Tai Lieu Chat Luong Source: The Six Sigma Black Belt Handbook Part One: The Six Sigma Management System This first part of The Six Sigma Black Belt Handbook focuses on the extension of Six Sigma into a management system that encompasses all levels of an organization Motorola University consultants have found that while implementing Six Sigma through individual projects has produced significant results in many organizations, sustainable, breakthrough improvements are realized by those organizations whose leadership has embraced Six Sigma and incorporated it into their vision, strategies, and business objectives - in short, adopted Six Sigma as the system for managing their organizations The Six Sigma Management System enables a leadership team to align on their strategic objectives, establish their critical operational measures, and determine their organizational performance drivers and then use those to implement, drive, monitor, and sustain their Six Sigma effort The four chapters in this part of the book will: z z z z z Introduce the Six Sigma Management System, and distinguish it from the Six Sigma metric and Six Sigma methodology Explain the background (Chapter 1), principles, and elements of the Six Sigma Management System (Chapter 2) Describe the Six Sigma leadership modes (Chapter 3) Provide insights into Six Sigma leadership (Chapter 4) Illustrate key tools used to implement the Six Sigma Management System Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website The Six Sigma Management System Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Source: The Six Sigma Black Belt Handbook Chapter • Introduction to Six Sigma Six Sigma has been labeled as a metric, a methodology, and now, a management system While Green Belts, Black Belts, Master Black Belts, Champions and Sponsors have all had training on Six Sigma as a metric and as a methodology, few have had exposure to Six Sigma as an overall management system Reviewing the metric and the methodology will help create a context for beginning to understand Six Sigma as a management system Management System ySix Sigma drives strategy execution yLeadership sponsorship and review yMetrics driven governance process yEngagement across the organization ManagementSystem System Management Methodology Methodology Metric Metric Methodology yConsistent use of DMAIC model yTeam based problem solving yMeasurement-based process analysis, improvement, and control Metric yMeasure process variation Figure 1-1 Six Sigma as a Metric, Methodology, Management System Six Sigma as a Metric Sigma is the measurement used to assess process performance and the results of improvement efforts - a way to measure quality Businesses use sigma to measure quality because it is a standard that reflects the degree of control over any process to meet the standard of performance established for that process Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Introduction to Six Sigma Chapter One Sigma is a universal scale It is a scale like a yardstick measuring inches, a balance measuring ounces, or a thermometer measuring temperature Universal scales like temperature, weight, and length allow us to compare very dissimilar objects The sigma scale allows us to compare very different business processes in terms of the capability of the process to stay within the quality limits established for that process The Sigma scale measures Defects Per Million Opportunities (DPMO) Six Sigma equates to 3.4 defects per million opportunities The Sigma metric allows dissimilar processes to be compared in terms of the number of defects generated by the process in one million opportunities Sigma 0.02 3.4 233 6210 66810 DPMO (Defects Per Million Opportunities) Figure 1-2 Sigma Scale A process that operates at 4.6 Sigma is operating at 99.9% quality level That means: z z z z 4000 wrong medical prescriptions each year More than 3000 newborns being dropped by doctors/nurses each year long or short landings at American airports each day 400 lost letters per hour A process that operates at the Sigma level is operating at 99.9997% quality level At Sigma, these same processes would produce: Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Introduction to Six Sigma Introduction to Six Sigma z z z z 13 wrong drug prescriptions per year 10 newborns dropped by doctors/nurses each year long or short landings at U.S airports each year lost letter per hour Mikel J Harry, one of the developers of Six Sigma at Motorola, has estimated that the average company in the Western world is at a Sigma level, while Sigma is not uncommon in Japan.1 Dave Harrold, in Control Engineering cites benchmark sigma levels broken down by industry and type of process: z z z z z z z z IRS phone-in tax advise - 2.2 Restaurant bills, doctors prescription writing, and payroll processing - 2.9 Average company - 3.0 Airline baggage handling - 3.2 Best in class companies - 5.7 U.S Navy aircraft accidents - 5.7 Watch off by seconds in 31 years - Airline industry fatality rate - 6.2 Clearly, the value of sigma is its universal application as a measuring stick for organizational and process quality With sigma as the scale, measures of as-is process quality and standards for should-be process targets for quality improvement can be set and understood for any business process Six Sigma as a Methodology The Six Sigma methodology builds on the Six Sigma metric Six Sigma