6 Sigma and Reliability Engineering Integration Webinar

Transcription

1 6 Sigma and Reliability Engineering Integration Webinar by Louis LaVallee Mike Silverman 5/7/2011 Ops A La Carte 1

2 DFSS / DFR Webinar Agenda Wednesday 4 May 1011 Agenda Introductions and Agenda Review 5 minutes DFSS DFR Integrating the two techniques Wrap-up Q&A 5/7/2011 Ops A La Carte 2

3 Presenter s Biographical Sketch 5/7/2011 Ops A La Carte 3

4 Presenter s Biographical Sketch Lou LaVallee Lou is a Senior Reliability Consulting with Ops A La Carte. Lou has over 30 years of experience as a sr. physicist, systems engineering mgr. and design quality and reliability engineering mgr. at Xerox. Lou has strong validation experience of design quality and reliability through product reviews and customer interaction. Lou is a master black belt and has trained several thousand engineers in quality engineering, critical parameter management, robust design, six sigma methods, QFD, systems engineering, and reliability engineering. Lou works in the upstate New York area. Lou is a Certified Reliability Engineer (CRE). 5/7/2011 Ops A La Carte 4

5 Presenter s Bio Sketch Mike Silverman Mike is founder and managing partner at Ops A La Carte Through Ops A La Carte, Mike has had extensive experience as a consultant to high-tech companies, and has consulted for over 500 companies in a variety of different industries including medical, defense, space, energy, consumer, telecom, and many others. Mike just completed his first book on Reliability entitled How Reliable Is Your Product: 50 Ways to Improve Product Reliability Mike has 25 years of reliability, quality, and compliance experience, the majority in start-up companies. Mike is also an expert in accelerated reliability techniques, including HALT and HASS. Mike and his team at Ops have authored and published over 50 papers and 30 seminars on reliability techniques and has presented these around the world including China, Germany, Japan, Taiwan, Korea, and Canada. Mike is the IEEE Reliability Society Santa Clara Valley Chapter President. Mike is a Certified Reliability Engineer (CRE) 5/7/2011 Ops A La Carte 5

6 Overview of Ops A La Carte Reliability Services and Training 5/7/2011 Ops A La Carte 6

7 COMPANY OVERVIEW Confidence in Reliability 5/7/2011 Ops A La Carte 7

8 Our Company is a privately-held professional reliability engineering firm founded in 2001 and headquartered in Santa Clara, California with offices in China, India and Singapore. was named one of the top 10 fastest growing, privately-held companies in the Silicon Valley in 2006 and 2009 by the San Jose Business Journal. is a solid company that has been profitable every quarter since its inception due to its outstanding reputation, customer value and scalable business model. 5/7/2011 Ops A La Carte 8

9 Our Team is made up of a group of highly accomplished Reliability Consultants. Each of our consultants has 15+ years of Reliability Engineering and Reliability Management experience in various industries. We tap a large network of labs, test facilities, and talented engineering professionals to quickly assemble resources to supplement your organization. 5/7/2011 Ops A La Carte 9

10 Ops Solutions Ops provides end-to-end solutions that target the corporate product reliability objectives Ops Individual A La Carte Consulting Ops identifies and solves the missing key ingredients needed for a fully integrated reliable product Ops Training Ops highly specialized leaders and experts in the industry train others in both standard and customized training seminars Ops Testing Ops state-of-the-art provides comprehensive testing services 5/7/2011 Ops A La Carte 10

11 assists clients in developing and executing any and all elements of Reliability through the Product Life Cycle. has the unique ability to assess a product and understand the key reliability elements necessary to measure/improve product performance and customer satisfaction. pioneered Reliability Integration using multiple tools in conjunction throughout each client s organization to greatly increase the power and value of any Reliability Program. 5/7/2011 Ops A La Carte 11

12 Testing Services Our own lab facility located in Northern California in the heart of Silicon Valley. We provide HALT/HASS services on a worldwide basis, using partner labs for tests outside California. Second oldest HALT facility in the world, established in 1995 (originally owned by QualMark) HALT equipment has all latest technology only lab in region Highly-experienced staff with over 100 years of combined experience in HALT and HASS Tested over 500 products in over 30 different industries Our HALT/HASS services are fully integrated with our other consulting services. Ops A La Carte 5/7/

13 Ops New Reliability Book How Reliable Is Your Product? 50 Ways to Improve Product Reliability A new book by Ops A La Carte LLC Founder/Managing Partner Mike Silverman The book focuses on Mike s experiences working with over 500 companies in his 25 year career as an engineer, manager, and consultant. It is a practical guide to reliability written for everyone in your organization. In the book we give tips and case studies rather than a textbook full of formulas. Available January 2011 in hardback for $44.95 or ebook for or For more info, go to 5/7/2011 Ops A La Carte 13

