sigma allows you to simultaneously meet both objectives at the same time.
Process improvement:
Measuring is necessary but not enough, in the long run, to stimulate people to make changes. The analysis of the defects per million and their corresponding sigma values will give an orientation about which processes have the greatest potential for improvement; once we have detected where the improvement potentials lie, we have to put into practice the instruments and capacities to improve these processes
Product improvement:
Six Sigma allows establishing a system of continuous improvement of products; But with Six Sigma we can go much further, as it is an excellent support for robust product design and for dynamic product simplification. Design engineers to develop their robust and simplified products need to know the capacity of the processes, with this they can reduce manufacturing costs while designing products with less variability in their manufacturing process.
A systematic for problem solving:
When a problem occurs in a process, the normal thing is that first we go to our past experience to find solutions or look for causes, then we go to analysis procedures such as Ishikawa, Pareto, etc. These methods do not always lead to optimal solutions.
Six Sigma provides us with a more precise and conclusive systematics with the application of the design of experiments, the appropriate use of regression analysis, SPC and many other statistical methods. The systematic measurement and resolution of problems using proven statistical techniques together with an adequate organization and training of people is what together guarantee the successes of Six Sigma.
ACTIVE PARTICIPATION OF THE MANAGEMENT AND CLIENTS:
Much has been said about the importance of the role of management in any activity of a company, this is so obvious that it goes without saying; but many improvement initiatives, slow or fast, long-term or short-term, with many or few resources,… have failed due to lack of… I don't know what direction. When we attend quality meetings, the importance of management participation is always cited…. Six Sigma facilitates this communication with management by teaching how to quantify the costs and benefits of improvement initiatives.
It is also important to inform the client of our initiatives or at least to exploit the advantages acquired, obtaining the client's evaluation of the results.
TRAINING AND TRAINING:
There is no doubt that an essential aspect of an initiative like Six Sigma is based on adequate training for all the people involved in the process.
A company that has decided to incorporate Six Sigma into its development processes must have all its employees adequately informed and provide them with the necessary training based on their level of participation. The attached table summarizes the type of training provided in a large company and to whom.
| Six Sigma for managers | "Whites" | "Yellows" | "Green" | "Black Belt" | Engineering | |
| Main content | BB requirements | Very basic | Basic | Normal | Full | Advanced |
| Duration | 2 days | 1 day | 2 days | 4 days | 13 days | 15 days |
| CEOs,
plant, division.. |
X | |||||
| Directors of engineering, production, quality | X | |||||
| Other directors | X | OR | ||||
| Design and production engineers | X | OR | OR | |||
| Quality Engineers | X | OR | ||||
| Procurement management | X | OR | ||||
| Accounting, personnel, planning, | X | OR | ||||
| Testing, marketing | X |
QUALITY MEASUREMENT:
Quality is quantifiable and must be quantified. To quantify quality, you have to express quality in numbers. Quantifying quality in numbers and carrying out concrete actions to improve it ensures that quality will improve.
This is true for the quality of manufactured products as well as for the quality of business processes of any kind. For the continuous improvement of designs, it is essential that design engineers continuously know how the quality of the designed product evolves.
Companies that adopt the Six Sigma system require that in their manufacturing processes they do measure a minimum of parameters to determine with them the sigma value of the plant. When a Six Sigma measurement system is established in a plant, the Sigma Value of the entire plant is obtained, of each of the main processes, of the individual products, of the most representative suppliers, of the different sections of the plant., etc. And all these values can be determined for each of the desired time intervals.
There is no doubt that this information is invaluable for selecting priority improvement initiatives. You can even advance us information about possible future failures
OBJECTIVES ESTABLISHMENT:
Difficult but achievable goals are the ones that give the best results. It should not be expected that in a year we can go from a quality level of 3 to 4 Sigma Value.
Some processes should improve by this order of magnitude, but the overall value will probably take longer to reach.
