Application of Project Simulation to Optimize Ares I X Upper Stage Simulator Project Implementation

Application of Project Simulation to Optimize Ares I–X Upper Stage Simulator Project Implementation APPEL PM Challenge Conference February 2008 Vince Bilardo NASA Glenn Research Center Chuck Sander ePM George Culver SAIC Ares I-X Mission – First Ares I Test Flight! Glenn Research Center CM/LAS Simulator USS to CM\LAS Interface Spacecraft Adapter and Service Module Simulators Upper Stage Upper Stage Simulator RoCS USS to RoCS Interface Frustum USS to FS Interface Avionics (internal) Forward Skirts Simulated fifth segment FS Four-segment motor PMC 2008 Ares I-X Upper Stage Simulator Team Glenn Research Center PMC 2008 Ares I-X Upper Stage Simulator Team 3 Purpose of Analysis Glenn Research Center 1. SAIC and ePM were tasked to investigate the Upper Stage Simulator manufacturing processes. a. ePM looked organizationally across entire manufacturing process. b. SAIC focused on manufacturing process details in GRC’s Ares Manufacturing Facility (AMF). 2. Following approach was used by both teams to conduct analysis: a. Solicited and received process flow and resource use data from customer (NASA USS Project Team). b. Used Simulation techniques to model process execution. 1) ePM used organizational simulation techniques with NASA SimVision® . 2) SAIC used SIMUL8 Discrete Event Simulation (DES) software. c. Iteratively revised models as process matured. 1) NASA SimVision informing the DES model and vice versa. 2) Modeling workshops informed the project team and the modelers. PMC 2008 Ares I-X Upper Stage Simulator Team 4 DES Model of USS Manufacturing Process Glenn Research Center Segment manufacturing process modeled with SIMUL8 DES software. Resources Flange Mfg Processes Segment Build Processes Skin Mfg Processes Painting and Secdry Struct Install PMC 2008 Ares I-X Upper Stage Simulator Team 5 Results: Sensitivity to Number of Workers Glenn Research Center 1. Process is highly sensitive to changes in number of both fabricators and welders. 2. To meet deadline with minimum personnel, data suggests good baseline with: a. 8 Fabricators per shift b. 6 Welders per shift Fabricators 10 8 6 10 8 6 10 8 6 Welders Charge Duration (days) 56.6 57.9 61.7 58.4 59.6 64.8 61.4 62.4 68.0 6 6 6 5 5 5 4 4 4 Each Charge duration reflects average of 50 independent trials PMC 2008 Ares I-X Upper Stage Simulator Team 6 Investigation of Non-Destructive Inspection Glenn Research Center Mandatory Inspection Points (MIPs) highlighted. PMC 2008 Ares I-X Upper Stage Simulator Team 7 Results: NDE Inspection Sensitivities Glenn Research Center 1. This sensitivity analysis assesses NDE defect rate, NDE repair time per defect, and NDE Inspection time. a. Subsequent charts show specific sensitivities for each factor. 2. Following results are average of 50 trials each of 72 cases obtained varying: a. NDE defect rate between 0.1% and 10%. b. NDE repair time at 1, 2, 4, or 10 hours. c. NDE inspection time at 8, 24 or 40 hours. 3. Aggregate results indicate Defect Rate is driving factor. a. Tornado Plot below indicates Defect Rate has higher relative importance. b. Model-Fit of all data indicates Defect Rate has highest coefficient. Relative Effect of Inputs on System Duration Pareto Plot of Estimates Term NDEDefectRate NDERepairTime NDEInspectTime t Ratio 10.702772 7.211215 2.517328 1. System Duration is 1,193 hrs + 159 hrs per defect percent + 106 hrs per NDE repair hour + 10 hrs per NDE inspection hour 2. Defect rate affects the system duration more than the other variables. PMC 2008 Ares I-X Upper Stage Simulator Team 8 Result: NDE Sensitivities to Baseline Glenn Research Center Duration (hours) 1. These charts show impact of changing each NDE factor and holding others factors constant to their baseline. 