14 Upcoming Seminars May 9-13 Santa Clara and via webinar TRACK 1: DFX TOOLS Design for Reliability (DfR) May 9-10 Design for Six Sigma (DFSS) May 11 Design for Mechanical Reliability (DfMR) May 12 Design for Warranty (DfW) May 13 Design for Software Reliability (DfS) May 13 TRACK 2: REL TOOLS: ALT/DOE/RCA Design of Experiments May 9-10 Best Accelerated Tests (BART) May Root Cause Analysis - May 13 Details for all are on the Ops Education Schedule for /7/2011 Ops A La Carte 14

15 FREE Webinars for 2011 Feb 17 Medical Device Seminar/Webinar, San Jose Mar 2 Warranty Webinar (coincides with Warranty Chain Management Workshop on March 17) Mar 3 Implantable Medical Seminar, Santa Clara Mar 22 Book signing for How Reliable Is Your Product, Santa Clara Apr 6 Solar Reliability Challenges May 4 DfSS vs. DfR Webinar (tied with Symposium and World Quality Conference) Jun 1 How to Use HALT with Prognostics (tied with Prognostics Conference) Details for all are on our site at 5/7/2011 Ops A La Carte 15

16 Upcoming Events May 25 SEMA (Solar) Event, San Jose. We will be giving a reliability presentation May 25 ASQ Medical, Sunnyvale. We will be giving a reliability presentation June 6 MD&M East, New York. We will be giving a one day seminar on medical reliability testing June 7-9 ARS, San Diego. We will be exhibiting and giving two presentations on reliability. June PHM Conference, Denver. We will be exhibiting and giving a presentation on reliability. Details for all are on our site at 5/7/2011 Ops A La Carte 16

17 6 Sigma and Reliability Engineering Integration Webinar by Louis LaVallee Mike Silverman 5/7/2011 Ops A La Carte 17

18 Attendees List Summary Never used DFR 50% Early DFR User 18% Major DFR User 32% Never used DFSS 42% Early DFSS User 33% Major DFSS User 25% Never used DFSS & DFR integrated together 80% Early DFSS & DFR User (poorly integrated) 13% Major DFSS & DFR User 7% 5/7/2011 Ops A La Carte 18

19 Variation Discussion 5/7/2011 Ops A La Carte 19

20 Variation for 6 and reliability Variation Definition: A difference between two of more similar things. Ideally, two cheeseburgers from McDonalds are identical. The first one in the morning and the last one in the evening are identical. Much later 5/7/2011 Ops A La Carte 20

21 Variation & its countermeasures Search for root cause & eliminate it Screen out defectives (scrap and rework) Feedback/feed forward control systems Tighten tolerances (control, noise, signal factors) Add a subsystem to balance the problem Calibration & adjustment Robust design (Parameter design/tolerance design) Change the concept Turn off the power 5/7/2011 Ops A La Carte 21

22 Polling Question 1 Have you used any of the previous variation countermeasures? Yes, we have used 1 method Yes, we have used > 1 methods No, but we have used other methods No, we have not used any methods I don t know 5/7/2011 Ops A La Carte 22

23 Standard Normal distribution (Probability density function) Mean = 0., std dev=1.0 Point of Inflection 1 Demarcations shown at 6 different standard deviation levels including fractional area under curve e.g % of the area under the curve is between +1 and 1 5/7/2011 Ops A La Carte 23

24 LSL All are in spec. Which distribution is preferred to minimize quality loss USL freq LSL target target target Average Quality Loss Q for nominal the best measure Q k{( x ) k loss _ coefficient 2 2 } $ / unit 2 target 5/7/2011 Ops A La Carte 24

25 Brownian Motion & Variation 5/7/2011 Ops A La Carte 25

26 Central Limit Theorem Simulation & Variation What happens to shape of distribution when something nonrandom occurs? What does it mean when you don t observe a normal distribution? 5/7/2011 Ops A La Carte 26

27 Polling Question 2 Have you ever experienced a non normal distribution? Yes, I have seen that once Yes, I see that often No I have never seen that I don t know 5/7/2011 Ops A La Carte 27

28 DFSS &Reliability Integration the process of seamlessly, cohesively integrating reliability and 6 Sigma processes together to maximize quality & reliability at the lowest possible cost 5/7/2011 Ops A La Carte 28

29 Design for Six Sigma 5/7/2011 Ops A La Carte 29

30 Polling Question 3 Does you current job require increasing Quality engineering expertise? Yes No I don t know 5/7/2011 Ops A La Carte 30

