Impacting Design Quality through Key Parameter Development & Management

1. Impacting Design Quality through Key Parameter Development & Management Using KPD&M during Technology & Product Development Processes to Prevent Design Problems

2. Key parameters control financial consequences… through Y & s Physical Law… Y= f(X) DY= f[DX+ D(X*N)] + error XY Xs Cp = (USL-LSL )/6s s Cpkl = (Y-LSL)/3s XY&s XNoise Cpku = (USL-Y)/3s T Y $L(Y) = k[s2 + (Y-T)2] …Financial Consequences Intro to KPD&M, Copyright 2010, PDSS Inc.

3. What does the word “Key” mean? Something that is… • New • Totally new to you & all your competitors, no one has fulfilled the requirement(s) or controlled the parameter(s) before – no experience! • Unique • The requirement(s) or parameter(s) have been fulfilled or controlled by others but not by you! • Difficult • The requirement(s) or parameter(s) are extreme & their fulfillment or control is very high in risk Intro to KPD&M, Copyright 2010, PDSS Inc.

4. Things that are NOT “Key”… Something that is… • Easy • Common • Old These are functions, part specifications & mfg. functions that we place under normal Q.C. metrics - Little or no SPC investment (low need to detect & prevent) - Cp & Cpk checked periodically - Use Six Sigma to react to problems in this area Intro to KPD&M, Copyright 2010, PDSS Inc.

5. Refining the term - Key • A function, part or material characteristic can be designated as Key = Under Watch! • Functions occur in the product or process as it is transforming mass & / or energy… it is what the product or process does. • Inherent in the design of the product or mfg. process • Characteristics are static dimensions, shape factors, surface finishes or bulk material properties • Key doesn’t just mean it is important! • It means there is high risk because… • unproven – we lack facts, little or no data - we don’t know! • unstable & must be “watched”! • dependent on different Design or Supplier’s capabilities Intro to KPD&M, Copyright 2010, PDSS Inc.

6. Key Parameters are like slippery bars of soap! – risk of “getting out of hand”! Risk! Cannot afford to call everything that is merely important a Key Parameter $$$ Intro to KPD&M, Copyright 2010, PDSS Inc.

7. The road to being designated Key… 1. Vary an X & measure the effect on Y…. DY/DX 2. Do so repeatedly & measure the variation around DY caused by each DXi = randomerror = e 3. Define the ratio between the Signal (DY/DXi)& the Noise e … this is called the F Ratio = Strength of each DXi on DY when compared to random noise in the replicated data 4. Establish if each Xi’s effect on Y is statistically significant… calculate the p value 5. Establish the Capability Index for Ys & Xs… calculate the Cp & Cpk values for Y & X under nominal (Cp) & stressful (Cpk) conditions! Intro to KPD&M, Copyright 2010, PDSS Inc.

8. Stability, F Ratio, p Value, Robustness, Tunability & Capability tell the Story! • If Xs & Ys are: • statistically significant…. Low p values (< 0.05) • have high F Ratios (>> 4) from Analysis of Variance (ANOVA) • possess unstable behavior (SPC trend & control issues) • have low Cpunder nominal conditions… then they are extremely risky & are designated as Keys! • These are our highest priority Keys to work on. • If these same Xs & Ys • possess high sensitivity to stressful noises after Robust Design • difficult to tune onto the desired target after Robust Design • have low Cpkunder stressful conditions… then they are still very risky & are designated as Keys! Intro to KPD&M, Copyright 2010, PDSS Inc.

9. Key >>> NUD!Guilty until proven Innocent! • Key parameters are under suspicion– we don’t trust them! • measured & watched for drift in mean • measured & watched for changes in s • A parameter or characteristic can come offour list of Keys…. Re-designated as ECO! • Proven stability over time (SPC Charting) • Ease of control under nominal & stressful conditions (Robust & Tunable performance) • Sustained capability (Cp/Cpk) as cost is reduced X & R Charts Intro to KPD&M, Copyright 2010, PDSS Inc.

10. What is Key Parameter Development & Management? Documenting a hierarchy of: Key requirements & the integrated set of measured functions, specifications & set points - down through a product architecture and its production & support processes. A proactiveprocess for: • Identifying • Connecting • Tracking • Refining • Preventing problems Intro to KPD&MM, Copyright 2010 PDSS Inc.