practitioners measure and assess process performance using DPMO and sigma They apply the rigorous DMAIC (Define, Measure, Analyze, Improve, Control) methodology to analyze processes in order to root out sources of unacceptable variation, and develop alternatives to eliminate or reduce errors and variation Once improvements are implemented, controls are put in place to ensure sustained results Using this DMAIC methodology has netted many organizations significant improvements in product and service quality and profitability over the last several years The Six Sigma methodology is not limited to DMAIC Other problem1 Harry, Mikel "Six Sigma: The Breakthrough Management Strategy Revolutionizing the World's Top Corporations." New York, N.Y Random House Publishers, 2000 Harrold, Dave, "Designing for Six Sigma Capability", Control Engineering, January 1, 1999 Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Introduction to Six Sigma Chapter One solving techniques and methodologies are often used within the DMAIC framework to expand the tool set available to Six Sigma project teams These include: z z z z z Theory of Inventive Problem Solving (TRIZ) Lean Ford 8Ds (Disciplines) Whys Is/Is Not Cause Analysis Utilizing the sigma metric and marrying this variety of approaches with the DMAIC methodology, the Six Sigma methodology becomes a powerful problem-solving and continuous improvement methodology Clearly, the use of a consistent set of metrics can greatly aid an organization in understanding and controlling their key processes So too, the various problem-solving methodologies significantly enhance an organization's ability to drive meaningful improvements and achieve solutions focused on root cause Unfortunately, the experience of Motorola University consultants has demonstrated that good metrics and disciplined methodology are not sufficient for organizations that desire breakthrough improvements and results that are sustainable over time In fact, conversations with organizational leaders who report dissatisfaction with the results of their Six Sigma efforts have shown their Six Sigma teams have sufficient knowledge and skill related to good use of metrics and methodology However, all too often, these teams have been applying the methodology to low level problems, and have been working with process metrics that don't link to the overall strategy of the organization It is this recurring theme that has driven Motorola University to develop the concept of Six Sigma as a management system, first introduced in the book "The New Six Sigma" Six Sigma as a Management System Six Sigma as a best practice is more than a set of metric-based problem solving and process improvement tools At the highest level, Six Sigma Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Introduction to Six Sigma Introduction to Six Sigma has been developed into a practical management system for continuous business improvement that focuses management and the organization on four key areas: z z z z understanding and managing customer requirements aligning key processes to achieve those requirements utilizing rigorous data analysis to understand and minimize variation in key processes driving rapid and sustainable improvement to the business processes As such, the Six Sigma Management System encompasses both the Six Sigma metric and the Six Sigma methodology It is when Six Sigma is implemented as a management system that organizations see the greatest impact These organizations are among those that have demonstrated that breakthrough improvements occur when senior leadership adopts Six Sigma as a management system paradigm z In 1999, ITT Industries implemented Value-Based Six Sigma (VBSS), the company's "overarching strategy for continuous improvement" (source: ITT Industries website) In the 2003 letter to the shareholders, the then-Chairman, President, and CEO, Louis J Guiliano, wrote, "VBSS gives us the tools and discipline we need to make fact-based decisions, to solve problems and to find solutions in a systematic and measurable way Now in its fourth year, the VBSS strategy is already making a huge difference to our customers The thrust of our VBSS projects has shifted from simple initial shortterm projects that drive out costs and waste, to more comprehensive projects that focus on making improvements that mean the most to our customers VBSS is generating many new ways of growing our business and increasing our capacity, as well as saving millions of dollars It is a key contributor to our robust cash flow performance As our VBSS project leaders grow in numbers and in expertise more than 10 percent of our 39,000 employees are now certified as Champions, Black Belts or Green Belts - they are increasingly focused on projects that are changing the way we business in profound and enduring ways Through the combined efforts of the Champions, Black Belts, Green Belts, and the teams they lead, we Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Introduction to Six Sigma Chapter One z z z z are seeing real progress in the quality of the products and services we are providing to customers We are shortening cycle times, reducing lead times, and eliminating excess inventories; we are meeting or exceeding on-time delivery commitments, and dramatically reducing defect rates Our customers have noticed these improvements, and we see this reflected in enhanced customer