Reliability Modeling and Optimization of New Product Development - PowerPoint PPT Presentation

Reliability Modeling and Optimization of New Product Development Process Mohammad Sadegh Mobin PhD Candidate in Engineering Management Department of Industrial Engineering and Engineering Management Western New England University Springfield, MA PhD advisor: Dr. Zhaojun (Steven) Li 2017 mobin.sadegh@gmail.com

Overview Reliability Modeling and Optimization of New Product Development Process  Part 1: Reliability Growth Planning (RGP) Modeling and Optimization  Part 2: Verification and Validation (V&V) Activities Planning and Optimization 1 mobin.sadegh@gmail.com

New Product Development (NPD) Consumer needs changes Changing Marketing technology environment changes Continuous Product Development High % of total revenue Not to lose market share To stay ahead of Competition 2 mobin.sadegh@gmail.com

New Product Development (NPD) Planning Product/Process design & development Production Product/Process V&V Business case Prototype/pilot (new idea) Production (Build components/ Concept design system) phase (System requirement Field Prototype test identification) performance (Test components/ Detail design System) Verification (Component requirement &Validation identification) (System and process V&V) 3 mobin.sadegh@gmail.com

NPD Challenges NPD programs are often plagued with: Cost overruns, Schedule delays, and Quality issues . Product Company Issues Year Source 787 The Wall Street Boeing Co Delay due to a structural flaw 2009 Dreamliner Journal Chevy Volt General Motors Cost overrun during design 2009 CNN Money The Design issues: An unanticipated test Honda/GE program glitch. A part of the gearbox HF120 Honda 2013 Flying failed during the test. Rebuild the turbofan engine and begin the test again. engine Delays in delivering engines. Quality United Technologies Defence- flaws and technical issues. Systemic F-35 Corp.’s Pratt and Whitney 2014 aerospace.com issues and manufacturing quality unit Bloomberg Business escapes. A failure in the main gear box and need US Marine Corps' for redesign of the component. Sikorsky 2015 HIS Jane’s 360 (USMC's) Problems with wiring and hydraulics systems. Budget constraints. 4 mobin.sadegh@gmail.com

NPD Challenges NASA ’s main projects that faced The Department of Defense (DoD) The set of 96 major new weapon system cost and time overrun: development programs (2000-2010) have:  The International Space Station.  an average development cost growth of Prime contract had grown: 25% 42% , (from $783M to $986M, the 3 rd increase in 2  an average delay of 22 months. years) . “ 50% of the DOD’s NPD programs faced cost  The NASA Ares-I launch system. overrun ” . Cost overrun: 43% “ 80% experienced an increase in unit costs from (from $28 billion original estimate to $40 billion ) initial estimates ” . 5 mobin.sadegh@gmail.com

Reliability Management Process 6 mobin.sadegh@gmail.com

Objective Model and optimize the NPD reliability process in terms of cost, time, and product reliability  Proposing a model to improve the reliability growth planning (RGP).  Providing a quantitative model to improve product V&V activities planning. Improve the NPD process by reducing:  Time to market delay (Scheduling)  Cost overrun (Budgeting)  Quality flaws (Reliable product) 7 mobin.sadegh@gmail.com

Part1: RGP Multi-Objective and Multi-Stage Reliability Growth Planning (MO-MS-RGP) mobin.sadegh@gmail.com

Reliability Growth Planning (RGP) A single stage Reliability Growth Plan 1000 One RG plan can be:  390 Days, MTBF (Days) Reliability goal = 73 100.0  Required test units  Required test time  Total cost: $50k Achieved MTBF: Another RGP can be: 10.00 70.0 days  490 Days, Initial MTBF:  Test units and test time 3 days  Total cost $70k 1.000 390 Days 490 Days Test Time Objective of RGP: To determine the number of test units, test time, and cost to maximize the reliability growth. 8 mobin.sadegh@gmail.com

Duane Reliability Growth Model  Duane Model (1964) o An empirical model, based on the learning curve , o Also known as power law model o Duane model in terms of cumulative failure rate: 𝒎𝒐 𝑫 𝒖 = 𝜺 − 𝜷 𝒎𝒐 𝒖 𝐷 𝑢 : The average failure rate 𝐷 𝑢 = 𝑂(𝑢)/𝑢 𝑂 𝑢 : The cumulative number of failures up to time 𝑢 during the reliability growth testing. 𝜀, 𝛽 > 0 , 𝛽 is known as growth rate 9 mobin.sadegh@gmail.com

