national aeronautics and space administration robert k. crain 281.483.09221
TRANSCRIPT
National Aeronautics and Space Administration
1
INCOSE TGCC May 2015MBSE and Data ArchitecturesRobert K. [email protected] Technologies
MBSE without a process-based Data Architecture is just a set of random characters…..Robert K. Crain 281.483.0922
Robert K. Crain 281.483.0922 2
Abstract
A Process-based Data Architecture provides an understanding of what information is needed to effectively execute the enterprise's business processes and provides a framework for effectively managing the enterprise's information environment. It provides a representation of data artifacts and data assets that classifies and defines all data entities, their attributes, and associations to facilitate knowledge of how data is produced, managed, and shared in different contexts of use. The Data Architecture provides identification, management, interoperability, and integration of information across business or organizational elements needed to support Product Data Life-cycle Management goals. It also ensure that data needed by programs and projects (e.g. for milestones, reviews, mission operations, anomalies or investigations, decisions, and outcomes) are identified and managed to provide traceability of data used in decision making. The Data Architecture is the foundation for a solid Model-Based Systems Engineering (MBSE) approach. A well-defined process-based data architecture will facilitate:
• Standardization (definitions, concern viewpoints, etc.) • Discovering document based process flaws• Defining process compliance audits• Interoperability and data exchange• Data Integration• Application (Tool) Schema and Methodology development• IT Architecture requirements definition (Applications, platforms, MDM, etc.)
Robert K. Crain 281.483.0922 3
Overview
MBSE without a process-based Data Architecture is just a set of random characters…..
– What is MBSE?– Enterprise Architecture Components– 10 step process Enterprise Architecture Definition– Tool Agnostic Data Architecture Definition– Data Architecture SE Vee Entity Relational Diagram (ERD)– Data Architecture Wiki (Reference)
Robert K. Crain 281.483.0922 4
What is MBSE?
MBSE without a process-based Data Architecture is just a set of random characters…..
What is MBSE?
– Four letters that represents….– Four Words that are comprised of even more letters
• Model • Based• Systems• Engineering
– Which is represented as an acronym…
M B S E
Robert K. Crain 281.483.0922 5
Letters, Words, Definitions, Acronyms
• In order for us to understand what MBSE is someone had toA. Define letters B. Arrange those letters into wordsC. Provide a definition of those wordsD. Combine the words together to provide a specific meaningE. Define the term acronym and how it is applied
MBSE
odelasedystemsngineering
Robert K. Crain 281.483.0922 6
MMUM?
So, what is Mfano mifumo ya uhandisi ya msingi (MMUM)???
Model Based Systems Engineering (MBSE) translated into Swahili.
Because we have the underlying definition of letters and arranging those letters into words, which have definitions, they can be translated into other languages.
Robert K. Crain 281.483.0922 7
Model Based Systems Engineering….
• Modelo de la ingeniería de sistemas basados• Modello base di sistemi di ingegneria• القائمة النظم هندسة نموذج• Μοντέλο που βασίζεται μηχανικής συστημάτων• 基於模型的系統工程• Odelmay Asedbay Ystemssay Engineeringway
All of these examples follow the same underlying “process” for creating an “architecture” to “model” words that have meaning.
The same needs to be true with Model Based Systems Engineering….
Robert K. Crain 281.483.0922 8
Translation
A rendering from one language into another
• Once we have defined what MBSE represents we can translate it into many different “languages”
• The same is true with developing a tool agnostic, process based Data Architecture– Once it is defined it can be represented into many different
applications (languages) and translated in tool agnostic representation for mapping and transformation
Robert K. Crain 281.483.0922 9
Systems Engineering Process
• Systems Engineering is Systems Engineering, the process hasn’t changed significantly throughout the years.
• What is changing is the methodology for implementing the processes.
Robert K. Crain 281.483.0922 10
Systems Engineering Methodologies
• Document-centric – Implementation of the Systems Engineering process using distributed
documents, often duplicating data
• Model/Data-centric – Implementation of the Systems Engineering process using distributed
tools/applications, object-oriented, define once use many
A process based, tool agnostic, data architecture is the key to a creating a sustainable, successful model-based methodology for implementing the various systems engineering domains.
