《BU BreakersSix Sigma v5.doc》由会员分享，可在线阅读，更多相关《BU BreakersSix Sigma v5.doc（79页珍藏版）》请在三一办公上搜索。
1、IntroductionSix Sigma was created in the 1980s by Motorola. The background was that the head of Motorola saw that many Japanese companies were successful due to their quality work. With this in mind Six Sigma was created focusing on reducing unwanted variation. Sigma is the Greek letter used to defi
2、ne the standard deviation of a population, in other words, measuring the variability or spread of data. Six Sigma is also a measure of variability. It is a name given to indicate how much of the data falls within the customers requirements. The goal of Six Sigma is to increase profits by eliminating
3、 variability, defects and waste. To achieve Six Sigma, a process must not produce more than 3.4 defects per million opportunities (DPMO). A Six Sigma defect is defined as anything outside of customer specifications.The Six Sigma value is directly related to the Cost of Poor Quality. At 2 sigma 69.14
4、6 % of all products meet customer requirements, at 4 sigma the number is 99.379 %. At 6 sigma the business is as perfect as it can get with 99.99966 % of the products meeting customer requirements. The aim of Six Sigma is to reduce the number of products that do not meet the customers requirements.
5、One way of doing this is to move the mean completely within the customer requirements, move the curve from A to B. Another way of increasing customer satisfaction is to reduce the entire variation of the process. In this way customer requirements are met more consistently. The basic idea of Six Sigm
6、a is the implementation of a strategy that focuses on process improvement and variation reduction. This should be done by Six Sigma improvement projects. There are two Six Sigma sub-methodologies: DMAIC and DMEDI. The Six Sigma DMAIC process (define, measure, analyze, improve, control) is an improve
7、ment system for existing processes. It focuses on real problems that are directly related to financial results. Within DMAIC there are many tools and techniques, including thorough statistical methods. With this methodology you will have results in 4-6 months and the improvement is sustained over th
8、e long term. DMAIC also spreads improvement throughout the organization. The Six Sigma DMEDI process (define, measure, explore, develop, implement) is an improvement system used to develop new processes.Define The first step in the DMAIC method is to define opportunities. In the define process the m
9、ain objectives are to identify the improvement opportunity, develop the business processes, define critical customer requirements and prepare the group to become an effective team. The define phase consists of 6 main activities: 1. Identify / Validate business opportunity2. Validate / Develop projec
10、t charter 3. Document and analyze processes4. Identify quick win and refine processes5. Determine customer requirements6. Build effective teamsThe key deliverables from the define phase are: Team charter Action plan Process maps Quick win opportunities Critical customer requirements Prepared teamThe
11、 project charter is a fundamental and critical component for a Six Sigma Team. The project charter consists of 6 components which can be seen in the picture below. In appendix A you can find a project charter for ABB Sace for the acquirement of new tools.After a draft of a project charter has been m
12、ade it can be evaluated using the SMART method. This is a checklist to insure that the charter is effective and thorough. All goals must be specific具体的, measurable可测量的, attainable可达到的, relevant相关的, and time bound有时间节点的. Process mapping is used to document and analyze processes. The two most common p
13、rocess mapping tools are the top-down chart and the functional deployment process map. With the top-down chart you document a core process and its sub processes. With the functional deployment process map功能展开过程图 you document sub processes, individual steps and decisions, and also who is responsible
14、for them, seen in the picture below. In appendix B there is an example of a process map done at ABB Sace. To be able to improve processes it is important to first understand customer requests. To determine customer requests companies often use marketing surveys together with interviews performed by
15、customer service. The customer feedback must be translated into critical customer requirements. The critical customer requirements must be specific, precise and measurable. The customer is not always an external customer. In many cases the customer is internal and the critical requirements can for e
16、xample be specifications or tolerances. How the project team works together is essential for the success of a project. There are several tools and techniques which enable teams to perform effectively but maybe the most important thing is to understand that it takes time for a group to transform into
17、 a team.Measure The objective of the measure phase is to identify measures that are necessary to evaluate if customer requirement are being met. You must also decide how to collect the data for measuring and how to calculate Six Sigma. The measure phase consists of the following activities:1. Determ
18、ine what to measure2. Measurement management3. Understand variation4. Determine Sigma performance5. Managing the measurement systemThe key deliverables from the measure phase are: Input, process and output indicators Operational definitions Data collection formats and plans Baseline performance; sig
19、ma, cost, time, other, team atmosphereDetermine what to measureThe first step is to understand the critical customer requirements. Good output indicators relate directly to critical customer requirements. This is the most essential thing to take into consideration or else the indicator is useless. A
20、 good indicator balances the ease of obtaining data, available time, and the cost of acquiring the data. One critical customer requirement can have several output indicators or vice versa反之亦然. After creating output indicators the next step is to develop process indicators. With these you can identif
21、y variations before a defect occurs. This will minimize customer dissatisfaction. If however the output is defective the organization will have the data to identify the root cause of the problem. The indicators should make it possible to analyze cause-and-effect relationships within the process and
22、lead the team to the root cause of problems. Input indicators measure the consistency of the inputs to the process. They should measure the critical requirements on the suppliers of the process. A relationship matrix can be uses to show the relations between output performance measures and input and
23、 process measures. The strength of the relationship is based on how likely changes in the input or process indicators will cause changes in the output indicator. Measure indicators must be chosen wisely. Choose as few indicators as possible since many measurements will lead to a higher cost for coll
24、ection and analysis. On the other hand too few indicators will limit the ability to find defects. Good indicators are: Related to critical customer requirements Simple to understand Help make appropriate actions Provide insight to cause-and-effect relationships Relevant to the process Driven by busi