Motivation for Multi-Stage RGP An example of multi-stage NPD plan 10 mobin.sadegh@gmail.com

Multi-Stage Reliability Growth Planning The schematic of multi-stage reliability growth planning  Challenges for multi-stage RGP in early product development stage: 1. How to allocate test units and time to individual stage. 2. How to determine the proportion of new technology introduction in each stage 11 mobin.sadegh@gmail.com

RGP Literature Review Single Multi Single - Multi- Author objective objective stage stage Yes No Yes No Duane (1964) [1] Yes No Yes No Crow (1974) [2] Lloyd (1986) [3] Yes No Yes No Robinson and Dietrich (1987) [4] Yes No Yes No Coit (1998) [5] Yes No Yes No Walls & Quigley (1999) [6] Yes No Yes No Walls & Quigley (2001) [7] Yes No Yes No Quigley and Walls (2003) [8] Yes No Yes No No No Krasich et al. (2004) [9] Yes Yes No No Johnston et al. (2006) [10] Yes Yes No Jin and Wang. (2009) [11] No Yes Yes Jin et al. (2010) [12] Yes No Yes No Jin et al. (2013) [13] Yes No Yes No Jin and Li (2016) [14] Yes No Yes No Jackson (2016) [15] Yes No Yes No Li et al (2016) [16] No Yes No Yes Significant contribution: Multi-objective & Multi-stage RGP 12 mobin.sadegh@gmail.com

Proposed MO-MS-RGP Model 1. Minimize failure rate at the final stage 2. Minimize total development time Objectives : 3. Minimize total test cost Stage 2 Stage 3 Stage 1 1- Total product test Reliability Goal MTBF at the time end of stage 3 MTBF at the Constraints: end of stage 2 MTBF 2- Number of available test units in each MTBF at the end of stage 1 development stage Initial MTBF Initial Initial (stage 3) MTBF MTBF (stage 2) (stage 1) 1- Number of test units for each Test Test Total time Decision time for time for subsystem in each stage variables: stage 1 stage 2 2- Test time for each subsystem in each stage 13 mobin.sadegh@gmail.com

MS-MO-RGP Mathematical Modeling mobin.sadegh@gmail.com

Proposed MO-MS-RGP Model Min: 𝜇 𝑜 = 𝑔 𝜇 𝑗−1 , 𝜇 𝑜(𝑗) , 𝛽 𝑗 , 𝑈 𝑗 𝑜 Min: 𝜐 = 𝑗=1 𝜐 𝑗 , 𝑗 = 1, … , 𝑜 𝑜 Min: 𝐷 = 𝑗=1 𝐷 𝑗 , 𝑗 = 1, … , 𝑜 Decision variables: 0 ≤ 𝜐 ≤ 𝜐 𝑣 s.t. • 𝑜 𝑗𝑘 𝑢 𝑗𝑘 • 𝑂 𝑚(𝑗) ≤ 𝑂 𝑗 ≤ 𝑂 𝑣(𝑗) , 𝑗 = 1, … , 𝑜 𝒈 𝟒 * Time (Yrs.) An optimal solution * Cost ($) (RGP) * Reliability (MTBF (YRS.)) * Number of test units for each sub-system in each stage * Planned testing time for each sub-system in stage 𝒈 𝟑 𝒈 𝟐 15 mobin.sadegh@gmail.com

Solution Methodologies Method 1: Creating a weighted composite objective function Shortcomings: 1. Difficulties in determining appropriate utility functions (weights). 2. Objectives have different scale and cannot easily be added up. Method 2: Consider one as main objective function and others as constraints Shortcomings: 1. Difficulties in determining boundary values. 2. Defining boundaries may reduce the solution space. Method 3: Multi-Objective Evolutionary Algorithms (MOEAs) e.g., MOPSO, NSGA, etc. 1. Simultaneously optimizing two or three (or more) conflicting objectives. 2. Effective methods in exploring feasible solutions and providing a population of approximately optimal solutions (Pareto-optimal frontier). 3. Apply evolutionary operators, e.g., crossover and mutation to generate variety of new solutions. 16 mobin.sadegh@gmail.com

Proposed Solution Methodology Overview of the proposed solution methodology Multiple Mathematical model: A set of Pareto- Objectives • Objective functions optimal solutions Evolutionary • Constraints • Algorithm Decision variables Inputs Optimal Data (Minimization objective functions) efficient Envelopment solutions Analysis (DEA) Outputs (Maximization objective functions) 17 mobin.sadegh@gmail.com