31 Evolution of 6 6 Term originally comes from statistics, where is Greek symbol for standard deviation. is scalar measurement of univariate variation. Six standard deviations between equal bilateral spec limits is implied. Statistics help us measure and understand both individual data points, averages, and variation. Primary focus is achieving improvements in quality and cost. Lean Focus is on eliminating non customer value added waste in a product, process or service. Result is reducing service cycle times, improving on time delivery performance, and reducing cost. Lean 6 Combines the speed and power of both Lean and 6 Only a fast and responsive process is capable of achieving high quality, and only a high quality process can sustain high velocity 5/7/2011 Ops A La Carte 31

32 Six Sigma has evolved from product focus (defect reduction) to project focus (cost reduction) to customer value (productivity) to enterprise performance (bottom line growth). Six sigma is always in a continuous improvement mode. Each time you look, a new best practice has been added. DFSS Design for Six Sigma is a similar set of quality improvement tool used when the process or product has not yet been developed. Instead of fixing problems within existing process, DFSS is designed to prevent problems before they occur by creating a design based on the principles of Six Sigma. DFR has evolved from industrial activities focused on problem solving downstream. It has evolved to support and model upstream engineering and scientific activities. Many products now have requirements defined for both reliability and maintainability. 5/7/2011 Ops A La Carte 32

33 DEFINE ACTIVITIES Identify Problem Complete Charter Develop SIPOC Map Map Business Process Map Value stream Gather VOC & VO Business Develop CCR & CBR s Finalize Product Focus ACTIVITIES in 1 st four phases DMAIC Process Road Map MEASURE ANALYZE Identify Key Input, Process and output Metrics Develop Operational Definitions Develop Data Collection Plan Validate Measurement System Collect Baseline Data Determine Process Performance/Capability Opportunity Propose Critical X s Prioritize Critical X s Conduct Root Cause Analysis on Critical X s Validate Critical X s Estimate the Impact of Each X on Y QuantifyOpportunity Prioritize Root Cause IMPROVE Develop Potential Solutions Develop Evaluation Criteria & Select Best Solutions Evaluate Solution for Risk Optimize Solution Develop To Be Process Map(s) and High Level Implementation Plan Develop Pilot Plan & Pilot Solution 5/7/2011 Ops A La Carte 33

34 DEFINE MEASURE ANALYZE IMPROVE Pareto Charts Project Selection Tools PIP Management Process Value Stream Map Various Financial Analysis Charter Form Stakeholder Analysis Communication Plan SIPOC Map High Level Process Map Non Value Added Analysis VOC and Kano Analysis RACI and Quad Charts SIPOC Map Operational Definitions Data Collection Plan Statistical Sampling Measurement System Analysis(MSA), Gage R&R Constraint Identification Setup Reduction Generic Pull Kaizen TPM Control Charts Process Capability TOOLS: 1 st 4 phases Pareto Charts C&E Matrix C&E/Fishbone Diagrams Brainstorming Detailed As Is Process Maps Basic Statistical Tools Supply Chain Accelerator Analysis Non Value Added Analysis Hypothesis Testing FMEA Box Plots Interaction Plots Simple & Multiple Regression ANOVA (1 way & 2 way) Logistic Regression Analysis of Means Box Cox Transformations Chi Square Analysis MultiVari Plots Brainstorming Benchmarking Process Improvement Techniques Line Balancing Process Flow Improvement Replenishment Pull Purchasing and Sales Strategy Poka Yoke FMEA Solution Selection Matrix To Be Process Maps Piloting and Simulation Response surface Methodology Queuing Theory 5/7/2011 Ops A La Carte 34

35 6 Fundamental Concept Y = f(x) Y s Dependent Output Effect Symptom Monitor X 1, X N Independent Input process Cause Problem Control In reliability engineering for example, Y is the stochastic variable, times to failure, and F(x) is the failure mechanism, or mechanistic model. Variation is implied, as well as response 5/7/2011 Ops A La Carte 35

36 Recent Newsworthy DFR & DFSS candidates Japan Nuclear Power Plant reactors and generators BP deep water oil rig blowout preventer malfunction Toyota accelerator pedal recall GM, Ford, Chrysler automobile reliability & recall issues Hurricane Katrina levee and flood wall failures 20 Airline Crashes worldwide in 2010 Space Shuttle Challenger explosion I 35W Mississippi River bridge Collapse San Bruno Ca. gas line explosion Cruise ship engine failures off San Diego coa Southwest Air/Boeing aluminum panel break away (rivet holes too big) 5/7/2011 Ops A La Carte 36

37 Six Sigma with lean is the integration of two powerful Business Improvement Approaches SIX SIGMA Precision + Accuracy + VOC LEAN Speed + Low Cost + Flexibility On target Performance 5/7/2011 Ops A La Carte 37

