Process Improvement shared by Ahsan Saleem. Email massfrompak@gmail.com

1. $ix $igma Remarkable Results and Rave Reviews Is it really more fun than a Root Canal?

8. Cost of Poor Quality (% of Revenues) versus  Level GE: $8 - $12 Billion/Yr 25% 15% 10% 5% 2%

16. Motorola’s Assumption the Process Mean Can Shift by as Much as 1.5 Standard Deviations Chapter 4: Six Sigma for Process and Quality Improvement

17. Six Sigma Long Term Shift & Drift LSL USL 1.5  1.5  Nominal Average Average

19. What Is Six Sigma and the 1.5 shift? The Original Concepts And Theories To quote a Motorola hand out from about 1987 ... 'The performance of a product is determined by how much margin exists between the design requirement of its characteristics (and those of its parts/steps), and the actual value of those characteristics. These characteristics are produced by processes in the factory, and at the suppliers. Each process attempts to reproduce its characteristics identically from unit to unit, but within each process some variation occurs. For more processes, such as those which use real time feedback to control outcome, the variation is quite small, and for others it may be quite large. A variation of the process is measured in Std. Dev, (Sigma) from the Mean . The normal variation, defined as process width, is +/-3 Sigma about the mean. Approximately 2700 parts per million parts/steps will fall outside the normal variation of +/- 3 Sigma. ( see chart #2 ) This, by itself, does not appear disconcerting. However, when we build a product containing 1200 parts/steps, we can expect 3.24 defects per unit (1200 x .0027), on average. This would result in a rolled yield of less than 4%, which means fewer than 4 units out of every 100 would go through the entire manufacturing process without a defect. ( see chart #3 )Thus, we can see that for a product to be built virtually defect-free, it must be designed to accept characteristics which are significantly more than +/- 3 sigma away from the mean. It can be shown that a design which can accept TWICE THE NORMAL VARIATION of the process, or +/- 6 sigma, can be expected to have no more than 3.4 parts per million defective for each characteristic, even if the process mean were to shift by as much as +/- 1.5 sigma ( see chart #2 ) In the same case of a product containing 1200 parts/steps, we would now expect only only 0.0041 defects per unit (1200 x 0.0000034). This would mean that 996 units out of 1000 would go through the entire manufacturing process without a defect. To quantify this, Capability Index (Cp) is used; where: A design specification width of +/- 6 Sigma and a process width of +/- 3 Sigma yields a Cp of 12/6 = 2. However, as shown in ( see chart #4 ), the process mean can shift. When the process mean is shifted with respect to design mean, the Capability Index is adjusted with a factor k, and becomes Cpk. Cpk = Cp(1-k), where: K factor= Process Shift Design Specification Width The k factor for a +/- 6 Sigma design with a 1.5 Sigma process shift ... 1.5/6 = 0.25 and the Cpk = 2(1- 0.25)=1.5 Cp= Design specification Width Process Width

20. Six Sigma is not a panacea: Motorola popularized the benefits of having six standard deviations between the process' nominal and each specification limit. If the process remains centered on the nominal, it has a Cpk (process capability index) of 2.0. This means a one part per billion nonconformance rate in each tail (above the upper specification and below the lower specification). Motorola allowed for a 1.5-sigma process shift-- which any decent statistical process control chart should detect very quickly, by the way-- which would make Cpk 1.5, and the nonconformance rate 3.4 ppm. Again, there is nothing wrong with this, but there is nothing new about it either. Walter Shewhart and his contemporaries identified the issue of process capability decades ago, and Henry Ford was seeking ever-more-precise manufacturing equipment during the 1910s and 1920s! Ford, in fact, had to hire Carl Johannson (of the famous Jo blocks, or gage blocks) to get the precision measurement systems necessary to support his operation. During the 1920s, Ford boasted of owning Jo blocks with 1-microinch (25.4 nanometer) steps; these dimensions now come to mind in microelectronics manufacturing. In summary, "Variation is the enemy" (we've known that for decades). Design for manufacture (DFM) includes consideration of the variation from the tools that will actually have to make the product. "Design for Six Sigma" is basically DFM, which also was a cornerstone of Henry Ford's manufacturing methods. A Six Sigma process with a 1.5 sigma shift in the process mean. Cpk=1.5 and the nonconformance rate is 3.4 parts per million. Six Sigma process capability with the process centered on its nominal (100). Cpk=2.0 and the nonconformance rate is 2 parts per billion.

23. The DMAIC Model The Problem Project Scope The Customer Metrics Current Process map Deliverables Process KIV Process KOV Collect Data Feed to SPC Variation Key Metrics GR&R (Validation) Process Capability Yield Sigma Level Pareto Charts Multilevel Pareto Root Cause Control Charts Fishbone D. FMEA Process Maps Major Obstacles Needed Resources Multi-Vari Optimization DOE PM Train Operators Visual Aids Gauges & Fixtures Control Plans Monitoring Standardization Documentation Audits & Reports Prevention Mistake Proof Sustain the Gain Process Input: x Output: Y=f (x) Define Improve Measure Analyze Control

24. Benefits of Using $ix $igma Productivity Improvements Culture Change Customer Retention Product/Service Development Cycle Time Reduction Market-Share Growth Cost Reduction DMAIC

25. What makes it Attractive? Tool to Plan & Deliver Values To Customers Sets a Performance Goal for Everyone Accelerates the Rate Of Improvement Measurable Results Tied to the Bottom-line Promotes Learning & Cross-Pollination Executes Strategic Change Generates Sustained Results $ix $igma

27. Some Six Sigma Tools 6  Voice of the Customer Process Design/Re-design Process Management Creative Thinking SPC DOE