Lean Engineering Product Development Professor Debbie Nightingale September 25, 2002 2 Deborah Nightingale, MIT ? 2002 Lean Engineering Learning Points ? Lean applies to engineering ? Engineering requires a process ? Different from manufacturing ? Lean engineering process eliminates waste and improves cycle time ? Make sequential processes flow seamlessly ? Managing iteration to avoid unplanned rework ? Efficient and standard process enables better engineering ? Integrated Product and Process development (IPPD) is critical for lean enterprise 3 Deborah Nightingale, MIT ? 2002 Process is Important in Engineering ? For this discussion, “Engineering” is defined as preliminary and detailed design and analysis, process design, and validation and verification Concept Development System-Level Design Detail Design Testing and Refinement Production Ramp-Up From Ulrich & Eppinger, Product Design and Development, 1995 Phases of Product Development Most relevant to processes in these phases 4 Deborah Nightingale, MIT ? 2002 Lean Engineering Requires a Process ? Engineering processes often poorly defined, loosely followed (LAI Case Studies) ? 40% of design effort “pure waste” 29% “necessary waste” (LAI Workshop Survey) ? 30% of design charged time “setup and waiting” (Aero and Auto Industry Survey ) Pure Waste Value Added Necessary Waste Inspiration ? “Invention is 1% inspiration and 99% perspiration” - TA Edison ? “Product development is 1% inspiration, 30% perspiration, and 69% frustration” - HL McManus 5 Deborah Nightingale, MIT ? 2002 Application of Lean to Engineering - Traditional Womack and Jones Understand Process Eliminate Waste Radical Change ? Precisely specify value by specific product ? Identify the value stream for each product ? Make value flow without interruptions ? Let the customer pull value from the producer ? Pursue perfection 6 Deborah Nightingale, MIT ? 2002 Process enables innovation and cuts cycle time Process repeatable without errors Perfection Driven by needs of enterprise Driven by Takt time Customer pull Iterations often beneficial Iterations are wasteMake process flow Information & knowledge Parts and material Identify Value Stream Harder to see, emergent goals Visible at each step, defined goal Define Value EngineeringManufacturing Engineering & Manufacturing Have Similarities and Differences Source: Lean Aerospace Initiative 7 Deborah Nightingale, MIT ? 2002 Engineering Value is Emergent Adapted From Chase, “Value Creation in the Product Development Process”, 2001. Time Value Risk Info Activities accumulate information, eliminate risk, use resources Value Realized Process Outcome 8 Deborah Nightingale, MIT ? 2002 Program Phase % of Programs Over Cost From Hoult et al., “Cost Awareness in Design: The Role of Data Commonality”, 1995. No Database Commonality Some Best Practice Engineering Requires the Seamless Flow of Information and Knowledge 0 5 10 15 20 25 30 35 R&D Concept Def. Concept Asses Prelim. Design Detail Design Fab&test Sales O&S ? Information can be an IT problem - solutions exist, but are not easy ? Knowledge is a people problem - requires communication - this is hard! 9 Deborah Nightingale, MIT ? 2002 Communication Key to Flow and Pull ? Flow cannot be achieved until engineering processes move and communicate without errors or waiting ? 62% of tasks idle at any given time (detailed member company study) ? 50-90% task idle time found in Kaizen-type events (case studies) Task Active Task Idle ? Pull achieved when engineering cycle times are as fast or faster than the customer’s need or decision cycle 10 Deborah Nightingale, MIT ? 2002 Category % Reduction Cycle-Time Process Steps Number of Handoffs Travel Distance 75% 40% 75% 90% ? Scope: Class II , ECP Supplemental, Production Improvements, and Make-It- Work Changes Initiated by Production Requests ? Value stream simplified, made sequential/concurrent ? Single-piece flow implemented in co- located “Engineering cell” ? Priority access to resources 849 BTP packages from 7/7/99 to 1/17/00 Source: Hugh McManus, Product Development Focus Team LAI - MIT Co-Location Improves Integration 11 Deborah Nightingale, MIT ? 2002 The Seven Info-Wastes Unnecessary serial production; Excessive/custom formatting; Too many iterations 7. Processing Haste; Lack of reviews, tests, verifications; Need for information or knowledge,data delivered 6. Defective Products Late delivery of information; Delivery too early (leads to rework) 5. Waiting Lack of direct access;Reformatting 4. Unnecessary Movement Information incompatibility; Software incompatibility; Communications failure; Security issues 3. Transportation Lack of control; Too much in information; Complicated retrieval; Outdated, obsolete information 2. Inventory Creation of unnecessary data and information; Information over-dissemination; Pushing, not pulling, data 1. Over-production Source: Lean Aerospace Initiative 12 Deborah Nightingale, MIT ? 