Six Sigma
Upcoming SlideShare
Loading in...5
×
 

Like this? Share it with your network

Share

Six Sigma

on

  • 2,256 views

 

Statistics

Views

Total Views
2,256
Views on SlideShare
2,254
Embed Views
2

Actions

Likes
1
Downloads
92
Comments
0

2 Embeds 2

http://www.lmodules.com 1
http://www.slideshare.net 1

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

Six Sigma Presentation Transcript

  • 1. Carnegie Mellon Software Engineering Institute Pittsburgh, PA 15213-3890 Six Sigma & Software/Systems Process Improvement Jeannine M. Siviy Software Engineering Measurement & Analysis Initiative Sponsored by the U.S. Department of Defense © 2002 by Carnegie Mellon University 11th ICSQ, October 2001. - page 1
  • 2. Carnegie Mellon Software Engineering Institute Outline / Objectives Six Sigma Overview Applications Survey Initiative Synergy Illustration © 2002 by Carnegie Mellon University SEPG 2002 - page 2
  • 3. Carnegie Mellon Software Engineering Institute Six Sigma Is… A Philosophy A Metric An Improvement Framework © 2002 by Carnegie Mellon University SEPG 2002 - page 3
  • 4. Carnegie Mellon Software Engineering Institute Six Sigma Philosophy Improve customer satisfaction by reducing and eliminating defects Greater Profits © 2002 by Carnegie Mellon University SEPG 2002 - page 4
  • 5. Carnegie Mellon Software Engineering Institute What is a Defect? Six Sigma: • Any product, service, or process variation which prevents meeting the needs of the customer and/or which adds cost, whether or not it is detected. Personal Software ProcessSM: • Defects or faults are the result of errors or mistakes. At a minimum, count a defect every time the program is changed during compile or test, where the change might be one character or multiple statements [Humphrey 95] SMPersonal Software Process and PSP are service marks of Carnegie Mellon University. © 2002 by Carnegie Mellon University SEPG 2002 - page 5
  • 6. Carnegie Mellon Software Engineering Institute Six Sigma Metrics “3.4 ppm” – the most-cited metric Other Measures • Defect Rate, parts per million (ppm) • Sigma Level • Defects per Unit (dpu) • Defects per Million Opportunities (dpmo) • Yield © 2002 by Carnegie Mellon University SEPG 2002 - page 6
  • 7. Carnegie Mellon Software Engineering Institute “3.4” and “Sigma” Metrics 1 New Car Buyer’s target: Lower Upper Target Spec Spec • 35 miles per gallon (mpg) • 29-41 mpg acceptable 29 35 41 Two Choices: Car 1 Car 2 35 +/- 2 mpg 35 +/- 1 mpg (35-29) / 2 = 3 (35-29) / 1 = 6 (41-35) / 2 = 3 (41-35) / 1 = 6 3 Sigma 6 Sigma ~3/1000 outside limits virtually 0 outside limits © 2002 by Carnegie Mellon University SEPG 2002 - page 7
  • 8. Carnegie Mellon Software Engineering Institute “3.4” and “Sigma” Metrics 2 Historical data: 1.5*standard deviation shift over time Car 2: Lower Upper Target Spec Spec •Mean shifts to 33.5 •“Mean - 6*Std Dev” now 29 35 41 extends below lower spec 27.5 33.5 39.5 •Extension corresponds to 3.4 Shifted Range ppm if normal distribution © 2002 by Carnegie Mellon University SEPG 2002 - page 8
  • 9. Carnegie Mellon Software Engineering Institute Six Sigma Process with Mean Shift Assumptions: • normal distribution • process mean shift of 1.5σ from nominal is likely • process mean and standard deviation are known • defects are randomly distributed throughout -3s σ -3σ s +3σ units s -6σ -6σ +6s σ +6σ • parts and process steps are independent © 2002 by Carnegie Mellon University SEPG 2002 - page 9
  • 10. Carnegie Mellon Software Engineering Institute Example Sigma Levels IRS - Tax Advice (phone-in) 100,000 PPM per Part or Process Step Restaurant Bills Doctor Prescription Writing 10,000 Payroll Processing Air Line Baggage Handling 1,000 Average Company Orders Placed on Factory 100 US Navy Aircraft Accidents Note: Sigma 10 Levels vary +/- 1s with source Domestic Airline publication date Flight Fatality Rate 1 2 3 4 5 6 7 (308537) (66807) (6210) (233) (3.4) Sigma Level (ppm for shifted process) [Harrold 98], [Harry 00] © 2002 by Carnegie Mellon University SEPG 2002 - page 10
