Product Data & Lifecycle Management (PDLM) -- Project Management Implications Paul Gill NASA Marshall Space Flight Center, Huntsville, AL [email protected] Lisa Murphy Atura Integration, Huntsville, AL [email protected] NASA PM Challenge 2011
Product Data & Lifecycle Management (PDLM) -- Project Management Implications
Paul GillNASA Marshall Space Flight Center, Huntsville, [email protected]
Lisa MurphyAtura Integration, Huntsville, [email protected]
NASA PM Challenge 2011
Goals • NASA space flight programs and projects are now expected to plan for Product
Data and Lifecycle Management (PDLM). • PMs will understand more about
• what PDLM is, • why they are being asked to address it, • how to exploit it, and • where to go for information and support.
• Give two ConOps: IFA and DDT&E• Use actual experience from CxP to illustrate challenges
[Content below to be addressed after rest of presentation is done.]• While no PM wants to pay more than needed for manufacturing, the likelihood that proactive
management of product definition data can avoid the cost, time, and risk of recreating data for analysis, modeling, simulation, training, dependent designs (e.g., GSE), and facilities modifications may provide a more cogent motivation to exploit PDLM.
• Finally, we will review the current state of PDLM services at NASA and identify how PMs, lead engineers, designers, systems engineers, and procurement personnel can go about finding the support they need.
Concept of Operation 1: Generic In-Flight Anomaly (IFA)
• 10 years ago we developed a Flag Ship Class spacecraft. • Nearing the end of a very long cruise mode, the
vehicle must be configured for planetary arrival While coming out of cruise, a critical component experiences an operational anomaly.
• Mission team has 12 hours to fix the problem prior to entering into orbit or the mission will be lost. Built-in monitoring system on the central electronics unit indicates a device on the processor card is not functioning properly.
The Question at hand: What data will be needed, and how do we plan for it a decade or more beforehand?
IFA Data Needs < 4 hrs: Partial List• As-designed/as-purchased/as-tested/as-built/as-flown
product structure and definition• Circuit card schematic • Specifications (e.g., materials, acceptance testing)
• Where (else) used• Location and status of spares• Firmware, software, parameters• Circuit card testing and failure history
• Impact analysis of failure (e.g., FMEA)• Failure history of components in similar settings• History of component/card/sub-system behavior over
course of mission• Trades/Design Rationale
System or Function
Communication & Control
Part/AssemblyObject
Mission Vehicle Instance
Only one stream is Product Data
IFA Support Requires Multiple Streams
Part/assy object on current
mission (“as flown”
Assembly & Verification – this instance
Fabrication &
Procurement
Other instances (e.g., prior Deliveries or flights)
DDT&E datafor thisdesign
Part/AssemblyObject
(As-flown)
ADP Object |
Description
Data from ADP for this specific
object on this flight
Design and analysis data
for design history
Data from handling or operations conducted after DD250, e.g., VAB
PRACA items
Delivered data from site that manufactured
this specific part
Data about other
instances of this part and
experience on other missions
IFA ConOp Considerations
• Data created by different groups at different physical locations, at different times, in different formats, and for other purposes And, it’s ten years later
• Need a small, particular subset of all data about this part – and need it in context From different contractors, different centers From different points in a long development cycle From communities with different vocabularies From tools now superseded by later versions
How many IFA scenarios does your Project have?
Q: What is 13 GB?