11. A bit of history…. • From Dogma & Faith… • Dogma = Cheaper & Faster – We must Hurry!!! • Faith = We You will make it! • To Doubt & Experimentation… • Doubt = We have risk & uncertainty – we need facts! • Experimentation = We can & will take the time to learn! X Learning generates facts which reduce uncertainty & lower risk… …but not by rushing & cutting corners. Intro to KPD&M, Copyright 2010, PDSS Inc.

12. Some personal history… • Dogma & Faith results circa. 1972-1994 compared to Doubt & Experimentation results from 1995-1999 at Kodak… • A focused effort was implemented in Kodak’s digital printing business unit on the DigiMaster Project: • Our 1st ever Commercial Systems Engineering Org., Center of Excellence & SE Process • Comprehensive integration of Key Parameter Development & Mgt. approach into Phase-Gate PDP • Clear definition of “Key” Customer needs • Heavy emphasis on Reliability Development using Robust Design • Detailed “Design for X” focus on produceability & serviceability • Strong Project Manager, rigorous PM methods & dedicated, accountable PDT (functional Centers of Excellence supported it) Intro to KPD&M, Copyright 2010, PDSS Inc.

13. The DigiMaster 9100 digital printing system Over 10,000 parts < 30 major subsystems (chemo-opto-mechatronics) Req’d 6s image quality across 15 measurable attributes Req’d 150K MIBSC within 95% Confidence Limits

14. Historic example of Prod. Dev. Team performance before the use of SE & KP enabled work flow… Pre-SE, KPD, DFSS, etc. Targets Change over to supplier mtl.s & parts Independent SS Changes to improve System Performance Late integration of tweeked SSs & purchased accessories System integration too early… MTBF 50-60% of Target Gate 4 Gate 5 Gate 2 Gate 3 Time

15. Development Teams can improve results with SE, Key Parameter Development & select DFSS tools… SE/KPD/DFSS Actual +2s UCL Pre-SE, KPD, DFSS, etc. Target >95% of Target -2s LCL MTBF 50-60% of Target Gate 4 Gate 5 Gate 2 Gate 3 Time

16. Duane plot from an actual project

17. Macro-effects Created a formal systems engineering organization - clear SE roles, with SE tools, tasks & deliverables tied to Gate Requirements measured with performance score cards – not checklists Enhanced the SE team to actively use specific KP tasks with DFSS tools to complete the tasks, directly assisting sub-teams – producing the right SE Gate deliverables Micro-effects Key Parameter Mgt.: clear definition of Key reqts. flow down & rigorous measurement of capability flow-up (Cp & Cpk trace-ability) System Integration, system sensitivity analysis & reliability testing only AFTER subsystem & subassy. robustness optimization was completed reliability development vs. assessment What made the difference? Complete story is in Ch. 7: Systems Architecting, Engineering & Integration using DFSS & Key Parameter Development

18. What does System Development look like as a flow of work over time? The Super-set of System Engineering Macro- Functions: Architecting Engineering Integration Assessment & Validation System Performance Balancing Subsystem Interface Development & Robustness Optimization Internal & External Needs System Modeling System Functions System KPM Database Transfer to Production, Service & Tech. Support System Integration & Stress Testing System Internal & External Validation System Architecture System Reqts. Flow of System Architecting, Engineering, Integration & Assessment Tasks

19. Process Map of Major System Architecting, Engineering, Integration & Assessment Tasks Define System Reqts Define System Functions Define System Architecture Partition System into Subsystems Create & build KPM Database Generate System FMEA Lead System Integration Meetings Develop System Noise Map Balance Interface Sensitivities – create latitude Define System Integration DOEs &Test Plans Integrate System Test Rigs & Data Acq. System Conduct System Integration Stress Tests Balance System Performance Conduct Reliability Assessments Validate System Performance Transfer KPM Database to Mfg. & Support

20. Key Parameter Enabled Systems & Design Engineering: Key Parameter Dev. process & enabling DFSS tools Design Optimize Concept Verify Key Parameter Management Process Full KPD&M details: Ch.s 8-13of DFSS text Requirements Development Process Concept Design Process Sequential Design of Experiments Process Reliability Definition, Modeling, Development & Assessment Process Design for “X” Process - Manufacturing, Assembly & Cost; Service Maint. & Support - Environment, Health, Safety, Legal & Regulatory