2. Model Baseline includes: a. b. c. d. NDE Defect Rate = 5% NDE Inspection = 8 hrs NDE Repair per Defect = 4 hrs Baseline shown with markers in each chart NDE Repair Time Sensitivity 4000 3500 NDE Defect Rate Sensitivity 3500 3000 2500 2000 1500 1000 500 0 0% 2% 4% 6% 8% 10% 12% NDE Defect Rate Total Duration Inspect Duration Repair Duration NDE Inspection Time Sensitivity 3000 2500 Duration (hours) Duration (hours) 0 2 4 6 8 10 12 3000 2500 2000 1500 1000 500 0 2000 1500 1000 500 0 0 5 10 15 20 25 30 35 40 45 NDE Repair Time per Defect (hours) Total Duration Inspect Duration Repair Duration Inspection Time (hours) Total Duration Inspect Duration Repair Duration PMC 2008 Ares I-X Upper Stage Simulator Team 9 DES Results and Conclusions Glenn Research Center 1. Process is sensitive to number of fabricators and welders. a. Suggested baseline using 8 fabricators and 6 welders per shift. 2. Flanges, Gussets, & Lugs process matches closely to planned schedule when two full shifts are used. 3. Segment manufacturing Charges match closely to planned schedule when two full shifts are used. 4. Previous analysis has shown process is sensitive to NDE defect rate, NDE repair time, and NDE inspection time. a. Each percent of NDE defect rate adds 159 hours to process duration. b. NDE Repair time increasing by 1 hour adds 106 hours to process duration. c. NDE Inspection time increasing by 1 hour adds 10 hours to process duration. PMC 2008 Ares I-X Upper Stage Simulator Team 10 What SimVision® Does Glenn Research Center Inputs Milestones Model Simulation Predictions Critical Path Project schedule Cost (Work, rework, wait, communication) Quality Risks (Work, communication) Position and people work load Activities and Activity Characteristics Activity Dependencies Position Characteristics Organization Task Assignment Meetings Decision-Making Policy Communication Policy Organizational Dynamics • Individual behavior • Organization theory • Performance measures • Discrete event simulation SimVision permits effective front loaded project design by facilitating planning and organizational design with meaningful scenario analysis. PMC 2008 Ares I-X Upper Stage Simulator Team 11 NASA SimVision® Model of USS Project Glenn Research Center Design and Fabricate Flight Test Article 1 Element Control Board (Tues 1 - 3 pm) Technical ntegration Meeting (Wed 9:00 - 11:00 am) Project Management Staff Meeting (Mon 11-12 pm) Interstage ntegration Meeting Thurs 9:30 - 11 am) ARES I-1 USS Element Manager Ground Processing Manager Technical Review Board (TRB) (Thurs 1 - 3 pm) CLV US Fit Test Article Manager GSE & Logistics Lead Structural / Mechanical Design Lead Fabrication Stand Up Meeting (Daily 9-9:30 am) Manufacturing Lead Element Test & Integration Lead Procurement GSE Design Team UDAC Design Teams Segment Design Team Fabricators Welders Inspectors Stand Aplha Stand Beta Stand Gamma Stand Delta Clock / Mate Space B-333 Team Plumbrook B-333 Test Q4 2006 UDACs SRR Q1 2007 Q2 2007 Q3 2007 Q4 2007 Q1 2008 Q2 2008 Q3 2008 UDAC-2 (9/3 - 11/30 2006) UDAC-3 (11/26/06 - 3/19/07) UDAC-4 (2/28 - 5/22 2007) PF 2 PF-2 Beta Occupied PF-2 UDAC-5 (6/25 - 9/04 