Robert K. Crain 281.483.0922 11
Model-Centric Domains For SE
• Model Based Project Control• Model Based Engineering
– Model Based Manufacturing and Operations– Model Based Systems Engineering– Model Based Design
Robert K. Crain 281.483.0922 12
Enterprise Architecture Framework
SECU
RITY
AR
CHIT
ECTU
RE
TECHNOLOGY (IT) ARCHITECTURE
APPLICATION ARCHITECTURE
DATA ARCHITECTURE
BUSINESS (PROCESS) ARCHITECTURE
Architecture Description
Business Mission Based Standards and Processes for meeting the Agency Vision
Data Processed Based Data Model (Objects, Attributes, Associations, Audits)
Application Enabling tools designed to perform domain specific functions
Technology IT Networks and Infrastructure
Security Application and Technology layer built in controls to protect data
Focus of this
Presentation
13
NASA Agency Business Process Sample
Robert K. Crain 281.483.0922
Number Process/Procedure TitleNPR 2830.1 NASA Enterprise Architecture ProceduresNPR 7120.9 NASA Product Data and Life-Cycle Management (PDLM) for Flight Programs and ProjectsNPR 7120.9 Handbook NASA PDLM HandbookNASA/SP-2010-3403 Schedule Management HandbookNASA/SP-2010-3404 Work Breakdown Structure (WBS) HandbookNPR 7120.5 NASA Space Flight Program and Project ManagementNPR 7120.5 Handbook NASA Space Flight Program and Project Management HandbookNPR 7120.7 NASA IT and Institutional Infrastructure Program and Project Management Requirements NPR 7120.8 NASA Research and Technology Program and Project Management RequirementsNPR 9420.1 Budget FormulationNPR 7123.1 Systems EngineeringNPR 7123.1 Handbook NASA Systems Engineering HandbookNPR 7150.2 NASA Software Engineering RequirementsNPR 7500.1 NASA Technology Commercialization ProcessNPR 8000.4 Agency Risk Management Procedure RequirementsNPR 8705.2 Human-Rating Requirements for Space SystemsNPR 8705.5 Probabilistic Risk Assessment (PRA) ProceduresNPR 8705.5 Guidebook PRA Procedures Guide for NASA Managers and PractitionersNPR 8705.6 Safety and Mission Assurance (SMA) Audits, Reviews, and AssessmentsNPR 8820.2 Facility Project Requirements
Program Mgmt, Systems Engineering, Operations, PDLM, Enterprise Architecture
Robert K. Crain 281.483.0922 14
Data Architecture Definitions
• Provides an understanding of what information is needed to effectively execute the enterprise's business processes and provides a framework for effectively managing the enterprise's information environment.
• Links information behavior (i.e., accessing, using, and sharing data), information management processes, and information support staff to other aspects of the enterprise.
• Provides a (process-based) representation of data artifacts and data assets that classifies and defines all data entities, their attributes, and associations to facilitate knowledge of how data is produced, managed, and shared in different contexts of use.
• Provides identification, management, interoperability, and integration of information across business or organizational elements needed to support program PDLM goals.
• Ensure data needed by programs and projects (e.g., for milestones, reviews, mission operations, and anomalies or investigations, decisions, and outcomes) are identified and managed to provide traceability of data used in decision making.
15
EA Definition Process Flow Overview
Robert K. Crain 281.483.0922
Business/Process
Architecture Definition
Data Architecture Framework Definition
Data Object Application Assignment
Application Schema
Development
Application Methodology
Definition
Disparate Data Object Identification
Data Integration Framework Definition
Technology Architecture Development
Define System of Interest
Architecture
Application Architecture
Selection
1 2 3 4 5
678910
Tool Agnostic Data Architecture Program Project Specific Tool/Application Development
Shared Data Integration Service
Focus of this Presentation
System Definition
16
EA Process Flow Step Description
Robert K. Crain 281.483.0922
1- Business Architecture Definition Based on the position of the system-of-interest within the hierarchy of the overall architecture determine the specific guiding Business Architecture rules for Systems Engineering and Project Management process requirements. The Program/Project could adopt established Agency standards, tailored Center standard, tailored Program/Project standards, or decide to create new standards that meet or exceed the parent standard(s). For Programs/Projects these documents would be included as Applicable Documents in Systems Engineering Management Plan (SEMP).