21. Allocated Reqt.s Flow-down & Measured Capability Roll-up VOC Needs Customer Satisfaction Verification & Preventive / Contingent Action Process Product Reqts. Product CFR Cp & Cpk Subsystem Reqts. Subsystem CFR Cp & Cpk Subassembly Reqts. Subassembly CFR Cp & Cpk Component Reqts. Component Spec. Cp & Cpk Flow down of the reqts to be fulfilled through the measurement of KFRs & KPs Roll-up of Cp & Cpk through the measurement of KFRs & KPs Mfg. Process Reqts. Mfg. Process Cp & Cpk

22. Requirements Development Process …Flow-down of NUD / Kano requirements to be fulfilled VOC Needs Enabling Tools. Methods & Best Practices: • Customer Interviewing • KJ Analysis • NUD Screening & Kano Analysis • QFD • Requirements Trace-ability & Documentation (DOORS, etc.) Product Reqts. Subsystem Reqts. Subassembly Reqts. Component Reqts. Mfg. Process Reqts.

23. System Concept Design Process Step 1: External Needs Gathering, Processing & Validating the Voices of the Customer, Marketing, Technology & Business Step 2: Internal Requirements & Constraints Generating & documenting a system of NUD / Kano requirements in a Key Parameter Mgt. data base Step 3: Innovation, Architecting & Solutions Concept Generation, Feasibility Evaluations & final Concept Selection

24. Metrics for Requirements Can be Compared to Measures of Sample Data What is Required? Customer Level (USL – LSL) System Level (USL – LSL) Subsystem Level (USL – LSL) Subassembly Level (USL – LSL) Component Level (USL – LSL) Mfg. Process Level (USL – LSL) What is Measured? Customer Level (Avg & σ) System Level (Avg & σ) Subsystem Level (Avg & σ) Subassembly Level (Avg & σ) Component Level (Avg & σ) Mfg. Process Level(Avg & σ) From this comparison we can document performance Capability Intro to KPD&M, Copyright 2010, PDSS Inc.

25. Reqt. Allocation & KP Measuring down through the System to Subsystems, Sub Assemblies, Parts & Mfg. Processes! Key Reqt.s Allocation & Linkage Product or System Level Sub System Level Sub Assy Level Part Level Capability Assessment & Traceability Mfg. Level Intro to KPD&M, Copyright 2010, PDSS Inc.

26. sKFR Product Functional Capability • (USL-LSL): tolerance rangefor a KFR response within the product (Sys/SSys/SAys) • (USL-LSL): as stated in the Reqts. Document • 6s = six times the sample std. dev. of a Key Functional Response KFR in the design • “s” measures functional variation • “s” is composed of both mfg. and customer-base variation in product usage and environments Intro to KPD&M, Copyright 2010, PDSS Inc.

27. sKTF Part Specification Capability • (USL-LSL): tolerance rangefor aKTF spec. on a component / assembly • (USL-LSL): directly traceable to both Product &ManufacturingKFRs • 6s = six times the sample std. dev. of a KTF Part specification • “s” measures dimensional, surface finish, bulk material propertyor material variation • “s” is composed only ofunit-to-unit Part variation Intro to KPD&M, Copyright 2010, PDSS Inc.

28. sKFR ManufacturingProcess Capability • (USL-LSL): tolerance rangefor aKFR spec. on a production machine • (USL-LSL): directly traceable up to PartKTF Spec. • 6s = six times the sample std. dev. of a KFR specification • “s” measures Process functional variation • “s” is composed only offunctionalmfg. variation Intro to KPD&M, Copyright 2010, PDSS Inc.

29. Required KP Mgt. Data for any form of Capability Assessment Gage R&R • All KFRs, KPs or KTF Spec.s must have a capable metrology process documented & in use • Each KFR, KP or KTF Spec. is placed under SPC so the Cp can be routinely quantified for Phase-by-Phase growth & Life Cycle stability characterization • All KFRs typically have a target of Cp = 2 & Cpk of 1.5 I & MR Chart Capability Study Intro to KPD&M, Copyright 2010, PDSS Inc.