38 LEAN 6 SIGMA 5/7/2011 Ops A La Carte 38

39 6 + Lean + Reliability 6 approach is a compilation and integration of best quality engineering practices, business processes, and tools from the past 80 years. Lean includes the 6 approach plus additional efficiency enhancements like reducing waste, creating higher speed processes, lower costs, and improved flexibility. Reliability engineering is a set of tools, processes, and activities to augment and complement the lean 6 approach. It includes design for reliability (DFR), reliability characterization, reliability optimization, life cycle cost minimization, forecasting, maximizing maintainability, durability, availability, safety, etc. There is overlap in some areas as might be expected, like the common use of quality function deployment (QFD) to capture and translate voice of customer, design of experiments (DOE), Modeling & Simulation, 5/7/2011 Ops A La Carte 39

40 Function Structure Behavior Decomposition Geometry/shape Materials, indented lists BOM, eng. drawings Structure Mapping Synthesis Mapping Function Behavior Physics/Chemistry Fundamental models Theoretical models Empirical models Simulation models Flow down linkages Mapping Dysfunction Symptoms Failure modes Failure mechanisms Ideally zero measures 5/7/2011 Ops A La Carte 40

41 6 Sigma : The Importance of Measurement Processes must be measured to establish a baseline (current condition) against which future improvements can be quantified If you can t measure it, you can t manage it. W. Edwards Deming ( ) Can you prove that the data were produced by your measurement system without distortion or contamination and without affecting the process being observed? Is data valid and meaningful? Thousands of available measurement systems add to the complexity of experimentation. 5/7/2011 Ops A La Carte 41

42 Measurements for Experimentation What to measure: Measurements are selected not given. Measurement Criteria: Continuous numbers Complete Fundamental Practical Monotonic with control factors Related to basic energy transfer mechanism of system Measurement capability Engineering Focus Picks up meaningful information about dysfunction Types of Measurement Data: Binary (go/no go) Categorical (ordinal, nominal, ranks, ) Counts or frequencies Counted fractions, percentages Scalar, Vector quantities Complex, matrix 5/7/2011 Ops A La Carte 42

43 Lack of monotonicity : displacement response for paper stacks Many different ways to get the same number X Stack 1 Stack 2 X X Stack 3 Stack 4 X X 5/7/2011 Ops A La Carte 43

44 5/7/2011 Ops A La Carte 44

45 Building reliability into concept selection 5/7/2011 Ops A La Carte 45

46 Original List of Criteria for car horn concepts Ease of achieving decibel (sound level) Ease of achieving Hertz (sound frequency) * Resistance to corrosion (water, pollutants) * Resistance to vibration, shock, acceleration/deceleration, wear and tear * Resistance to temperature cycling and extremes * Low power consumption * Ease of maintenance Small size *Long service life Low manufacturing cost Ease of installa on * Long shelf life Criteria Added During the Round 1 Discussion Quick response me *Small number of parts simplicity of design Ease of opera on (accessibility, emergency response) Ease of integra on into the automobile subsystems Low weight * Reliability issues during concept selection 5/7/2011 Ops A La Carte 46

47 DFSS Project Focus and Selection 1. Identify potential projects with high return/ease of implementation 2. Use accelerated approach on initial projects (1 week) 3. Use analytical methods to estimate potential benefits (ROI, NPV) and rank order alternate projects 4. Avoid project scope creep 5. Successfully complete projects, and fold them back into further training 6. Determine cost savings from project completion. 5/7/2011 Ops A La Carte 47

48 Slope (sensitivity) and Root mean squared error parameters Simple Metal Helical Compression Spring Force vs Displacement Force F 0,0 x (mm) Ideally, all points fall on dashed line Origin 0,0 Noise factors add variation Variation may exceed tolerable limits 5/7/2011 Ops A La Carte 48

49 Case Study: Four hypothetical I V curves with two noise conditions Q=charge density 5/7/2011 Ops A La Carte 49

50 I V staircase charging data 5/7/2011 Ops A La Carte 50

51 Marginal Means Plot For Photoreceptor DOE 5/7/2011 Ops A La Carte 51

52 Take Aways from P/R experimental Design Ideal function development was used successfully to improve robustness and reliability Y= f(x) = X+ ANOVA & regression tools provided optimum conditions in 1/6 time of time to failure testing. +3 db gain in S/N ratio 2x reliability improvement. Development engineers and their management are always looking for efficient ways to reduce expenditures for system reliability testing. This method works well for many situations. 5/7/2011 Ops A La Carte 52

53 Polling Question 4 Could you use DOE to improve your golf game? Yes No I don t know 5/7/2011 Ops A La Carte 53

54 5/7/2011 Ops A La Carte 54

55 Quality Function Deployment Whats become Hows Matrix approach becomes a powerful framework for systems engineering, quality engineering and reliability engineering 5/7/2011 Ops A La Carte 55