  • 11. Carnegie Mellon Software Engineering Institute Statistical Thinking • Everything is a process • All processes have inherent variability • Data is used to understand variation and to drive decisions to improve the processes Special Cause Variation Data Spread due to Original Mean Common Cause Variation New mean after improvement (Spread due to common cause variation will re-establish itself.) [ASQ 00], [ASA 01] © 2002 by Carnegie Mellon University SEPG 2002 - page 11
  • 12. Carnegie Mellon Software Engineering Institute Everything Is a Process Example: Software Engineering Require- Design Code ments Compile Test Ship © 2002 by Carnegie Mellon University SEPG 2002 - page 12
  • 13. Carnegie Mellon Software Engineering Institute Six Sigma Improvement Framework Define Measure Analyze Improve Control © 2002 by Carnegie Mellon University SEPG 2002 - page 13
  • 14. Carnegie Mellon Software Engineering Institute Six Sigma Toolkit Define Measure Analyze Improve Control • Benchmark • 7 Basic • Cause & Effect • Design of Statistical • Baseline Tools Diagrams Experiments Controls: • Contract/Charter • Defect • Failure Modes & • Modeling • Control • Kano Model Metrics (i.e., Effects Analysis • Tolerancing Charts • Voice of the “ppm”) • Decision & Risk • Robust • Time Series Customer • Data Analysis Design methods • Voice of the Collection • Statistical Business Forms, Plan, Inference Non-Statistical • Quality Function Logistics • Control Charts Controls: Deployment • Sampling • Capability • Procedural • Process Flow Techniques • Reliability adherence Map Analysis • Performance • Project • Root Cause Mgmt Management Analysis • Preventive • “Management -5 Why’s activities by Fact” • Systems -4 What’s Thinking © 2002 by Carnegie Mellon University SEPG 2002 - page 14
  • 15. Carnegie Mellon Software Engineering Institute Design for Six Sigma (DFSS) New solutions • Rather than analysis of existing processes Focus • Customer and business • Emphasis on critical-to-quality characteristics Useful tools • Modeling, simulation, lean, systems thinking © 2002 by Carnegie Mellon University SEPG 2002 - page 15
  • 16. Carnegie Mellon Software Engineering Institute Applications Six Sigma applications for Systems and Software Engineering are emerging © 2002 by Carnegie Mellon University SEPG 2002 - page 16
  • 17. Carnegie Mellon Software Engineering Institute Survey of Applications 1 Allied Signal • 1997 air supply control system shutdowns • Black Belt project team commissioned to find solution Motorola • Inspection data analysis & unit test optimization • Design of experiments methods & test cases • Complexity analysis & resource allocations • Quantitative risk management via uncertainty modeling General Electric • DFSS • Six Sigma & Extreme Programming [Harry 00], [Stoddard 00], [Kelliher 01] Motorola & GE presentations available at http://seir.sei.cmu.edu © 2002 by Carnegie Mellon University SEPG 2002 - page 17
  • 18. Carnegie Mellon Software Engineering Institute Survey of Applications 2 Honeywell • PSPSM/TSPSM & Six Sigma - “TSP provides the data needed to apply Six Sigma” JP Morgan • Capability Maturity Model® (CMM®) & Six Sigma - “…Six Sigma methodology is beneficial on all levels of maturity.” NCR • CMM & Six Sigma - “…helps organizations working towards Level 4 & 5 deliver the best business results.” [Pavlik 00], [A-M 99], [Demery 01] Presentations available at http://seir.sei.cmu.edu © 2002 by Carnegie Mellon University SEPG 2002 - page 18
  • 19. Carnegie Mellon Software Engineering Institute Initiative “Synergy” CMM® • Level 1-3 • Level 4-5 CMM IntegrationSM (CMMISM) Personal Software ProcessSM (PSPSM) Team Software ProcessSM (TSPSM) ®Capability Maturity Model and CMM are registered in the U.S. Patent and Trademark Office. SMCMM Integration, CMMI, Personal Software Process, PSP, Team Software Process and TSP are service marks of Carnegie Mellon University. © 2002 by Carnegie Mellon University SEPG 2002 - page 19