TANGIBLE RESULTS:
Acceptance of some old principles:
Six sigma is nourished by proven quality principles such as: Juran's non-quality costs, Deming's methodology of “measure, analyze, improve and control”, the robust design of Taguchi, SPC, design of experiments, regression analysis, etc…
All these methodology supported by a system of measuring and presenting the measurement values, the Black Belt infrastructure and not least the necessary enthusiasm is what has allowed us to reap results with Six Sigma.
Six Sigma Measurement:
INTRODUCTION:
Quality is quantifiable, and it must be quantified. To measure quality, you have to express quality in figures and act according to the measured values. These two simple principles give rise to a methodology for continuous quality improvement.
This is true for the manufacture of products, for the provision of services and for the design of new products.
To establish a Six Sigma systematic, it is necessary to establish certain measurement parameters whose set will provide the sigma value of the processes, products, suppliers, workshops, departments, etc. It is very useful to express these values on a graph as a function of time.
The following will detail how sigma values are calculated from measurements of defects per million. Before starting the measurement system, all employees must be informed of the system and the plans and goals. This is especially important for process owners.
HOW WE QUANTIFY QUALITY IN FIGURES:
Each of the parameters that we are going to measure may involve one or more opportunities, so we express the defects per million opportunities in the following way:
Dpmo = x 1,000,000
Defect in the broad sense of probability of defect.
When several dpmo are counted you can calculate a global dpmo and this dpmo can be converted to sigma value
A LITTLE BASIC STATISTICS BEFORE YOU START:
Most of the production processes follow a normal distribution, with a frequency distribution following the Gaussian bell and with a probability that some values are outside the upper and lower limits; this probability is what we understand by "probability of defect"
Our process will be all the more reliable the more focused on the limits and the narrower and higher the hood. A flattened and off-center hood is a consequence of high probability of defects. Graphically the area of the
Gaussian bell that lies outside the area marked by the upper and lower limits is precisely the defect probability.
In the Normal distribution tables we will find precisely a relationship between this area and the distance Z defined as:
Where Z is the "sigma value"; X the mean and s the standard deviation.
The relationship between the "probability of defect" (area of the Gaussian curve that falls outside one of the upper or lower limits) and Z (distance from the mean value to this limit) for a
Normal distribution is found in the corresponding tables.
Normal and Six Sigma Distribution
Two limits are often given in the LS and LI specifications, therefore we have to consider both areas that are outside the curve.
The total probability of defect will be the sum of the probability of exceeding the upper limit plus that of exceeding the lower limit. In this case, to calculate the Z value, both probabilities are added.
The Z number is what in Six Sigma we call “sigma value” when we only have an upper limit, as is the case in the figure. When there is an upper and a lower limit, we calculate an equivalent sigma number by adding the defect probabilities of both extremes and with this value we search for the Z value.
Z value and Six Sigma
A distinction must be made between the probability of a defect and the sigma value "short" and "long".
The short sigma value is equal to the long sigma plus “sigma shift”.
The “sigma shift” value, by convention, which is taken in the absence of other data:
"Sigma shift" = 1.5
By convention, when we talk about probability of defect we talk about the long term and when we talk about the sigma value we talk about the short term.
When we talk about the short term, only random phenomena influence.
When we speak in the long term, random phenomena and disturbances intervene (in telecommunications we speak of noise and signal).
In Six Sigma what we are interested in eliminating in the first place is noise, the signal is more controllable, but it should not be touched without first controlling the noise.
WHAT SHOULD WE MEASURE:
The answer will come from experience and knowledge of the specific process we are studying.
Some general rules for selecting the measurement variables to be measured:
- Important variables for the business (product characteristics, labor and material content including scrap and cycle time…). What is profitable to improve and those variables that you want to modify with improvement programs. Those variables that have to guarantee that Programs in place deliver results - those needed to ensure that improvements are lasting.