2- Data Architecture Framework Definition Based on the Program/Project Business Architecture definition a Data Architecture “Framework” can be adopted, tailored or developed. The Data Architecture “Framework” identifies the Data Objects, attributes, and associations (correspondence) applicable to Program/Project. Existing DAFs can be tailored and expanded to add optional Center/program/project specific attributes and or associations. Deviations/tailoring to the common model should be documented and well understood so the data can evidentially be transformed for data exchange, integrated reports, audits, and metrics. Unique program/project specific additions should also be documented in the Program/Project’s Data Architecture Dictionary. The resulting DAF should be documented/referenced in the in the Systems Engineering Management Plan (SEMP).
3- Application Architecture SelectionIdentify Authoritative Data Repository Application for Data Objects. Applications MUST be able to “expose” data to other Applications for Data Integration. These applications are typically server-based, not desktop instances. Applications are typically related to a specific Systems Engineering or Program Management function (i.e. Requirements Management, Risk Management, Schedule Management, Budget Management, etc.).
4- Data Object to Application AssignmentThe identified Data Objects, from the Data Architecture “Framework”, necessary to perform the Project Management and Systems Engineering functions, per the Business Architecture, should be assigned to their authoritative repository (application). Typically authoritative data objects reside in only one application.
17
EA Process Flow Step Description Con’t
Robert K. Crain 281.483.0922
5- Application Schema DevelopmentEach application needs to incorporate the subset of the Data Architecture “Framework” based on the assignment of Data Objects to that Authoritative Application. The Program/Project can take liberties with the Data Object names and attribute labels, however the underlining definition must be maintained. An XML schema for the NASA Data Architecture Framework can be tailored and parsed, based on Program and application specifications and used to establish the traceable Data Framework.
6- Application Methodology Development A Methodology, how you implement the process, should be developed and documented for application’s function(s) based on the assigned data object. In theory, there should not be duplications of implementations, multiple ways of capturing the same data. This methodology needs to be governed to ensure the application’s model artifacts are consistent.
7- Disparate Data Objects Identification An analysis of the Applications and their assigned Data Objects will result in the level of data disparity within the program/project. The Program/Project should document anywhere there is an association between two data objects that are in different, disparate applications
8- Data Integration Framework Definition The purpose of the Data Integration Framework (DIF) is to perform as middle-ware, which joins the disparate data object to maintain the structure of the model. The DIF is the back-end Application-to-Application interface. The DIF front-end provides the user interface to record the associations across data objects/applications and to generate user defined reports and analysis. The Data Integration Framework should comply with the previously defined Data Architecture Framework to develop a Common Model. This common model will facilitate the transformation of data from Application-to-Application and generating reports.
9- Technology Architecture Development The Technology Architecture includes the Information Technology (IT) infrastructure, which includes the Data Integration Framework (DIF). To be successful the Technology Architecture must have real-time, or near real-time, access to the exposed application data items.
10- Define the System of Interest Architecture After the Business, Data, Application, and Technology Architectures have been defined the Model-Based Project Management and Systems Engineering efforts can begin for the System-of-Interest Architecture. It is important for the System-of-interest to be modeled per the described data architecture and application methodologies. Inter-application data items should be registered in the Data Integration Framework to extend the data model from one application to the other application(s) within the Technology Architecture.
18
SE Tool/Application Language
Robert K. Crain 281.483.0922
Odelmay Asedbay Ystemssay Engineeringway
Modelo de la ingeniería de sistemas basados
القائمة النظم هندسة موذج
Μοντέλο που βασίζεταιنμηχανικής συστημάτων
基於模型的系統工程
DOORS
Magic Draw SysML
Data Architecturexml
Translation Mapping
xsdMS Project
Cradle Model Based Systems Engineering
Windchill
Robert K. Crain 281.483.0922 19
Tool Agnostic Data Architecture
• Model/Data-centric – Implementation of the Systems Engineering process using distributed
tools/applications, object-oriented, define once use many– Based on a process-based Data Architecture
Schedule
You
Organization
Work Breakdown Structure
Budget
Requirements
Verification
Product Under Test
Facility
Support Equipment H/W S/W
Test Procedure Document
Interfaces
Configuration CAD
Deviation Variance
GMIP
Results
Test Execution
Waiver
PRACA
Robert K. Crain 281.483.0922 20
Robert K. Crain 281.483.0922 21
Sample Data Objects
Data Objects• A data object is a logical cluster of attributes in a data set
that relate to a common business process.