56 System level VOC Copy quality as good as Lithography Line density, background density HOQ #1 Module level Transferred mass/area Patch reflectance Toner dispense rate Line density, background density HOQ #2 Printer Toner Dispenser

57 Subsystem level Toner Dispense rate Auger speed Auger clearance HOQ 3 Component level Drive gear pitch diam Auger diam Housing diam Auger speed Auger clearance HOQ 2 Printer Toner Dispenser

58 Injection Molding Manufacturing Process Level From HOQ # 4 Pressure Temperature Shot weight Drive Gear Pitch diam Auger diameter Housing Diameter HOQ # 5 There may be many requirements and engineering specs. Only those that are new, difficult, extremely important, have a negative correlation, or offer/enable a competitive opportunity with be carried through. Critical few only

59 House 1 Pencil example 5/7/2011 Ops A La Carte 59

60 Successful Lean Six Sigma Deployment Requires Company Wide Involvement Centralized Coordination & Training Operations Leadership Owns Vision, direction integration, business results Lead Change Project owner Implements solutions Owns financial Results Part time as part of job Project Sponsors Green Belts Participates on BB teams/ leads projects Part time on projects All Employees Understand Vision Apply concepts to job & work area Project Team members Provides project specific support Can be yellow or green belt Part time on prrojects Deployment managers Black Belts Leads BU performance Improvement Prioritized Projects Full time position Master Black Belts Train black belts/green belts Coaching Leads Lean 6 Sigma projects Full time position Leads six sigma projects Trains and coaches project teams Full time position 5/7/2011 Ops A La Carte 60

61 DFR Design for Reliability 5/7/2011 Ops A La Carte 61

62 Polling Question 5 Does you current job require increasing Reliability engineering expertise? Yes No I don t know 5/7/2011 Ops A La Carte 62

63 DFR is a process that describes the entire set of tools and activities that support the effort to increase a product s reliability and are applied from the early concept stage of a design all the way through to product obsolescence. Reliability Assessment Define Reliability Requirements early in design cycle. Think preventive rather than reactive. Don t create the problem. Technology readiness no new inventions required so that a predictable schedule can be maintained Control complexity and newness of work processes, modules, parts, software, design rules, manufacturing processes, to control initial reliability Selection of activities and tools to enable delivery of good reliable products. Training (and certification) for both management and engineers in deficient areas, eg. DFR. 5/7/2011 Ops A La Carte 63

64 How Do You Set up a DfR Program? A reliability assessment is our recommended first step in establishing a reliability program. This mechanism is the appropriate forum for selecting the best tools for each product life cycle phase. 5/7/2011 Ops A La Carte 64

65 Reliability Program Assessment Initiate a Reliability Program Determine next best steps Reduce customer complaints Select right tools Improve reliability Program Plan $ Profits Goal market share Gap Analysis Benchmarking satisfaction field failures complaints Statistical Data Analysis Assessment Interviews 5/7/2011 Now $ unreliability? Unknown Reliability? Ops A La Carte A detailed evaluation of an organization s approach and processes involved in creating reliable products. The assessment captures the current state and leads to an actionable reliability program plan. 65

66 Reliability Maturity Matrix Measurement Category Management Understanding and Attitude Reliability status Problem handling Cost of Reliability as % of net revenue Feedback process DFR program status Summation of reliability posture 5/7/2011 Stage I: Uncertainty No comprehension of reliability as a management tool. Tend to blame reliability engineering for reliability problems Reliability is hidden in manufacturing or engineering departments. Reliability testing probably not part of organization. Emphasis on initial product functionality. Fire fighting; no root cause analysis or resolution; lots of yelling and accusations. Warranty: unknown Reported: unknown Actual: 20% None. No reliability testing. No field failure reporting other than customer complaints and returns. No organized activities. No understanding of such activities. We don t know why we have problems with reliability Stage II: Awakening Recognizing that reliability management may be of value but not willing to provide money or time to make it happen. A stronger reliability leader appointed, yet main emphasis is still on an audit of initial product functionality. Reliability testing still not performed. Teams are set up to solve major problems. Longrange solutions are not identified or implemented. Warranty: 3% Reported: unknown Actual: 18% Some understanding of field failures and complaints. Designers and manufacturing do not get meaningful information. Organization told reliability is important. DFR tools and processes inconsistently applied and only when time permits. Is it absolutely necessary to always have problems with reliability? Ops A La Carte Stage III: Enlightenment Still learning more about reliability management. Becoming supportive and helpful. Reliability manager reports to top management, with role in management of division. Corrective action process in place. Problems are recognized and solved in orderly way. Warranty: 4% Reported: 8% Actual: 12% Accelerated testing of critical systems during design. System level modeling and testing. Field failures analyzed and root causes reported. Implementation of DFR program with thorough understanding and establishment of each tool. Through commitment and reliability improvement we are identifying and resolving our problems. Stage IV: Wisdom Participating. Understand absolutes of reliability management. Recognize their personal role in continuing emphasis. Reliability manager is an officer of company; effective status reporting and preventive action. Involved with consumer affairs. Problems are identified early in their development. All functions are open to suggestion and improvement. Warranty: 3% Reported: 6.5% Actual: 8% Refinement of testing systems only testing critical or uncertain areas. Increased understanding of causes of failure allow deterministic failure rate prediction models DFR program active in all areas of division not just design & mfg ing. DFR normal part of R&D and manufacturing. Failure prevention is a routine part of our operation. Stage V: Certainty Consider reliability management an essential part of company system. Reliability manager is on board of directors. Prevention is main concern. Reliability is a thought leader. Except in the most unusual cases, problems are prevented. Warranty: 1.5% Reported: 3% Actual: 3% The few field failures are fully analyzed and product designs or procurement specifications altered. Reliability testing done to augment reliability models. Reliability improvement is a normal and continued activity. We know why we do not have problems with reliability. 66