30. NUD VOC Need #1 NUD VOC Need #2 System Reqt. System Reqt. System Reqt. Y=System KFR Y=System KFR Y=System KFR Subsystem-to-System Level Transfer Functions Y = f(x1, x2, …xn) Su bsystem Reqt. Subsystem Reqt. Subsystem Reqt. Y =SS KFR Y =SS KFR Y =SS KFR 1 1 1 Subsystem Reqt. Subsystem Reqt. Subsystem Reqt. Y =SS KFR Y =SS KFR Y =SS KFR 2 2 2 Subsystem Reqt. Subsystem Reqt. Y =SS KFR Y =SS KFR 3 3 Subassy-to-Subsystem Level Transfer Functions: Y = f(x1, x2, … xn) Subassy. Reqt. Subassy. Reqt. Subassy. Reqt. X =SAssy. KFR X =SAssy. KFR X =SAssy . KFR 1 1 1 Subassy. Reqt. X =SAssy. KFR 2 Component Reqt.s Component Reqt.s Component Reqt.s X =Comp. KTF Spec.s X =Comp. KTF Spec.s X =Comp. KTF Spec.s n n n Component Reqt.s X =Comp. KTF Spec.s n Mfg. Process Reqt.s Mfg. Process Reqt.s Mfg. Process Reqt.s X =Mfg.. KTF Spec.s X =Mfg.. KTF Spec.s X =Mfg.. KTF Spec.s n n n KPD&M Flow-Down Map Intro to KPD&M, Copyright 2010, PDSS Inc.

31. Modeling & Simulation • M&S was in place & was pretty good - but it left KP knowledge gaps - & not just a few!!! • Could not predict physics-based interactions between controllable engineering parameters very well… Xi * Xj = ??? • Could not predict physics-based interactions between controllable engineering parameters AND NOISE PARAMETERS = unwanted sources of variation… Xi * Noise = ??? • From variation in production parts, assembly & materials • From variation in disruptive sources external to the system • From variation in deteriorative sources internal to the system • Weibull, Exponential, Gamma, Rayliegh, Lognormal, Normal, etc.???? Intro to KPD&M, Copyright 2010, PDSS Inc.

32. 2 Major Matrices dominate the KP Dev. Process! On the Parameters Side: • The Designed Experiment (DOE) • NUD Transfer Functions (Key Y = f(Xs)) measured, Ranked & Prioritized On the Requirements Side: • The Houses of Quality from NUD-based QFD • Translated, Ranked, Prioritized & Allocated Key Customer Needs Intro to KPD&M, Copyright 2010, PDSS Inc.

33. Sequential Designed Experiments Process Concept Design Optimize Verify Tolerance Balancing DOEs Multi-vari Studies Screening DOEs System Stress Test DOEs Modeling DOEs Robust Design DOEs …Iterate… Optimization DOEs Building your knowledge of statistically significant Key Parameters using a sequential DOE strategy

34. DOE choices in Product Commercialization There are 7 major types of Designed Experiments 1. Multi-vari studies - (correlation & hypothesis forming studies) 2. Screening Experiments - (sorting controllable factors & noise factors for significance) 3. Modeling Experiments - (quantifying Y = f(x) relationships) 4. Mean Optimization Experiments - (adjust mean performance to hit a desired target) 5. Robustness-to-Noise Experiments - (reduce s in the presence of noise) 6. System Stress Testing Experiments - (identify sensitivity across interfaces & system boundaries) 7. Tolerance Balancing Experiments - (refine cost vs. quality in subsystems, subassemblies & parts) “Everything should be as simple as possible – but not simpler…”

35. Identifying Key Functional Response & Key Adjustment Parameter Relationships Key Functional Response (KFR) Ideal Key Adjustment Parameter (KAP) Intro to KPD&M, Copyright 2010, PDSS Inc.

36. Key Functional Robustness Parameters:KAPs & KFRPs: How they affect a KFR - Robust & Tunable Performance! KFRPs are KFR Variance Reducers KAPs are KFR Mean Shifters Robust against Variation Mean Adjusted to VOC Target Intro to KPD&M, Copyright 2010, PDSS Inc.