22
3 As - Attributes, Associations, Audits
Robert K. Crain 281.483.0922
Data Object metadata– Attributes
– A quality or characteristic inherent in or associated with a business process– String– Fixed-String (16/60/256/etc.)– Enumeration (Category)
» Single Pick (sp)» Multiple Pick (mp)
– Boolean/File– Date (MM/DD/YYYY)
– Association• Definition of relationships between Data Objects
– Bi or Uni-Directional semantics– From/To, Parent/Child hierarchy
– Audits/Metrics• Measurements of Process Compliance• Quality of Data• Measurement of Goals
Robert K. Crain 281.483.0922 23
Associations (Relationships)
Data Object Associations (ex. Requirement Data Object )• Definition of relationships between Data Objects
24
Requirement Association “Peacock” Chart
Robert K. Crain 281.483.0922
Graphical Representation of the Requirement Data Object Association rules.
Common Attributes (All Data Objects)
Robert K. Crain 281.483.0922 25
Used for standardized Life-cycle maturation, Configuration Control levels, data source identification, attribute formatting, etc.
• Inherited by all Data Objects
Unique Attributes (Requirement Data Object)
Robert K. Crain 281.483.0922 26
Metadata about the Data Object (Requirement example)• Unique to the Data Object
27
Audits (Measurements, Process, Metric)
Robert K. Crain 281.483.0922
Measurement, Data Quality, and Process Audits (ex. Requirement Data Object)
28
Integrated Peacock Charts
Robert K. Crain 281.483.0922
Numerous “Rabbit Holes” to venture
down depending on stakeholder’s
concern
Data Objects and Semantics provide
the logic for navigating the Data Architecture Model
Requirement Data Object Peacock Chart integrated with the
Verification Data Object Peacock Chart
Verified By
29
Systems Engineering “Vee”
Robert K. Crain 281.483.0922
Standard Systems Engineering Vee process executed• Left to Right• Top to Bottom• Recursive for each Architecture Level
30
SE Vee Entity Relational Diagram (ERD)
Robert K. Crain 281.483.0922
ARCHITECTURE
INTERFACE
MISSION
Achieved By
PRODUCT
Joined Thru
DVO
VERIFICATION
REQUIREMENT
ANALYSISRISK
TBX
Studied By
Studied By
Causes
MISSION-
Class Manifested
RESULT
RESULT
RESULT
Causes
Generates
HAZARD CONTROL
FMEA
CIL
Detailed By
Closes
Closes
Tracked For
Causes
PRACA
Closed By
Closed By
Closed By
VARIANCE
Generates
ResolvedBy
DOCUMENT
Documented By
O&M REQUIREMENT
LCC
RESULT
Closed By
SUPPORTSupported By
SCHEDULE
Run During
Performed During
RelatedTo
CAD
RelatedTo
GMIP
Inspected By
CAPABILITY
FUNCTION
OPERATION
Results In
OBJECTIVE
Related To
Performs
Specified By
ASSET
Specifies
Maintained By
Analyzed By
Controlled By
Basis Of
Satisfied By
Deviation/Waiver
NGO Guides
Refined By
MOP
TPM
Measured By
Measured By
Monitored By
Applicable
Modeled By
Monitored By
Assigned To
Realized By
ARCHITECTURE
Built From
Basis Of
Tracked For
Basis Of
Analyzed By
References
References Performs
Performs
Requires
Verified By
Modeled By
Provides
Causes
Studied By
Allocated To
Studied By
Produces
Constrained By
Maintained By
Related To
RelatedTo
MOE
SE Data Architecture Process Data Model
31
Data Architecture Definition
Robert K. Crain 281.483.0922
Identify “Mission-Based” Business Processes to derive the Data Architecture.• Data Objects• Data Attributes• Data Associations• Audits
Process Driven Model-Based Domains
NPR 7123.1 SE Requirements Example
Robert K. Crain 281.483.092232
Examine Official Process Documents and derive Data Objects, Attributes, and Associations
Systems Engineering Process
33
Process Based Data Objects Associations Identified
Process Based Data Object Associations
Robert K. Crain 281.483.0922
MBSE Approach provides ability to capture and formalize relationships between Data Objects that already exist in “real world” practice.