2007) PF-2 ainting PF-2 Design and Procure Design & Proc Comple te PF-2 Parts Fab PF-2 anges PF-2 Rolling PF-2 all Fit F-2 Top Flange -2 Final Welding PF-2 2nd Structure Assmbly Complete PF-1 / PF-2 SS Stack & luation Trial 1 (Bldg 50) PF-1 / PF-2 SS Stack & luation Trial 2 (Bldg 50) PF-1 / PF-2 SS Stack & luation Trial 3 (Bldg 50) PF-1/2 Trial Stack Complete DIS-2 Delta Occupied DIS 2 DIS-2 DIS-2 Assmbly Complete DIS 2 Xport B50 to Plum Brook Pathfinders DIS-2 Design and Procure Design & Proc Complete DIS-2 Parts Fab DIS-2 anges DIS-2 Rolling DIS-2 all Fit DIS-2 Top Flng -2 Final Weld IS-2 2nd Structure ainting DIS-1 DIS-1 Assmbly Complete DIS 1 Xport B50 to Plum Brook DIS-1 Alpha Occupied DIS 1 ainting IS-1 2nd Struct DIS-1 Design and Procure Design & Proc Complete DIS-1 Parts Fab DIS-1 anges DIS-1 Rolling DIS-1 all Fit DIS-1 T Flng S-1 Fin Weld Inte rnal Access ther 2nd Internal Access Fab Fab Complete BDM Mounting Fab Structure Complete BDN Mounting Other 2nd Structure Fab Fab Complete US-6 Beta Occupied US 6 US-6 Segment US-6 Prelim Des GRC Proc Data Dwg US-6 S-6 Flange US-6 Final Des GRC Mfg Dwg US-6 Parts Fab S-6 Rolling Welding 6 Wall Fit S-6 Top Flange Final Welding US-6 6 2nd Structure Painting -6 to US-5 LT, C&M US 6 Complete - Arrive Bldg 333 SS-3 Stacking US-7 Gamma Occupied US 7 -6 to US-7 LT, C&M US-7 Segment US-7 Prelim Des GRC Proc Data Dwg US-7 S-7 Flange US 7 S-7 Rolling S-7 Top Flange 7 Wall Fit US-7 to SA LT, C&M Final Welding 7 2nd Structure US-7 Painting US-7 Final Des GRC Mfg Dwg US-7 Parts Fab Welding Complete - Arrive Bldg 333 SA Aplha Occupied SA SA A 2nd Structure SA Segment Proc SA Comple te A Painting SA Flange SA Prelim Des GRC Data Dwg SA Final Des GRC Mfg Dwg SA Parts Fab Welding SA Rolling SA Top Flange - Arrive Bldg 333 o SM LT, C&M Final Welding A Wall Fit SM Beta Occupied SM SM Complete M 2nd Structure M Painting mple te SS-3 Stack SM Segment SM Prelim Des GRC Proc Data Dwg SM SS-3 Stacked erSeg-3 Install and Checkout M Flange SM Final Des GRC Mfg Dwg SM Parts Fab Welding SM Rolling SM Top Flange inal Welding - Arrive Bldg 333 Destack and Ship to Hangar Wall Fit SS-3 Avionics Delivered Basis US-5 Delta Occupied US 5 US-5 Segment US-5 Prelim Des GRC Proc Data Dwg US-5 S-5 Flng US-5 Rolling US-5 all Fit US-5 T Flng S-5 Fin Weld 5 2nd Structure US-5 Painting -5 to US-4 LT, C&M US-5 Final Des GRC Mfg Dwg US-5 Parts Fab US 5 Complete - Arrive Bldg 333 SS-2 Stacking Weld US 4 US-4 Gamma Occupied US-4 US-4 Segment US-4 Prelim Des GRC Proc Data Dwg US-4 Final Des GRC Mfg Dwg US-4 Parts Fab S-4 Flng Weld US-4 Rolling US-4 all Fit US-4 T Flng S-4 Fin Weld US-4 2nd Structure S-4 Painting to US-3 LT, C&M US 4 omplete Arrive Bldg 333 ♦ Ares GRC Sch 06Oct31.mpp ♦ USS Resource Loading_2006-11-08.xls US 3 to US-2 LT, S-3 Painting C&M US 3 US-3 Delta Occupied US-3 S-3 Flng US-3 Segment US-3 Prelim Des GRC Proc Data Dwg US-3 Final Des GRC Mfg Dwg US-3 Rolling US-3 all Fit US-3 T Flng S-3 Fin Weld US-3 2nd Structure US-3 Parts Fab omplete Arrive Bldg 333 Weld US 2 US-2 Alpha Occupied US-2 US 2 S-2 Flng US-2 Rolling US-2 all Fit US-2 T Flng S-2 Fin Weld US-2 2nd Structure S-2 Painting US-2 Segment US-2 Prelim Des GRC Proc Data Dwg US-2 Final Des GRC Mfg Dwg US-2 Parts Fab to US-1 LT, C&M omplete Arrive Bldg 333 Complete SS-2 Stack SS-2 Stacke d Weld SuperSeg-2 Install and CHeckout S-2 Destack and Ship to Hangar SS-2 Avionics Delive red SSR Gamma Occupied SSR SSR SSR Painting SSR Segment SSR Prelim Des GRC Proc Data Dwg SSR