loyalty and market share position More than that, our employees gain satisfaction from working on these teams." General Electric started its quality focus in the 1980s with WorkOut Today, Six Sigma is providing the way "to meet our customers' needs and relentlessly look for new ways to exceed their expectations Work-Out® in the 1980s defined how we behave Today, Six Sigma is the way we work Six Sigma is a vision we strive toward and a philosophy that is part of our business culture It has changed the DNA of GE and has set the stage for making our customers feel Six Sigma." (source: General Electric website) Raytheon has used Six Sigma to cut billions of dollars in costs, improve cash flow and profits by millions, improve supplier and customer relationships,and build internal knowledge networks "Raytheon Six Sigma™ is the philosophy of Raytheon management, embedded within the fabric of our business organizations as the vehicle for increasing productivity, growing the business, and building a new culture Raytheon Six Sigma is the continuous process improvement effort designed to reduce costs." (source: Raytheon website) Honeywell views its Six Sigma initiative (called Six Sigma Plus) as the way to maintain its position with its customers as a premier company "At Honeywell, Six Sigma refers to our overall strategy to improve growth and productivity as well as a measurement of quality As a strategy, Six Sigma is a way for us to achieve performance breakthroughs." (source: Honeywell website) Valley Baptist Medical Center (Harlingen, TX) has incorporated Six Sigma Quality as one of its Seven Strategic Initiatives The hospital has been recognized with a number of national awards, including the "Top Performer" in the country for the overall quality of physician care in the emergency room (Award presented by independent marketing research firm Professional Research Consultants.) James G Springfield, President and CEO of Valley Baptist Health System, says they have made Six Sigma the company's system for operations (source: Valley Baptist Health System website) Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Introduction to Six Sigma Introduction to Six Sigma These and other organizations have discovered that successful practice of Six Sigma requires the adoption of a management system to strategically guide their Six Sigma programs The next chapter will explore the principles behind the Six Sigma Management System, as developed by Motorola University consultants to guide their clients to build strategic management systems and achieve breakthrough results Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments Measurement System Analysis in Non-Manufacturing Environments 557 z z z z tation People-dependent: processes are not fully automated, so training and employee turnover are consistently large issues Undocumented: processes evolve and mutate over time (as people move in and out), and often not have documented steps Significant external "noise": process output is often significantly affected by inputs that may not be easily controlled by the business, such as economic conditions, competitive actions, etc Impossible or difficult to get repeat measures: traditional measurement studies require getting repeat measures on the same unit or process event (e.g delivery of service) and this can be difficult or impossible for many non-manufacturing situations From all of these factors, an interesting dynamic emerges: there is a huge improvement opportunity in many non-manufacturing processes, but the execution of projects can be significantly more challenging than in manufacturing applications Definitions of roles, responsibilities and process steps are often quite nebulous This means that a significant degree of emphasis during a non-manufacturing Six Sigma project will lie in the Define and Measure phases - prior to taking data for baselining purposes Some clarity must be obtained up front, and the measurement system is a large part of that effort In fact, before the classical MSA study begins, there are two prerequisites that must be met to clarify the measurement system: Match the process output metric to the customer metric Create an operational definition for the metric(s) If a classical MSA study cannot be done because of the last characteristic listed above (inability to get repeat measures), these two prerequisites become in essence the validation of the measurement system and become even more important Prerequisite: Matching Metrics Six Sigma is an outside-in approach to process improvement: it begins with the customer's stated requirements of what is needed from the provider, which is then translated into an internal business process that fulfills the customer need When it comes to measurement systems, first Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments 558 Chapter Twenty-One and foremost is the requirement that the metric used to characterize the internal process output matches (or at least reflects) how the customer is judging the process output This is true even if the customer is internal the next downstream user of the process output under consideration Case Study: On Time Delivery A major manufacturer routinely delivered truckloads of their products to a large distributor in their network on a weekly basis The truck driver would call the warehouse supervisor at the distributor's location on the night before to