25、ness goals Reveal trends Easy to collect Clearly defined Lead to process improvement Statistically valid RepeatableMeasurement managementThe second part of the measure phase is to manage measurements. This can be done with a data collection methodology. The methodology begins with the identification
26、 of the measurements which was covered in the last section. After identification there are four steps:The operational definitions answer the questions what, how, how much, and who. What to measure is answered by giving a description of the criteria used for the measure. The methodology to collect th
27、e data answers how, the amount of data; how much, and the responsibility; who. Operational definitions provide everyone with the same meaning of the measurement. When developing a measurement plan one must ensure that the data is meaningful, valid, and collected at the same time. The measurement pla
28、n establishes a method for collecting the data. This can be done by creating a data collection plan template, see example below. The measurement plan should also tell how the data will be used and how it will be displayed.Data is often classified into to types; continuous连续的 and discrete离散的. Classif
29、ication is important for deciding analysis tools, sample size, type of control chart, and method of calculating sigma. 变量 特征The third step is collecting the data. When collecting data you must always follow the plan and note any movement away from it. It is also important to be consistent in the col
30、lecting, to avoid unfairness and not analyze the data as it is being collected. There are many methods to collect initial data. The two most common are checksheets with which you log data with a simple mark, or time stamps that record the time each activity begins and ends. Sampling data lets you dr
31、aw conclusions about the entire population while saving time and money. There are two main sampling situations; systematic process sampling and random sampling. 描述的The last step in the data collection methodology is to display and evaluate the data. When displaying the data you can find errors and v
32、ariations. There are several different ways of displaying the data. In evaluating the collection methods you must determine if the methods have provided consistent and representative data. You must also evaluate if the method will give you similar results if the collection is repeated and if you hav
33、e received the information needed. Understand variation In all processes there will exist variation. Variation means that the process doesnt produce the exact same result every time. It is important to measure and understand variation to help identify the level of performance and what to change in o
34、rder to reduce variability and defects. The goal of Six Sigma business improvement is to centre a process well within the customer requirements by reducing the variation. To determine if a goal is achieved one must know the customer requirements and the mean and standard deviation of the process. By
35、 root cause analysis the cause of variation can be identified and permanently deleted. Control charts can be used to monitor a process and manage variation. They provide an easy visual indicator of the process performance. With the control chart you can determine if a process is dealing with common
36、or special cause variation. Common cause variation is random and expected in the system while special cause variation is unpredictable and irregular. These variations are often referred to as ”in control” or “out of control”. Determine sigma performanceSigma performance is measured by determining th
37、e percentage of a process that falls within the critical customer requirements. The four steps in the calculation are:1. Determine critical customer requirements and identify them as output variables.2. Determine the type of data to be collected for each performance measure that will be used to calc
38、ulate sigma.3. Determine the nature of the collected data (short or long-term). This since output measures can vary over time. 4. Calculate sigma performance. The method to be used for calculating sigma performance depends on if the data is discrete or continuous. One way of determining process perf
39、ormance is with the process yield收益. The process yield is the part of process outputs that meet the critical customer requirements. With the Z value method the critical customer requirements level is used together with the mean and the standard deviation to calculate a Z value. The Z value is then t
40、ransformed into a yield with a Z table giving the Sigma performance level of the process. In the example below the critical customer requirement is 10 minutes. The mean is 6.8 and the standard deviation is 1,2. This gives the Z value 2.67 which in a Z table is transformed into 99.6 %; in other words
41、, the probability of the process being without a defect is 99.6 %.This method can be used even with 2-sided limits, upper and lower critical customer requirements. For discrete data the DPMO (defects per million opportunities) method can be used. With this the number of defects are counted and compa
42、red with the total production. The relationship between DPMO and process sigma is:Process yield = 1 (DPU/O) where DPU = D/NOtherwise a conversation table can be used to transform DPMO and sigma performance. Most measurements have a target value and acceptable limits of variation around the target. H
43、ow well the expected values fall within these tolerances is a measure of the process capability. Process capability is calculated from the mean (), standard deviation () and the upper and lower tolerances. There are two different measurements, Cp and Cpk. Cp only measures the dispersion 散步 of the va
44、lues whereas而 Cpk also measures the shift from the target, how well the process is centred on the target. For both Cp and Cpk the value should be as high as possible. A goal is often to have the measurements be larger than 1.33. If they are less than 1 the variability of the process is greater than
45、the limits. Managing the measurement systemMeasuring systems must be analyzed to make sure that there are not any variations in the system. Variations in the measuring system can make process seem worse than they are. This will result in a waste of resources trying to improve a process when the actu
46、al problem is the variation in the measuring system. The total variation that is observed is the sum of the process variability and the measurement variability. AnalyseIn the analyze opportunity phase you analyze the opportunity to identify a problem. The goal is also to identify the root causes of
47、the problem and eliminate them. The main activities in this phase are:1. Process stratification层别 and analysis2. Determine potential root causes3. Validate root causes4. Experimental design5. Manage creativityThe key deliverables from the analyze phase are: Data analysis Process map Validated root causes Problem statementProcess stratification层别 and analysisThe goal of the Six Sigma process improvement is to permanently eliminate root causes to defects that cause customer dissatisfaction. Root cause analysis finds the underlying source of a defect s