37. Reliability Development Process Design Optimize Concept Verify Reliability Requirements Definition -System -Subsystem / Subassembly - Component Reliability Modeling – Probabilistic Simulations Reliability Development Tasks - FMEAs, CAE/CARD, DOE, Robust Design, Tolerance Design Reliability Assessment Tasks • Life Tests, Accelerated Life Tests • HALT, HASS, HAST, Destructive Tests

38. Design for “X” Process Design Optimize Concept Verify DfX Requirements -System (Product & Production Processes) -Subsystem / Subassembly - Component / Materials Design for X Tasks - Benchmarking, DFMA, Design for Cost, VA/VE… DfX Assessment Tasks • HSER DOEs & Related Tests

39. KPD enhanced Team Performance Score Cards Preventive Peer Reviews Contingent Design Reviews Reactive Gate Reviews Measuring the use of tools, completion of tasks and the fulfillment of Gate Deliverable requirements…

40. Gate Deliverable Scoring linkage from Tool & Task Scorecards Tool Scoring Items Data Integrity Quality of Tool Use Tool Results vs. Task Reqts Task Scoring Items % Task Fulfillment Task Results vs. Gate Reqts Avg. Tool Score Gate Deliverable Scoring Items Risk Accrual against Gate Reqts. Confidence in Data Score

41. Summary – building KP Dev. capability & maturity • Companies who have deployed KPD&M are slowly realizing they can’t use it right if they don’t have SE functional excellence & governance in their Phase-Gate process… • With KPD&M integrated into SE the results are much better • Ad hoc systems work in product commercialization processes keeps you from being great… KP enhanced SE Process, Roles, tool-task-deliverables… SE capability maturity Formal SE Process & Roles Ad hoc SE

42. The ARDEC Story:Defining a process for Pro-active KPD&M • What approaches are available for conducting KP Development & Management? • Are the steps during Development different from those conducted when defining KPs after Launch? • Technology & Product Development? • Post-launch Production & Ongoing Life cycle Management out to Discontinuance? Intro to KPD&M, Copyright 2010, PDSS Inc.

43. A New Technology & Product Development Process was constructed: Vector Similar to the definition of a Vector, the ARDEC T&PDP will serve as a course or compass direction for navigating ARDEC IPTs through technology and product development projects doing the right things at the right time. Webster - Vector: a quantity that has magnitude and direction and that is commonly represented by a directed line segment whose length represents the magnitude and whose orientation in space represents the direction; b:a course or compass direction c: a course to be taken by an aircraft. ARDEC ‘s Technology & Product Development Process (T&PDP) = Vector Intro to KPD&M, Copyright 2010, PDSS Inc.

44. Foundations of Vector • Vector is built upon a wide variety of benchmarks that were “value-mined”… • 8 major Corporations: • 6 texts from product development consulting firms: • NASA / DoD TRL models • Latest version of the DoD 5000.2 Ford Intro to KPD&M, Copyright 2010, PDSS Inc.

45. Best elements integrated to design the T&PDP process…. Benchmarks VOC NUD Reqts. Benchmarks, Hybridization and Pugh Concept Selection Process used to document Value Selection - led to the design of Vector Block Diagrams Intro to KPD&M, Copyright 2010, PDSS Inc.

46. Block Diagrams: Defining What to do…. • Results 3. Major Activities • Tasks 5. Deliverables • Readiness 1. Entrance Criteria • Completeness 6. Exit Criteria • Intent 2. Objectives • Enablers 4. Enabling Best Practices Intro to KPD&M, Copyright 2010, PDSS Inc.

47. And when to do it…. The Vector Process is constructed of Blocks of Major Activities.. Vector Technology Dev. Process… 9 Blocks of Major Activity Groups defined & documented: Vector EMD Process… 10 Blocks of Major Activity Groups defined & documented: Intro to KPD&M, Copyright 2010, PDSS Inc.

48. Each Block contains a designed Work Flow… adaptable to the type of Project Block of Major Activities Block of Major Activities Block of Major Activities Block of Major Activities Block of Major Activities MS Project Network Diagrams will illustrate serial / parallel flow paths of Major Activities within each Block… Including linkage between the Actions & their enabling Tool sets. Intro to KPD&M, Copyright 2010, PDSS Inc.

49. Aligning the Blocks to TRLs & MRLs – Vector added KP depth-of -rigor & clarity of the TRL / MRL definitions & detailed deliverables EMD 10 TD5-6 TD3-4 TD 7 TD 8 TD3-4 TD5 TD6-7 TD 8 EMD 10 EMD 10 EMD 9 Intro to KPD&M, Copyright 2010, PDSS Inc.

50. Technology Dev. Phases & Gates were defined from the 9 TD Block Diagrams…. Phase 2: Technology Concept Dev. Phase 1: Technology Project Plan & Requirements Dev. Phase 3: Technology Sub-level Dev. & Optimization Phase 4: Technology Integration & Final Optimization Intro to KPD&M, Copyright 2010, PDSS Inc.