confirm the scheduled delivery time for the next day On numerous occasions, the supervisor would request that the delivery be delayed until the following day due to labor issues or backed-up work in their warehouse The driver would enter the revised delivery date into the manufacturer's system and comply with the new request However, the distributor's purchasing agent who placed the order was not aware of the revised delivery date since it had been negotiated between the driver and warehouse supervisor Consequently, Purchasing judged performance of the supplier based on their compliance to original schedule, and the manufacturer judged performance based upon their compliance to the revised schedule The mismatched measurements were easy to rectify, once the problem was identified; however, up to this point the distributor, had deemed the supplier a poor performer, and the business relationship was quite strained Lesson: Be sure that the customer measures the process output the same way as the process owner Case Study: On Time Departure (Airline Industry) A classic case of mismatched measurement systems can be found in the airline industry Most air passengers, when asked, consider departure time as the time the wheels leave the runway Most airlines, however, measure departure time through an electronic system that monitors when the aircraft parking brake is released Therefore, from the airlines' perspective, the flight departs on time when the brake is released at the loading gate and the plane begins to taxi For those passengers who have experienced a flight that moves away from the loading gate and parks on a remote section of the tarmac for extended time periods, on time depar- Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments Measurement System Analysis in Non-Manufacturing Environments 559 ture would not be an accurate characterization of their experience In an example of unfortunate timing, a news story reported the airline industry's contention that departure and arrival performance had improved - even though customer satisfaction measurements regarding departure/arrival performance had significantly declined during this same time period Lesson: this example of mismatched metrics is difficult to correct by itself, since the aircraft electronic monitoring system is also used to calculate pilots' and flight attendants' pay However, a mitigation technique would be to monitor both operational and customer satisfaction metrics simultaneously, to avoid optimizing one metric and sub-optimizing another Prerequisite: Operational Definitions Once a metric (or metrics) that satisfies both the customer needs and the requirements of the process owner is identified, another pre-requisite to the MSA validation study is the creation of an operational definition for that metric An operational definition is a specific description of how the process output is to be measured, including by whom, with what measurement tool, and in what data capture format (tick sheets, direct computer entry, etc.) Customer-based specifications must be identified in the operational definition, since these become the standards by which the goodness or badness of the process output are judged The purpose of the operational definition is to ensure that everyone affiliated with the measurement process has a clear understanding of how the output will be measured and recorded, so that variability from people interfacing with the measurement system is minimized Example Operational Definition for On Time Delivery The units measured will be the time differential (in minutes) between the customerrequested schedule and the actual arrival of the product at the customer's location The system will calculate the differential, comparing the time at which the customer signs the electronic ticket in the hand-held unit carried by the driver to the 'requested delivery time' field on the order The target Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments 560 Chapter Twenty-One value is minutes differential; the customer will accept arrivals within +/- 15 minutes of the target as 'on time' MSA: Repeatability & Reproducibility (R &R) Studies The discipline dictated by the Six Sigma approach requires that measurement systems be validated before collecting process data Now that the metric has been defined and the operational definition written, the Measurement System Analysis can be performed to quantify the sources of variation in the measurement system Since the process data will be used to answer some very important questions such as 'How well is the process performing?' and 'What is driving the process output to vary?', it is critical to ensure that the data produced by the measurement system are "precise" (as consistent as possible.) MSA is, therefore, an important element needed before characterization of the y = f(x) equation can be done Just as total observed variation can be broken into two pieces (variation due to the measurement system and variation due to the process), measurement system variation can also be further broken down Two common sources of variation are called repeatability and reproducibility A study, known as an R & R study, is conducted to understand these sources These sources and example studies are described below However, the first decision before conducting the study is the type of data being used Total Observed Variation = ‘AR&R ‘AR&R