A: The amount of memory required to open the top-level 3D CAD model of the Crew Module (only) at Orion’s Preliminary Design Review (PDR)
Here’s a hint
ConOp 2: Development Data Deluge• Before we have an IFA, have to get through DDT&E• We are seeing some very large amounts of data created
during Design & Testing alone Scale of product, types of analysis & testing, procurement
strategy all affect this – but no one is immune• Illustrative Cases from CxP
Core Input for Analysis: OML Analysis, Testing & Simulation Deluge Sample Documents & CAD Models: Ares & Orion at PDR
Documents: Ares 1 PDR reviewed ~500 documents and two drawingsWith ~38,000 documents in Ares Windchill Project Folders
CAD: 16 months later at Orion PDR, LMSSC delivered ~11,000 discrete 3D models for Service Module, Crew Module, Launch Abort System
LMSSC had ~250,000 versions, iterations, or variants in their Windchill vault
ConOp 2: DDT&E
Reviews Data does nothing but grow over phases
Integrated Stack OML Challenging to integrate CAD models from different
suppliers Designs at different maturities Not a design object; uses pre-release models Requires special CAD settings and practices
Analysis, Testing & Simulation Volumes of data created for and used for analysis,
verification, and testing
DDT&E: Ares/Orion OML
MSFC Ares Vehicle Integration responsible for integrated stack OML 3D CAD envelope model
• in Low-, Medium-, and High-Fidelity versions for each Design Analysis Cycle (DAC)
Proliferation of demand for OML or data from OML for other design & analysis uses, including:J2 LH2 blade ejection coneAcoustic wind tunnel 4% modelClearance analysis simulationIU/SA compartment for Human FactorsProtuberance dimensionsFairing panel separation dynamicRe-entry configuration for US
Sensor locationsGNC node pointsInboard profileIS-gimbalThrust oscillation modelsRoCs nozzle placementAntenna locations
Source: list of requested models or data from OML CAD models in DAC 2 for Ares 1.
Design Interactions
Ares1 OML Data Exchange: Multiple Sources & Heavily Manual
LMSSC
Boeing
ATK
ESMD ICEProject Folders
JSC DDMS
MSFC DDMS
KSC DDMS
Orion
FirstStage
UpperStage
Source: CxP CAD WG May 2009
Ares Vehicle Integration
(@MSFC)
Manual Processing
CxP Lvl II DIO (@ JSC)
Ground Ops KSC
More Things to Do With CAD Models• 3D prototyping• Verify/analyze design for requirements or standards compliance• Conduct “-ilities” analyses• Create motion models (oscillation, rotation)• Create time-based visualizations (e.g., of assembly processes)• Use in models and simulations (e.g., VRML)• Plan verifications & validations; prepare before & after comparisons• View, manipulate, annotate, mark-up, e.g., for TIMs, Reviews
• Mass properties: mass, CG, surface area, volume, Parts lists, used-on
• Produce illustrations, “viewables” or other representations • Communications, Public Affairs, General Information• Training & Procedures, Documents, & Manuals
12
• ICD skeleton models • Flat pattern for sheet metal parts • Bulk items, (e.g. Spray-on Insulation,
Propellant• Deployed models
• Dynamics models • Pipe Assembly Models• Harness subassembly models • Layout models • Mass properties models
A Taste: Analysis & Testing
• LMSSC test plans included telemetry ranging from 5 MB/sec (slow) to 20 MB/sec (fast) per channel
• Engineering Task Description Sheets (from CAIT) show dependencies on 507 different data packages
• Ares initiated a risk that they would not have enough storage for the testing data expected• [&&&CHECK NOTES RE SIZE OF STORAGE]
• And there would be much, much, more: • Imagery• Simulation data sets (inputs/outputs), simulation
testing set-up/configurations• Assembly, Installation, & Interference checking
Why Product Data & Lifecycle Management?Because we need to answer questions such as:1. How much should we risk (conversely, how much are
we willing to pay) to ensure the relevant data exist and are accessible, discoverable, and understandable to support an IFA?
2. Where should we invest our attention and resources to manage data during development?a. What data do we need from our contractors?b. In what formats do different users need the data?
These concerns led to changes to NPD 7120.4 to include Product Data and Lifecycle Management, and development of PDLM NPR.