2002 Making Processes Flow System Requirements Choose Preliminary Configuration 1 ET: 8/50 days HIP: 60/457 hrs CT: 50 hrs C: $4500 V: 33 Perform Aero Analysis 3 ET: 7/50 days HIP: 42/457 hrs CT: 39 hrs C: $1075 V: 20 Create Ext & Mech Drawings 2 ET: 3/50 days HIP: 15/457 hrs CT: 12 hrs C: $475 V: 17 Determine Structural Rqmts 5 ET: 3/50 days HIP: 21/457 hrs CT: 18 hrs C: $675 V: 8 Create Structural Configuration 4 ET: 5/50 days HIP: 25/457 hrs CT: 22 hrs C: $950 V: 13 Perform Loads Analysis 8 ET: 7/50 days HIP: 41/457 hrs CT: 37 hrs C: $1525 V: 25 Perform Stability & Control Analysis 7 ET: 8/50 days HIP: 50/457 hrs CT: 45 hrs C: $4100 V: 18 Perform Weight Analysis 6 ET: 4/50 days HIP: 23/457 hrs CT: 20 hrs C: $1325 V: 19 Create Manufacturing Plan 11 ET: 12/50 days HIP: 79/457 hrs CT: 59 hrs C: $2225 V: 34 Perform SS&L Analysis 10 ET: 5/50 days HIP: 43/457 hrs CT: 38 hrs C: $2975 V: 38 Develop Finite Element Model 9 ET: 4/50 days HIP: 21/457 hrs CT: 19 hrs C: $1350 V: 22 Develop Design Report/ Pres. 12 ET: 4/50 days HIP: 37/457 hrs CT: 30 hrs C: $2225 V: 35 Mg t R e v i e w , F o rm a ttin g F o rm a ttin g Mf g Re v i e w Eng Revi ew Mg t Re v i e w major value tasks From Millard, “Product Development Value Stream Analysis and Mapping”, 2001 ? Value Stream Mapping and Analysis required for understanding ? Process mapping and Design Structure Matrix methods most powerful for process improvement ? Process mapping customized for PD developed 13 Deborah Nightingale, MIT ? 2002 Results: Engineering Release Process ? Reduced Cycle time by 73% ? Reduced Rework of Released Engr. from 66% to <3% ? Reduced Number of Signatures 63% Traditional Lean Cycle Time Std Dev Time ? Value stream mapped and bottlenecks found ? Process rearranged for sequential flow ? Waiting and delays removed Source: Lean Aerospace Initiative 14 Deborah Nightingale, MIT ? 2002 Complexity may Require Iteration ? Engineering release process prior state New Requirement Schedule Review PR’s Write EDA Basic Layout FAMSCO Write PS Assign Task Detailed Layout Layout from Config STRESS Assy Drawing Detail Drawing CHECKBOARD RELEASE CENTER SIGNOFF NEAR DCCInvestigate C/A Board C/A 15 Deborah Nightingale, MIT ? 2002 Complex Engineering Processes Require Efficient Iterations AND Flow ? Understand how iterations reduce risk and/or handle emergent knowledge ? Don’t set up iterations that have large time lags that can cause unnecessary rework ? Within an iteration and between iterations make information flow efficiently ? Answer may be faster and more efficient iterations, not necessarily fewer ones 16 Deborah Nightingale, MIT ? 2002 Manage Iteration Sequential Process Discovery Emergent knowledge Complex process Rote work Held knowledge Simple process Choose Approach Balance Factors Make Simple Processes Sequential; Manage Iteration of Complex Ones 17 Deborah Nightingale, MIT ? 2002 Key Learnings ? Engineering process is important ? Efficiently execute “the fundamentals” ? Remove waste and improve cycle time ? Iterations are not necessarily waste ? When needed (and managed) add knowledge effectively and avoid unnecessary rework Good process is key to effective engineering so LEAN APPLIES! 18 Deborah Nightingale, MIT ? 2002 Integrated Product and Process Development (IPPD) A management technique that simultaneously integrates all essential acquisition activities through the use of multidisciplinary teams to optimize the design, manufacturing, and supportability of processes. 19 Deborah Nightingale, MIT ? 2002 Integrated Product and Process Development (IPPD) IPPD facilitates meeting cost and performance objectives from product concept through production, including field support. One of the key tenets is multidisciplinary teamwork through IPTs. Conceptualization and Design Test and Production Sustainment C o s t o f C h a n g e High High Low Low Number of Design C h anges Dollars Traditional IPPD Traditional vs IPPD Approach Deborah Nightingale, MIT ? 2002 21 Deborah Nightingale, MIT ? 2002 IPPD Key Tenets ? Customer Focus ? Concurrent Development of Products and Processes ? Early and Continuous Life Cycle Planning ? Maximize Flexibility for Optimization and Use of Contractor Approaches ? Encourage Robust Design and Improved Process Capability 22 Deborah Nightingale, MIT ? 2002 IPPD Key Tenets ? Event-Driven Scheduling ? Multidisciplinary Teamwork ? Empowerment ? Seamless Management Tools ? Proactive Identification and Management of Risk 23 Deborah Nightingale, MIT ? 2002 Benefits of IPPD ? Reduced overall time for product delivery. ? Reduced system (product) cost. ? Reduced risk. ? Improved quality. ? Improved response to customer needs. 24 Deborah Nightingale, MIT ? 2002 Integrated Product Team ? Build successful programs ? Identify and resolve issues ? Make sound, timely decisions Working together to: TEAM Team Leader FUNCTIONAL REPS * Program Mgmt * Engineering * Manufacturing * Logistics * Test & Eval ?Contracting ?Suppliers * User (All APPROPRIATE Areas) 25 Deborah Nightingale, MIT ? 2002 Multi-Program Enterprise Impacts ? Research examples where time/cost delays due to infrastructure issues beyond the specific program ? Access and availability of enterprise resources ? Space system testing example ? Use of commonality to support operations not just design 26 Deborah Nightingale, MIT ? 2002 Analysis of Spacecraft Test Discrepancies Over 23,000 discrepancies from over 20 programs, encompassing over 225 spacecraft 0 5 10 15 20 25 30 35 Employee- Operator Design Material Equipment Software No Anomaly Unknown Other Root Cause Category Communications Missions Other Missions Mean Confidence In terval ?Median On a per spacecraft basis almost 50% of discrepancies are caused by workforce and equipment issues common to many programs Source: LAI Product Development Team