Study’ Study’ Variation from the Measurement System + Variation from the Process ‘GR&R ‘GR&R Study’ Study’ Data Type Attribute (Discrete) Continuous Nature of Variation Nature of Variation Repeatability Reproducibility Repeatability Reproducibility Figure 21-4 Measurement System Sources of Variation Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments Measurement System Analysis in Non-Manufacturing Environments 561 Types of Data The first task is to identify the type of data being measured Data that are discrete can be measured as counts of occurrences, categories of results or pass/fail proportions Discrete data are often described as "classification" data, and are also known as process attributes Examples of discrete data in non-manufacturing processes include: z z z Number of damaged containers Customer satisfaction: fully satisfied vs neutral vs unsatisfied Error-free orders vs orders requiring re-work Data that are continuous can be measured by numerical values, typically in which a decimal could make sense Continuous data, by definition, can theoretically be divided into finer and finer increments of measurement, and so can be used to quantify the output Continuous data are also known as "variable" data Examples of continuous data in non-manufacturing processes include: z z z Cycle time needed to complete a task Revenue per square foot of retail floor space Costs per transaction From a process point of view, continuous data are always preferred over discrete data, because they are more efficient (fewer data points are needed to make statistically valid decisions) and they allow the degree of variability in the output to be quantified For example, it is much more valuable to know how long it actually took to resolve a customer complaint than simply noting whether it was late or not Measurement system analysis can be performed on processes using either data type - MSA for discrete or attribute data is known as 'AR&R', and MSA for variable data is known as 'GR&R' Sampling for MSA Studies When a measurement study is to be performed, careful consideration must be given to the sample being used for the study The measured elements of an MSA study should be representative of the full range of variation which could be seen from the process Thus, it is an engineered sample, Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments 562 Chapter Twenty-One containing much more variability for the number of observations gathered than what would typically be seen from an actual process sample It is very important to realize that the MSA study sample is NOT reflective of actual process performance, and cannot be used to characterize the process output Its sole intent is to test the measurement system over the full range of output which could occur - including out-of-specification (customer requirement) results Through the MSA, the precision, or consistency, of the measurement system can be quantified The MSA sample and the process baseline sample are completely independent of one another The size of the MSA sample depends upon the nature of the data being measured If the data are discrete (attribute), then more observations will be needed to characterize the measurement system than for continuous (variable) data For example, a continuous study may have a sample size of 30 to 40 while an attribute study may have 100 or more observations Types of Measurement Error Regardless of the nature of the data, the variation (error) from the measurement system can be further subdivided into two sources, known as R & R: I Repeatability error - when one operator is using the same gauge to measure the same element multiple times and obtains different results Repeatability error is often called equipment variation because it is usually caused by a problem with the measuring device itself Remedy: modify the gauge, or improve the environmental conditions for reading the gauge (lighting, location, etc.) True Value Gauge operator #1 Gauge operator #2 Gauge operator #3 Figure 21-5 Measurement System Variation Due to Repeatability Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments Measurement System Analysis in Non-Manufacturing Environments 563 II Reproducibility error - when multiple operators are using the same gauge to measure the same element multiple times, and the results are different Reproducibility error is often called appraiser variation, because it is usually caused by inconsistencies in measurement methods used by the operators Remedy: train the operators, using the detailed operational definition for the measurement system True Value Gauge operator #1 Gauge operator #2 Gauge operator #3 Figure 21-6 Measurement System Variation Due to Reproducibility Attribute R & R (AR&R) Attribute gauge studies are typically used when the measurement result is binary, such as defect/no defect or successful/unsuccessful, although rating scales can also be validated with this method Multiple measurement system operators are chosen to measure a sample set two or more times In this way, both repeatability (variation within the operator) and reproducibility (variation between the operators) can be quantified In an attribute study, a standard can be used for comparison with the results from the measurement system operators The standard is the 'truth' and any discrepancy from truth due to the measurement system is considered an error (or defect) For