omplete Arrive Bldg 333 SS-1 Stacking SSR Flng SSR Rolling SSR all Fit SSR T Flng SSR Fin Weld SSR Final Des GRC Mfg Dwg SSR Parts Fab Weld US 1 US-1 Beta Occupied US-1 S-1 Flng US-1 Rolling US-1 all Fit US-1 T Flng S-1 Fin Weld US 1 US-1 2nd Structure S-1 Painting -1 to IS-2 LT, C&M US-1 Segment US-1 Prelim Des GRC Proc Data Dwg omplete Arrive Bldg 333 US-1 Final Des GRC Mfg Dwg US-1 Parts Fab Weld IS 2 IS-2 Delta Occupied IS-2 o IS-1 LT, -2 Flng IS-2 Rolling IS-2 all Fit IS-2 T Flng S-2 Fin Weld S-2 2nd Structure 2 Painting IS 2 omplete Arrive Bldg 333 C&M IS-2 Segment IS-2 Prelim Des GRC Proc Data Dwg IS-2 Final Des GRC Mfg Dwg IS-2 Parts Fab Weld ( (FJ-1 / FTA1-0) Gamma Occupied SR Flng (FJ-1 / FTA1-0) TSR Painting FTA1-0) Transport om B50 to Plum Brook TSR Parts Fab TSR ng-Btm R Wall TSR ing-Top R Wall Fit-Top TSR Segment ward FTA 0 / FJ-1 and FTA-1/ FJ-2 Contract Weld Delivery Bldg 50 it-Btm FJ-2 / FTA1-1) Gamma Occupied SR Flng SR ng-Btm R Wall it-Btm SR ing-Top R Wall Fit-Top SR Painting SR (FJ 2 FTA-1) Deliv ery - Bldg 4 SR Segment IS-1 Alpha Occupied IS-1 SR Parts Fab Weld ( FTA-1) Delivery - Bldg 333 IS 1 IS-1 Segment IS-1 Prelim Des GRC Proc Data Dwg IS 1 IS-1 Final Des GRC Mfg Dwg IS-1 Parts Fab -1 Flng Weld IS-1 Rolling IS-1 all Fit S-1 T Flng S-1 Fin Weld S-1 2nd Structure 1 Painting omplete Arrive Bldg 333 Complete S-1 SSR Temp Stack SS-1 Temp Stacked SuperSeg-1 Install and Checkout S-1 Destack for SR Integration SS-1 SR / FTA-1 (FJ-1) Assy SR Inte gration and nstrall & Check Out SS-1 Stacked SS-1 Destack and Ship to Hangar SS-1 Av ionics De livered GSE Design - CM Interface GSE CM Interface Fab GSE Stand Design Beta Fixture Fab Beta Fixture Assy Beta Fixture Complete Gamma Fixture Assy Delta Fixture Fab Gamma Fixture Complete Delta Fixture Assy Delta Fixture Complete GSE Local Cart Design Local Cart 1 - Fabrication Local Cart 1 - Assembly Local Cart 1 Complete GSE Design - 1st Stage GSE 1st Stage Fab SE Fnl Assy Fixture Design B333 Uno - Fabrication B333 Uno - Assembly B333 Uno Complete B333 Dos - Assembly B333 Dos Complete B333 Tres - Assembly B333 Tres Complete Ship Segements to KSC Ground Support Equipment Gamma Fixture Fab B333 Dos - Fabrication B333 Tres - Fabrication PMC 2008 Ares I-X Upper Stage Simulator Team 12 Results: Baseline Simulation Case Glenn Research Center 1. Critical Path a. Ship to KSC Date: March 2009 – six months late! b. PF’s / DIS’s miss completion dates. c. Preliminary Segment Designs are showing schedule pressure. d. Final Segment Designs show float. Critical Path flows through “middle” segments 2. Organizational Impacts a. Coordination effort is significant for integrative tasks (Rolling, Seg Fitup). b. Space/Stand competition drives schedule delays. c. Welding tasks present an opportunity to compress/recover schedule. d. Welders and Fabricators experience > 1 month backlog (up to 6 welders and 16 fabricators required during peak periods). Stand Backlogs Welding Backlog Conclusions a. Accelerate Preliminary Designs and Final Designs for early Segments (e.g., US-2/3/4/6/7). b. Delay designs for later segments to provide resources to critical path. PMC 2008 Ares I-X Upper Stage Simulator Team 13 Results: Manufacturing Skills and Processes Glenn Research Center 1. Case Summary a. Ship to KSC Date: October 10 2008. b. Welder experience increases over time. c. Utilize Weld Machines for critical welds. d. WIP space / move tasks off stands. 