  • 20. Carnegie Mellon Software Engineering Institute CMM and Six Sigma* σ Optimizing 5 • Organization-wide 6σ improvements and control Process σ • Correlation between key process areas & 6σ methods improvement σ • 6σ used within CMM efforts Quantitative 4 Process measured and controlled Defined σ 3 • Defined processes feed 6σ Process characterized for the organization and is proactive σ 2 • 6σ philosophy & method focus Repeatable Process characterized for projects and σ • 6σ “drilldown” drives local is often reactive (but threaded) improvements Initial σ 1 • 6σ may drive toward and Process unpredictable and poorly controlled accelerate CMM solution Six Sigma is enterprise-wide. Six Sigma focuses on “critical to quality” factors. *Similar comments apply to CMMI © 2002 by Carnegie Mellon University SEPG 2002 - page 20
  • 21. Carnegie Mellon Software Engineering Institute CMMI & Six Sigma* “Within” CMMI • Quantitative Project Management (QPM) • Organizational Process Performance (OPP) • Organizational Innovation and Deployment (OID) • Measurement & Analysis (MA) • Capability Levels • Generic Practices “Around” CMMI • SEPG process improvement rollout • Assessment methods • Prioritization of process areas *Similar comments apply to CMM © 2002 by Carnegie Mellon University SEPG 2002 - page 21
  • 22. Carnegie Mellon Software Engineering Institute D M A I C Illustration – “Define” 1 Business Driver • Need 10% cost reduction in order to compete in the marketplace and stay in business Baseline data (PSP) • Productivity: 19 LOC/hr • 33% of development time spent fixing defects • Approximately 250 defects/KLOC © 2002 by Carnegie Mellon University SEPG 2002 - page 22
  • 23. Carnegie Mellon Software Engineering Institute D M A I C Illustration – “Define” 2 Goal: • Reduce or prevent defects to reduce cost Quantitatively speaking: • Reduce cycle time by 22 minutes/program • Reduce fix time by 1.3 minutes/defect • Reduce defects by 6/program • Reduce defect density to 190 defects/LOC … or a combination that produces 21 LOC/hr © 2002 by Carnegie Mellon University SEPG 2002 - page 23
  • 24. Carnegie Mellon Software Engineering Institute D M A I C Illustration – “Define, Measure” PSP-based Process Map: Design Code Compile Test • Requirements • Resources • Code • Executable Code • Estimate • Test Plan • Concept design • Detailed Design • Code • Executable • Functional • Test cases • Data: Code Code • Data: - Defect Quantity • Data: • Data: - Design Review - Fix time - Defect Quantity - Defect Quantity defects - Defect Injection - Fix time - Fix time - Fix time Phase - Defect Injection - Defect Injection - Phase duration - Phase duration Phase Phase - Phase duration -Phase duration © 2002 by Carnegie Mellon University SEPG 2002 - page 24
  • 25. Carnegie Mellon Software Engineering Institute D M A I C Illustration – “Analyze” Opportunities to reduce repair time • Defects removed in test: 78% of repair time • Defects injected in design: 25% of repair time • Defects injected in code: 56% of repair time • Syntax defects in general: 63% of defects Removed Defects Fix Time Injected Defects Fix Time, Baseline programs 1-6 Baseline programs 1-6 Code: 2% Test: 19% Design: 25% Compile: 20% Test: 78% Code: 56% © 2002 by Carnegie Mellon University SEPG 2002 - page 25
  • 26. Carnegie Mellon Software Engineering Institute D M A I C Illustration – “Improve” Improvement Plan at Program 6 • Syntax checklist • Well-timed reviews • Subcategories within defect types Defect Density Mean Comparison: Defect Density 450 450 400 400 350 350 Defects/KLOC defects/KLOC N&C 300 300 250 250 200 200 150 150 100 100 50 50 0 0 first second All Pairs 1 2 3 4 5 6 7 8 9 10 Pre Post Tukey-Kramer Program Improvement Improvement 0.05 Group © 2002 by Carnegie Mellon University SEPG 2002 - page 26