attribute measurements, a rule of thumb is that the agreement with truth should be 99% or better for the measurement system to be considered "good." Agreement of 95% or more is often considered marginal; if the agreement is less than 95%, the measurement system should be considered unacceptable AR&R studies can be done using statistical software packages which provide graphical output and other summary information; however, they are often done by hand due to the straightforward nature of the calculations Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments 564 Chapter Twenty-One Example An attribute measurement instrument with notoriously bad R & R is the "reason code" form Whether it is reason codes for customer complaints, product returns or help requests, this type of measurement instrument is rarely validated and usually wildly variable Process associates are asked to categorize events into pre-defined codes, and the meaning of the codes is seldom clear In one case, a company used reason codes to categorize the reason for product returns Customer service representatives were required to fill out an on-line form and pick from a pre-determined, numerical code that represented a return reason When the Six Sigma project reached the MSA phase, they investigated the "goodness" of this measurement system and found that of the 27 codes on the list, only were ever used They also discovered that there was a common return reason that was apparently not an issue when the reason codes were developed years before, so the representatives just picked a category to label these returns Through the Six Sigma project, reason codes were streamlined to seven categories, based upon input from the associates In addition, concise operational definitions were provided on-line to help them choose the appropriate categories The team then conducted an AR&R study They consulted an expert to determine the reason for the product returns This was considered the "truth." Three service representatives were then asked to determine the reason code for 20 different product returns Then, on a different occasion, the three representatives were asked to determine the return codes on the same 20 returns It was a blind study; the representatives did not know they were looking at the same product returns both times The Six Sigma team was then able to determine repeatability (did the representatives assign each product the same reason code?), reproducibility (were the same reason codes assigned from representative to representative?) and did they assign the right reason code per the expert The results of the study were 100% across all categories The team felt confident in their measurement system Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments Measurement System Analysis in Non-Manufacturing Environments 565 Gauge R & R (GR&R) GR&R studies are performed on continuous output data As in the AR&R study, multiple measurement system operators are used to measure sample data multiple times in order to quantify repeatability and reproducibility Since the data are continuous though, the variation in the measurement system is characterized by standard deviation calculations that are best left to statistical software packages GR&R results are usually expressed as the % of total variability coming from the gauge, or measurement system Since variability from the measurement system should be small relative to the total variability, a typical criterion for acceptability is 10% or less If the measurement system contributes between 10% and 30% of the total variability, it may be considered conditionally acceptable There are three approaches that can be used to calculate the percentage of variability attributable to the measurement system: % of Total Variability (%R&R) - used to determine if the measurement system variability is small enough to understand process variability % of Tolerance (%P/T) - used when the customer specification range is given (Upper Specification Limit - Lower Specification Limit) The criterion for acceptability for GR&R as a % of tolerance is the same as for % of Study: 10% or less is acceptable, and 10-30% is marginally acceptable % of Contribution - a special case of the variability calculation, which uses the statistical variation rather than the standard deviation (which allows the contributions to be arithmetically added) The criterion for acceptability is: < 2% variation is acceptable, 2-9% is conditionally acceptable Statistical software packages often provide the added benefit of quantifying the discrimination of the continuous measurement system This can often be seen as the number of distinct categories in the analysis output When the number of distinct categories is greater than five, the measurement system can discern over five groups within the data range and is often considered to have acceptable discrimination Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com) Copyright © 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website Measurement System Analysis in Non-Manufacturing Environments 566 Chapter Twenty-One A summary of typical acceptance criteria is shown in the chart below: AR&R GR%R %Total Variability GR&R % Tolerance GR&R %Contribution # of Distinct Categories Acceptable 99%+ Agreement < 10% < 10% < 2% > 10 95% -99% Agreement 10% - 30 % 10% - 30 % 2% - 9% - 10 Marginal Unacceptable < 95 % Agreement > 30% > 30% > 9%