67 Reliability Maturity Matrix Lets look at one row to get a better understanding. Measurement Category Stage I: Uncertainty Stage II: Awakening Stage III: Enlightenment Stage IV: Wisdom Stage V: Certainty Problem handling Fire fighting; no root cause analysis or resolution; lots of yelling and accusations. Teams are set up to solve major problems. Longrange solutions are not identified or implement ed. Corrective action process in place. Problems are recognize d and solved in orderly way. Problems are identified early in their developm ent. All functions are open to suggestio n and improvem ent. Except in the most unusual cases, problems are prevented. 5/7/2011 Ops A La Carte 67

68 Next Steps Determine current state of your organization (Summary of Assessment) Identify strong and weak areas Goal Setting Market Analysis to gather requirements Benchmarking Gap Analysis Develop plan and implement 5/7/2011 Ops A La Carte 68

69 RELIABILITY GOAL SETTING Establish the target in an engineering meaningful manner 5/7/2011 Ops A La Carte 69

70 Reliability Definition Reliability is often considered quality over time. Reliability is The ability of a system or component to perform its required functions under stated conditions for a specified period of time IEEE /7/2011 Ops A La Carte 70

71 Reliability Goals & Metrics Summary Reliability Goals & Metrics tie together all stages of the product life cycle. Well crafted goals provide the target for the business to achieve, they set the direction. Metrics provide: the milestones, the are we there, yet, and the feedback that all elements of the organization require to stay on track toward the goals. 5/7/2011 Ops A La Carte 71

72 Reliability Goals & Metrics Summary A reliability goal includes each of the five elements of the reliability definition. Probability of product performance Intended function Specified life Specified operating conditions Customer expectations 5/7/ Ops A La Carte

73 Reliability Goals & Metrics Summary A reliability metric is often something that organization can measure on a relatively short, periodic basis: Predicted failure rate (during design phase) Field failure rate Warranty Actual field return rate Dead on Arrival rate 5/7/ Ops A La Carte

74 Reliability Program and Integration Plan A Reliability Program and Integration Plan is crucial at the beginning of the product life cycle because in this plan, we define: What are the overall goals of the product and of each assembly that makes up the product? What has been the past performance of the product? What is the size of the gap? What reliability elements/tools will be used? How will each tool be implemented and integrated to achieve the goals? What is our schedule for meeting these goals? 5/7/ Ops A La Carte

75 Example of Design for Reliability (DfR) Tools Reliability Modeling and Prediction Thermal Analysis Derating Analysis Failure Modes and Effects Analysis (FMEA) Fault Tree Analysis (FTA) Design of Experiments (DoE) Human Engineering/Human Factors Analysis Highly Accelerated Life Test (HALT) Accelerated Life Test (ALT) HALT vs. ALT RDT and ORT Highly Accelerated Stress Screen (HASS) Root Cause Analysis (RCA) Restriction of Hazardous Substances (RoHS) Outsourced Engineering and Reliability Field Data Analysis 5/7/ Ops A La Carte

76 DfR Tools by Phase Phase Activities Tools Design Concept Architecture and High Level Design Initial System Testing Final System Testing Operations and Maintenance Define project reliability requirements (Reliability Program and Integration Plan) Modeling & Predictions Defect Detection at System Level Verify Reliability Metrics Continuous assessment of product reliability Benchmarking Internal Goal Setting Gap Analysis Reliability Modeling System Failure Predictive Analysis (FMECA & FTA) Human Factors Analysis HALT DVT RDT V&V FRACAS RCA 5/7/2011 Ops A La Carte 76