  • 27. Carnegie Mellon Software Engineering Institute D M A I C Illustration – “Control” Tracking performance • Quantitative goal statement • Hypothesized root causes • Countermeasures & contribution to impact • Key impact indicators Productivity (LOC/hr) Direct causes (from countermeasures): 30 25 • Fewer defects injected in code & test Goal = 21 20 • Defects removed earlier, faster (i.e., in 15 10 design & code) 5 0 Root cause (need new countermeasures): 1 2 3 4 5 6 7 8 9 10 Program Number • “Re-learning” curve © 2002 by Carnegie Mellon University SEPG 2002 - page 27
  • 28. Carnegie Mellon Software Engineering Institute D M A I C Illustration – Analysis Summary Tools used in full analysis included • Process Mapping • Descriptive statistics • Means comparisons & significance testing • Plots - Pie Charts - Trends - Phase profiles - Histograms - Pareto charts - Correlation plots • Cause & Effect Diagrams • “Management by Fact” Focus was exploratory, investigative • Ready for stability & control monitoring © 2002 by Carnegie Mellon University SEPG 2002 - page 28
  • 29. Carnegie Mellon Software Engineering Institute D M A I C Illustration – Scaling up Illustration Real Life § Quickly drilled down from § Drill down may be complex, high level cost goal to may span wide breadth of personal improvement organization § Defined process in place § May need to select or define process § Measures in place § May need to develop measures § Continuous incremental § Continuous incremental improvements improvements § Event-based “step- § Event-based “step-change” change” improvements improvements § Re-learning curve § Constantly changing skills, technologies § Personal data § Non-attributed data (e.g., team, project) § Used productivity as one § Excessive productivity focus of impact measures may drive unwanted behaviors © 2002 by Carnegie Mellon University SEPG 2002 - page 29
  • 30. Carnegie Mellon Software Engineering Institute Advancing the State of 6σ & SW/SE Repository of Examples -http://seir.sei.cmu.edu -concrete visualization -relationship to models, initiatives -variety of tools -many views –project, process, product –software, systems –maturity/capability levels Repository of Benefits © 2002 by Carnegie Mellon University SEPG 2002 - page 30
  • 31. Carnegie Mellon Software Engineering Institute Summary Customer satisfaction is key driver All efforts should link to business results © 2002 by Carnegie Mellon University SEPG 2002 - page 31
  • 32. Carnegie Mellon Software Engineering Institute Contact Information Jeannine Siviy Software Engineering Institute Measurement & Analysis Initiative jmsiviy@sei.cmu.edu 412-268-7994 © 2002 by Carnegie Mellon University SEPG 2002 - page 32
  • 33. Carnegie Mellon Software Engineering Institute References 1 [A-M 99] Abdel-Malek, Nabil and Anthony Hutchings, Applying Six Sigma Methodology to CMM for Performance Improvement, JP Morgan, European SEPG 1999, (slides available to SEIR contributors at http://seir.sei.cmu.edu) [ASA 01] American Statistical Association, Quality & Productivity Section, Enabling Broad Application of Statistical Thinking, http://web.utk.edu/~asaqp/thinking.html, 2001 [ASQ 00] ASQ Statistics Division, Improving Performance Through Statistical Thinking, Milwaukee: ASQ Quality Press, 2000. H1060 [Arnold 99] Arnold, Paul V., Pursuing the Holy Grail, MRO Today, June/July 1999, www.progressivedistributor.com/mro/archives/editorials/editJJ1999.html [Breyfogle 99] Breyfogle III, Forrest W., Implementing Six Sigma: Smarter Solutions Using Statistical Methods, John Wiley & Sons, 1999 [Bylinsky 98] Bylinsky, Gene, How to Bring Out Better Products Faster, Fortune, 23 November 1998 [Demery 01] Demery, Chris and Michael Sturgeon, Six Sigma and CMM Implementation at a Global Corporation, NCR, SEPG 2001, (slides available to SEIR contributors at http://seir.sei.cmu.edu) [Harrold 99] Harrold, Dave, Designing for Six Sigma Capability, Control Engineering Online, January 1999, http://www.controleng.com/archives/1999/ctl0101.99/01a103.htm © 2002 by Carnegie Mellon University SEPG 2002 - page 33