2. Organizational Impacts a. Manufacturing resource backlogs are < 2 weeks (less errors, faster work). b. Reduced facility competition increases pressure on assembly resources. c. Facility competition drives > 30 days work backlog. CP shifts to early segments 40% reduction in errors Conclusions a. Crew based assembly reduces schedule delay by ~4 months. Task (PF-2 b. Weld Machines further improve delay by Final Weld) Errors ~1 month. Baseline 4.8 c. Weld Machines alone reduce schedule Weld pressure by ~3 months. Machine 3 d. Provide WIP space; move Secondary Structure Assembly, Painting, and Clock and Mate to Bldg 333 (saves ~30 days of schedule delay). PMC 2008 Ares I-X Upper Stage Simulator Team Ignored 0.7 0.3 Corrected 1.5 1.1 Reworked 2.6 1.6 14 Manufacturing Organizational Requirements Glenn Research Center 1. Case Summary a. Ship to KSC Date: September 1 2008. b. Crew based segment manufacture. c. Utilize Weld Machines for critical welds. d. Increase Manufacturing and Inspection resources. Improved welding performance drives decision wait increases for inspection tasks as errors are produced over a shorter period. 2. Organizational Impacts a. Increased Welder (+4) and Fabricator (+2) resources reduces schedule. b. Stand competition is reduced but still an issue for Beta and Alpha as schedule compresses. c. Decision Wait and Coordination volume due to inspection delays drives schedule. Streamlined Inspection process and increased Inspection resources reduce backlog by more than 1 month. Conclusions a. Increase Fabricators (Total = 14) and Implement Crew Based Fabrication and Assembly Welders (Total = 8), or • 4 dedicated high skill/experience Welders using Weld b. Dedicate engineer to reduce decision Machines (Bldg 50) • 12 Fabricators (8-12 Bldg 50, 0-4 Bldg 333) cycle time to 24 hours, resource 1 Accelerate Inspection Process inspector/stand and improve approval • 1 dedicated inspection engineer process. • 24 hour review process for NDE 15 PMC 2008 Ares I-X Upper Stage Simulator Team Communication and Work Flow Sensitivities Glenn Research Center Conclusions a. Improved coordination reduces schedule pressure by 2 weeks. b. Focusing on early segment design while permitting fabrication of low risk items does not delay final delivery date to KSC. Manufacturing Dashboard reduces Coordination risks and requirements by ~30% Delay of Segment Fabrication to July 1, 2007 permits PF’s, DIS’s and other Fabrication (e.g., GSE) steps to proceed with reduced competition for resources. PMC 2008 Ares I-X Upper Stage Simulator Team 16 Summary of SimVision Findings Glenn Research Center 1. Resource backlogs and facility competition drive schedule delays. 2. Early design backlogs pressure start of fabrication for early segments. 3. Solving the facility and process issues shifts the bottleneck risk to the organization. a. Centralized decision making and formalized communication increases risk of delays and need for coordination. b. Coordination and communication risks rise as facility bottlenecks are removed. 4. Specializing resources and tying them to tasks reduces Schedule pressure but increases need for Coordination. a. Competition for facilities (stands and floor space) will put stress (overtime, rework, quality issues) on the organization to maintain schedule. b. System must be developed to coordinate the required highly choreographed manufacturing flow. PMC 2008 Ares I-X Upper Stage Simulator Team 17 Summary of SimVision Recommendations Glenn Research Center 1. Accelerate designs for Pathfinder-2 and early Segments (e.g., US2/3/4/6/7). 