77 Goal Setting Gap Analysis Assessment Benchmark Reliability Plan Metrics Block Diagrams Golden Nuggets CONCEPT PHASE FMEA DOE FTA Tolerance Analysis Component Selection Preventive Mainten. Predictions EOL Analysis Thermal Analysis Warranty Analysis Derating Analysis FEA POF Software Reliability DESIGN PHASE Test Plan HALT RDT ALT HALT-AFR Calculator RCA CLCA PROTOTY PE PHASE Vendor Assessmt Lessons Learned Outsourcing Warranty Returns HASS ORT OOBA Reliability Statistics Reporting Ops A La Carte EDA for Obsolesc 5/7/ MANUFACTURING PHASE

78 Next Step Execute the Reliability Program in accordance with the Program Plan Use your metrics to check how you are doing along the way Feed information from each step forward to integrate the techniques together Modify the design as needed based on your findings Release the product when you have satisfied your goals Closely monitor the product in production and in the field and feed misses back into this design and into the design process for future designs 5/7/ Ops A La Carte

79 Polling Question 6 How would you rate your product development process on a scale of 1 4? 1 (uncertainty no organized activity) 2 (awakening documented but not practiced) 3 (enlightenment implemented in key phases) 4 (wisdom implemented in all phases) I don t know 5/7/2011 Ops A La Carte 79

80 DFR & DFSS Integration 5/7/2011 Ops A La Carte 80

81 Reliability Improvement using DOE DOE techniques can be used to systematically investigate the variables (factors) that influence the life of the product, thereby providing the analyst with information that can be used to improve reliability. Noise factors can be assigned to an orthogonal array to help identify which stresses are most detrimental to lifetime response. DOE methods can be used in the production stage to estimate product reliability (e.g. warranty predictions). This can be done as a two step process in which DOE is used first to identify the control/noise factors that affect product reliability. Then the principles of accelerated life testing analysis can be used to create a model that enables predictions (under normal usage conditions) Problems: The lifetime responses are usually not normally distributed, limiting traditional analysis. Censored/incomplete data, which are not uncommon, cannot be handled by traditional means. Homogeneous variances assumption across levels of response are mostly not valid. Note pairwise matching of reliability activity and QE tools 5/7/2011 Ops A La Carte 81

82 Experimental design example DC Electric Motor Schematic windings 1 Low Resistance Failures (most common failure mode) Insulation resistance degrades over time or precipitous failure from environmental change, moisture, contamination 2 Mechanical Failures Inadequate lubrication, unbalance, vibration, and misalignment. (Gradual) 3 Overload failures Substantially more current drawn than rated load capacity. (Sudden) 5/7/2011 Ops A La Carte 82

83 DC Electric Motor Experimental Design Factor and levels Assignment to L 16 Orthogonal array with coded levels as +/ 1 5/7/2011 Ops A La Carte 83

84 Layout with time to fail response data are shown in table below Response: time to fail (hrs) Coded design variables 5/7/2011 Ops A La Carte 84

85 ANALYSIS of Electric Motor Example 2 parameter Weibull PDF is selected, as it fits the data with larger log likelihood value (smallest model fit error) The scale parameter in the Weibull PDF, above is usually selected for polynomial regression against the factors in the design: The parameters i in the general linear model are determined using regression and used to predict and in the reliability prediction. If time, cost and other resources allow, it is always recommended to conduct a follow up experiment to confirm the best settings for improved life. The gain in life from standard conditions can also be estimated. 5/7/2011 Ops A La Carte 85

86 Results of the final model with significant factor effects Failure time prediction data analysis: 5/7/2011 Ops A La Carte 86

87 DEFINE Phase (Reliability Activities) ACTIVITIES Decompose system into hierarchal functional elements (subsystems, subassemblies, components) with comparable complexity Define high level product Reliability Goals and requirements Flow down of reliability requirements from system to components Identify potential Failure modes FRACAS initiated Measurement system selection Allocation process for reliability of subsystems Benchmarking Similar Product s Reliability following decomposition Software reliability allocation Identify Reliability Software Tools (FRACAS, FMECA, ) FMEA for selected concepts Identify resources necessary for reliability engineering success Collect Voice of Customer Reliability data collection Identify potentially useful physics of failure models and update then to current situation Early predictions of future customer usage conditions Early prediction of future machine load profile Predictions of future environmental stresses. Mission defined, preliminary schedule defined, Mission statement published Human interactions effects prediction Liability management Major risks identified 5/7/2011 Ops A La Carte 87