Examine Process Based Data Objects and Derive Associations and Semantics
34
Data Architecture Wiki
Robert K. Crain 281.483.0922
https://nasa-ice.nasa.gov/confluence/pages/viewpage.action?pageId=9732
Currently Hosted In NASA ICE (Integrated Collaborative Environment)
Data Object (Schema) Menu
Robert K. Crain 281.483.092235
Currently there are 53 defined Data Objects classes
with 70+ defined types
Detailed Process Procedures
Data Arch. Definition
36Robert K. Crain 281.483.0922
Data Architecture Benefits
The Data Architecture is the foundation for a solid Model-Based Systems Engineering (MBSE) approach. A well-defined process-based data architecture will facilitate:
• Standardization (definitions, stakeholder concerns, viewpoints, etc.) • Identification of Document-based process flaws• Process compliance audits• Interoperability and data exchange between applications/tools• Data Integration • Application (Tool) Schema and Methodology development• Definition of IT requirements• More detailed, integrable Data Delivery Requirements (DRDs)
Ontology is tool agnostic and represents standard relationship represented in a document-centric environment. (Ex. Verifications verify Requirements which specify an
Architecture used to achieve a Mission that is guided by NGOs and Objectives)
37Robert K. Crain 281.483.0922
BACK-UP
38
Wiki Purpose
Robert K. Crain 281.483.0922
• The purpose of the Model Based Systems Engineering & Integration Wiki is to:– Provide an Integrated model for the components of
• NPR 7123.1 – NASA Systems Engineering Processes and Requirements• NPR 7120.5 – NASA Space Flight Program and Project Management Requirements• NPR 7120.9 - NASA Product Data and Life-Cycle Management (PDLM) for Flight
Programs and Projects – Provide a Tool agnostic, common Architecture Framework used as an integration
point for distributed data• Definition of Technical data structure
– Data Objects– Attributes– Associations
– Provide standardized processes and procedures for implementing Systems Engineering and Project Control using the defined Data Architecture
– Provide standardized metrics and audits– Provide a common framework for Data/Configuration Management
Common Attributes (All Data Objects)
Robert K. Crain 281.483.0922 39
Used for standardized Life-cycle maturation, Configuration Control levels, data source identification, attribute formatting, etc.
• Inherited by all Data Objects
Robert K. Crain 281.483.0922 40
Associations (Relationships)
Data Object Associations (ex. Requirement Data Object )• Definition of relationships between Data Objects
41
Requirement Association “Peacock” Chart
Robert K. Crain 281.483.0922
Graphical Representation of the Requirement Data Object Association rules.
Robert K. Crain 281.483.0922 42
Detailed Step-by-Step Process
Robert K. Crain 281.483.0922 43
Detailed Steps (Directorate Architecture)
Hyperlink to Data Object Definition (Attributes,
Associations, Audits, etc.)
Hyperlink to detailed step-by-step directions to
establish Data Architecture model
Robert K. Crain 281.483.0922 44
WBS Dictionary Example
WORK PACKAGE
WBS WBS
WBS Dictionary
WBS
Decomposed By
SCHEDULECOST
Decomposed By
WORK PACKAGE
WORK PACKAGE
Decomposed By
Decomposed By
Decomposed By
ControlledBy
ControlledBy
WBS
Decomposed By
1st Level WBS
2nd Level WBS
Charging LevelCost Charge
Number
Associated Cost and Schedule
Controls, basis of EVM
COSTControlledBy
Robert K. Crain 281.483.0922 45
Wiki Walk Through
https://nasa-ice.nasa.gov/confluence/pages/viewpage.action?pageId=9732
Please contact me if you require [email protected]
URL