2. Delay designs for later segments to provide resources to critical path. 3. Operate a crew based manufacturing organization. − 4 Welders with Weld Machines (1 additional Weld Machine dedicated to rolling machine). − 12 Fabricators (distributed across fabrication and assembly in Bldg 50 and 333). − 4 Segment Assemblers (Bldg 333 super segment assembly). − Increase welding skills (all critical welds made by weld machines) and develop welding experience (crew based assembly). 4. Begin fabrication of low risk items as early as possible (> 2 weeks). 5. Plan for in process holding space for parts (2 weeks), flanges (8) and segments (1). 6. Plan to relocate Clock and Mate tasks for middle and late segments to Bldg 333. 7. Dedicate 1 engineer (on site), resource 1 inspector/stand and improve approval process (<24 hrs) to minimize schedule impact of inspection tasks. PMC 2008 Ares I-X Upper Stage Simulator Team 18 Project’s Take Away Recommendations Glenn Research Center 1. Adopt mechanized welding approach with goal of minimizing weld defects and subsequent rework. • Project Response: Accepted. Agrees with EWI recommendations. Mechanized MIG welding now baselined for Skin-Skin and Flange-Skin welds. 2. Shift design resources to finish common segment design ASAP. • Project Response: Accepted. Common Segments US-2, 3, 4, 6, 7 focus of first incremental USS Critical Design Review. Three serial CDR’s planned out. 3. Adopt crew based approach to maximize crew skill level. • Project Response: Accepted. Professional welders obtained via support contract. 4. Implement schedule dashboard to increase visibility of segment manufacturing flow to team and labor on floor. • Project Response: Accepted. Dashboard one of several metrics used to communicate and track schedule. Daily Standup and Material Review Board (MRB) meetings setup. 5. Implement flange storage recommendations. • Project Response: Accepted. Lay down area in west end of AMF to be utilized. 6. Study moving clock/mate/match drill and secondary structure to Building 333. • Project Response: Partially accepted. Mate space carved out in AMF, plus additional processing (painting, internal access structure installation) required after clock/mate/match drill that is better suited for AMF. PMC 2008 Ares I-X Upper Stage Simulator Team 19 A Look Backward: Benchmarking Simulation Against Reality Glenn Research Center Simulation Prediction Optimum Staffing: 4-6 welders, 8-12 fabricators Two shift operation required to meet schedule. Accelerate early Common Segment design and delay Complex Segment design. Increase welding skills and experience, utilize mechanized welding to minimize defect rate. Plan for in process holding area for machined parts (flanges, tangs, lugs). Dedicate staff to improve AMF floor coordination and reduce rework decision making time to 24 hrs max. PMC 2008 Actual Result Validated: Steady State Staffing (Per Shift) 6 welders, 9 fabricators Validated: Project implemented two shift ops from outset. Validated: Three serial critical design cycles/reviews implemented, to feed three serial manufacturing Charges. Validated: Had to contract out to obtain welders with sufficient skill. Once obtained, mechanized welding was no longer needed! Validated: Need for more floor space drove set up of Temp Storage Facility. Validated: Floor Director position created per shift; Segment Lead Engineer positions created; MRB set up to meet daily. Ares I-X Upper Stage Simulator Team