  • 34. Carnegie Mellon Software Engineering Institute References 2 [Harry 00] Harry, Mikel, Six Sigma: The Breakthrough Management Strategy Revolutionizing the World’s Top Corporations, Doubleday, 2000 [Kelliher 00] Kelliher, Timothy P., Daniel J. Blezek, William E. Lorensen, James V. Miller, Six Sigma Meets Extreme Programming, General Electric Corporate R&D, (paper available to SEIR contributors at http://seir.sei.cmu.edu) [Pavlik 00] Pavlik, Rich, Cary Riall and Steve Janiszewski, Deploying PSPSM, TSPSM, and Six Sigma Plus at Honeywell, Honeywell Air Transport, SEPG 2000, (slides available to SEIR contributors at http://seir.sei.cmu.edu) [Purcell 00] Purcell, Leitha, Experiences Using Six Sigma in a SW-CMM® Based Process Improvement Program, Northrop Grumman, SEPG 2000, (slides available to SEIR contributors at http://seir.sei.cmu.edu) [Pyzdek 01] Pyzdek, Thomas, The Six Sigma Handbook, McGraw-Hill Professional Publishing, 2001 [Stoddard 00] Stoddard, Robert W., Implementing Six Sigma in Software, Motorola, Inc., Software Engineering Symposium 2000, (slides available to SEIR contributors at http://seir.sei.cmu.edu) © 2002 by Carnegie Mellon University SEPG 2002 - page 34
  • 35. Carnegie Mellon Software Engineering Institute Additional Reading 1 Books (General Six Sigma Topics, not software-specific): Breyfogle III, Forrest W., Implementing Six Sigma: Smarter Solutions Using Statistical Methods, John Wiley & Sons, 1999 Breyfogle, III, Forrest W., Cupello, James M., Meadows, Becki, Managing Six Sigma, John Wiley & Sons Harry, Mikel, Six Sigma: The Breakthrough Management Strategy Revolutionizing the World’s Top Corporations, Doubleday, 2000 Pyzdek, Thomas, The Six Sigma Handbook, McGraw-Hill Professional Publishing, 2001 Web pages & Web sites: American Society for Quality Six Sigma Forum, http://www.sixsigmaforum.com/concepts/var/index.shtml International Quality Federation, www.iqfnet.org (Follow the black belt links) Six Sigma Academy, www.6-sigma.com Software Engineering Information Repository: http://seir.sei.cmu.edu (Follow links to Measurement area then to Six Sigma) SEI Software Technology Review: http://www.sei.cmu.edu/str/descriptions/sigma6.html © 2002 by Carnegie Mellon University SEPG 2002 - page 35
  • 36. Carnegie Mellon Software Engineering Institute Additional Reading 2 Journals (URL’s subject to change without notice) 6 Sigma Pro, Quality Digest, May 2000, http://www.qualitydigest.com/may00/html/sixsigmapro.html 6 Sigma Con, Quality Digest, May 2000, http://www.qualitydigest.com/may00/html/sixsigmacon.html Arnold, Paul V., Pursuing the Holy Grail, MRO Today, June/July 1999, http://www.mrotoday.com/mro/archives/Editorials/editJJ1999.htm Card, David, Sorting out Six Sigma and the CMM, IEEE Software, May/June 2000 Dusharme, Dirk, Six Sigma Survey: Breaking Through the Six Sigma Hype, Quality Digest, November 2001, http://www.qualitydigest.com/nov01/html/sixsigmaarticle.html Paul, Lauren Gibbons, "Practice Makes Perfect," CIO Enterprise Magazine, 15 January 1999, http://www.cio.com/archive/enterprise/011599_process.html Harrold, Dave, Designing for Six Sigma Capability, Control Engineering Online, January 1999, http://www.controleng.com/archives/1999/ctl0101.99/01a103.html Harrold, Dave, Optimize Existing Processes to Achieve Six Sigma Capability, Control Engineering Online, March 1999, http://www.controleng.com/archives/1999/ctl0301.99/03a301.html Lahiri, Jaideep, The Enigma of Six Sigma, The Net Business of the India Today Group, http://www.india-today.com/btoday/19990922/cover.html Pyzdek, Thomas, Why Six Sigma is Not TQM, Quality Digest, February 2001, http://www.qualitydigest.com/feb01/html/sixsigma.html Pyzdek, Thomas, Ignore Six Sigma at Your Peril, Quality Digest, April 2001, http://www.qualitydigest.com/april01/html/sixsigma.html © 2002 by Carnegie Mellon University SEPG 2002 - page 36