What’s Happened:
In 2008, Office of Chief Engineer takes lead on PDLM1. In 2009, updated to NPD 7120.4 to include PDLM2. Started working on PDLM NPR (approved 1/2011)3. Interoperability work (CAD, model exchanges)4. PDLM Steering Committee formed
Definition….• Product Data Management (PDM). A framework that enables
organizations to manage and control engineering and technical information, specifically data surrounding the product's design, definition, and related engineering, test, manufacturing, and logistics processes and is a key element of PLM…
• Product Life-cycle Management (PLM). The process of managing the entire life cycle of a product from its conception, through design and manufacture, to service and disposal. PLM integrates people, data, processes, and business systems and provides a product information backbone for companies and their extended enterprise…
Scope & Coverage• Single Project & Tightly Coupled Space Flight Programs• Entire lifecycle for all types of product-related data
• [See NPR}
Recent experience has shown:• 3D CAD powerful, but requires special attention
• Cannot wait until ADP to get models if you have insight-oversight• Collaborative design requires robust, frequent data exchange
• Requiring same version, build of same tool not sufficient• Must look at who is doing what• Ask who needs it, why and when
• Data exchange standards lag industry practice• So far, proprietary models only sure why to get all of data
• Need to consider software along with hardware in product definition
PDLM NPR Summary (a)
Projects & ProgramsResponsible for Process and Data ArchitectureWrite a Plan and update often
• Authoritative data are identified, captured, cataloged• Agile, flexible, sound practices for data management• Critical product data receives timely attention to acquire what is
needed, assure integrity, reflect maturity state(s) and authority• Know who needs what, when, format – across lifecycle
• MDAA is responsible for seeing the PMs meet requirements
PDLM NPR Summary (b)
Information Systems/Infrastructure (OCIO, Center Director)Assure that infrastructure adequate
• Seek to effectively re-use solutions to common problems, improve performance
Tools are known and providers committed to supportSecurity has received due attention
Project Manager – not center– is responsible for producing plan, building commitments
• Work with Center or other providers to come to agreement on what services, for whom, and how
Continues for now distributed PDM/PLM tool model• No one group assigned to provide agency-wide PDLM
Practical Matters: Plans, Tools & Data Acquisition
Content of PDLM Plan overlaps traditional Project plans such as CM, DM, Records Management, SEMP, program/project plans
• Multiple uses of same applications/similar data• Must initiate plan early and then update regularly• Identify needs, project future needs, coordinate with IT supplier
Data acquisition is critical to PDLM• Challenging to write DRDs that support CAD data exchange• Need to consider the data needed during design and IV&V• Also what is at physical delivery of product, engineering changes
Few NASA personnel have hands-on experience with the new data-centric, model-centric, technology direction
Generally, NASA Projects Face:Distributed Production & Use over an Extended Lifecycle• Need to exchange and use PRE-RELEASE product data
• Mixture of internal and external sources – Centers, primes, partners, universities
• High analysis demands, highvolumes of ancillary data
• Long project life cycles
• Need for IFA reach-back
• Ten independent Centerswith local solutions
• NASA cannot dictate how things are done at primes
Rockets as ProductsDifferent Specifications Needed to Get Data for Different Needs• Do derivative designs such as
tooling, test stands• Sub-contract part of design work• Do design integrations• Conduct design review• Take over design change
authority• Do modeling and simulations • Do physical integration &
verification (e.g., at test site or VAB)
• Re-bid production
More Reasons to Care
• 2D drawings from NASA’s standard CAD tool (PTC Pro/Engineer Wildfire) are made from 3D models
• To integrate the design of the 787 Dreamliner from their four design groups, Boeing • Had 16 Terabytes of data in their master repository• Packaged and delivered quarterly 150 applications for the
distributed design teams to use• Some of the 24 different extensions to CAD models identified by
MSFC CAD standard (only some of which are released):• Interface Control Document (ICD) skeleton
models • Envelope part models (e.g., OML) • Flat pattern for sheet metal parts • Bulk items, (e.g. Spray-on Insulation,
Propellant• Generic of family table part instance
• Deployed models • Dynamics models • Pipe Assembly Models• Harness subassembly models • Layout models • Mass properties models
Questions?