88 MEASURE Phase (Reliability Activities) Activities Failure mechanism evaluation Identify potential measures needed for failure analysis Identify potential measures needed for CP degradation testing Identify potential measures for time to failure and tendency to fail Collect & Understand measures made during previous life test Develop operational definitions of possible functional failures Measurement system & hardware reliability, maintainability, durability for life testing Validate cost opportunity for reliability improvement Reliability data collection plan Identify Constraints for reliability testing (time, cost, samples) Safety measures identified Warranty measures identified Built in test measurement considerations defined Measures for redundancy defined Measures for prognostics defined Potential Critical parameters and critical specifications identified Downstream Manufacturing and service measurements identified considering lower cost and lower complexity Data collection Methods Selection of what to measure for reliability improvement Validation of measurements, Measurement errors/uncertainty 5/7/2011 Ops A La Carte 88

89 6 Roles & Responsibilities in Training Progressive steps Content Trainee Manager Champion Training Admin SME Pre Training Assignment Selecting Project to develop Study Material in advance SME/Mgr Engagement Class Participation Attend the course Exercises in class Post Training Register theme Project Implementation Consult & direction by SME Progress Documentation Continuing Education Proceed to advanced levels Follow up by dedicated staff Become in house trainer Advise other people 5/7/2011 Ops A La Carte 89

90 Summary Design for reliability (DFR) and Design for six sigma (DFSS) are modern methods for both problem prevention and problem solving. Both have scientific, engineering, and statistical underpinnings. Both have evolved over the past few decades from more humble beginnings. DFSS has evolved into a well architected system of management principles, analytical methodologies, and reporting systems. DFR is on a similar path. The DFR focus is on the preparation of a system or platform design to perform required functions for a specified period of time, but also to help solve problems associated with functional failures when they arise. Integration of DFSS methods with DFR methods will take some time and effort but it is inevitable. Engineering positions requiring both reliability and quality engineering skills are becoming more commonplace. This will force engineers and management to become proficient in both. 5/7/2011 Ops A La Carte 90

91 Questions? 5/7/2011 Ops A La Carte 91

92 Contact Info Thank you to everyone who attended this afternoon's webinar. We hope you found it informative. If you did not make it, make sure you go to our webinars page, where you can download the slides as a PDF and a recording of the session. Further questions may be sent to: Mikes@opsalacarte.com Loul@opsalacarte.com (408) /7/2011 Ops A La Carte 92

93 Background slides if Needed 5/7/2011 Ops A La Carte 93

94 Weibull Life Distribution frequency Time The probability density function (pdf) is a mathematical function that describes the time to fail distribution. The pdf can be represented mathematically or on a plot where the x axis represents time, as shown above. One of the most popular life distribution is the 3 parameter Weibull distribution shown below: the scale parameter, η, defines where the bulk of the distribution lies. The shape parameter, β, defines the shape of the distribution and the location parameter, γ, defines the location of the distribution in time. t is the time measure, e is exponential function 5/7/2011 Ops A La Carte 94

95 Variability & ALT Analysis Table 1: Common Acceleration Models Model Arrhenius Eyring Model Inverse Power Rule for Voltage (simplified Eyring model) Exponential Voltage Model Electromigration Model Coffin Manson Model Example Application Model failure mechanisms that depend on temperature (e.g., chemical reactions) Model multiple stresses. Many other models are simplified versions of the Eyring model. Capacitor dielectric breakdown Capacitor dielectric breakdown Semiconductor metallization line failure Mechanical crack growth 5/7/2011 Ops A La Carte 95

96 Roadmap Comparison DMAIC Define Determine Project Scope, Objectives, Resources, Constraint DMEDI Define Very Similar to DMAIC Measure Determine Customer Groups, Determine CCR s, Obtain Data to Quantify Existing Process Performance Measure Define Customers and Needs Using VOC and QFD, Determine CCR s Analyze Analyze Data to Identify Tangible Root Causes of Defects Improve Intervene in the Process to Improve Performance Control Implement a Control System to Maintain Performance over Time Explore Develop Design Concepts, and High Level Design Develop Develop and Optimize Detailed Design Implement Validate Design with Pilot, Establish Controls, Full Scale Implementation 5/7/2011 Ops A La Carte 96

97 Sample Variance & Standard deviation

98 Propagation of variance d b a y d y b y a y Sensitivity terms Taylor series expansion without covariance terms Y=f(a,b, d)

99 Six Sigma in Practice What does it mean? 99% Performance 6 Sigma Performance % Meet Spec 202 Billion Pieces of Mail Delivered by USPS per year 2.02 Billion pieces Lost per Year pieces lost per year 24/7 Power Delivery to Your Home 87 Hours without Power every Year Less than two minutes without power per year 510 Million Prescriptions Worldwide per year 5.1 Million Wrong Prescriptions per Year 1719 Wrong Prescriptions per Year 27 Billion Credit Card Transactions per year 0.27 Billion wrong transactions per year wrong transactions per Year Ref: Scott Burr, 2005, Design for Six Sigma Hubenthal Burr Associates