Simulation Conceptual Modeling Study Group

1
Simulation Conceptual Modeling Study Group

• Welcome to the Fall 05
• Simulation Interoperability Workshop Meeting

2
Agenda

• Welcome and Introduction
• The SCM SG as I see it
• Study group Discussions

3
History of SCM SG

• Meetings
• Kickoff meeting during Spring 03 SIW
• Follow-up Meetings during Euro and Fall 03,
  Spring and Fall 04, Spring 05 SIWs (112
  Attendees)
• Progress
• IAW TOR, established SISO reflector (now
  Discussion Group) with 79 subscribers
• IAW TOR, issued data call on SCM SG reflector for
  empirical evidence on simulation conceptual
  modeling
• Article in SISO Simulation Technology Magazine
• Initial Draft of SCM Vision Statement
• 3rd Compilation of SCM Bibliography

4
Rationale

• The Simulation Conceptual Modeling Study Group
  has been formed to conduct preliminary
  investigation on the best practices of simulation
  conceptual modeling and to establish
  recommendations for pursuit of the topic within
  the scope of the SISO, if appropriate.

5
Community Description

• MS Community Potentially Benefiting from SCM SG
• MS sponsors/users
• MS requirement managers
• MS simulation designers
• MS software developers
• MS VVA agents
• Base Object Model Product Development Group
• VVA FEDEP Overlay Product Development Group
• (A large and diverse group)

6
Community Need

• A set of Best Practices does not exist for
  Conceptual Modeling
• MS community fragmented on use of terminology
• Novice MS designers, and developers left to
  develop ad hoc solutions for the transformation
  of sponsor/user requirements into simulation
  designs and software code
• MS VV Agents left with numerous difficulties in
  establishing credibility of finished products

7
Simulation Development Elements
8
Conceptual Model

• Synthesized combination of simplified and
  abstracted parts of real world
• It represents an idealized approximation of the
  original complex real world system
• An transformation process underlies simulation
  development
• The real world is too complex to be understood
  much less modeled by humans
• Most practical approach is to abstract key
  elements forming reasonable approximation of real
  world
• The simulationist responsible for ensuring the
  approximation is adequate for the intended
  purpose

9
Role of Conceptual Model

• A simulation developers method of translating
  modeling requirements into a detailed design
  framework- Pace
• Result of the Art of Modeling or the process of
  analysis, abstraction, simplification, and
  synthesis within a simulation development project

10
Precepts of Conceptual Modeling

• A conceptual model should be malleable and adopt
  a coherent form appropriate for each particular
  simulation development phase
• Highly unstructured, informal models are
  sufficient during early simulation definition
• Structured and enhanced formal models appropriate
  during simulation development

11
Approach
1.
2.
3
1.
2.
Domain Experts
Problem
Domain
Publications
Regulations
Structured, Synthesized
Information
Legacy Systems
Domain Analysis
12
The Informal Model

• Written using natural language and contains
  assumptions made during its construction
• Plays fundamental role during the period of
  activity when the modeler conceives, programs,
  debugs, and test models
• Helps users and colleagues comprehend basic
  outline of the model from their perspective on
  how the real world operates

13
Approach contd.
Problem
Domain
Domain Analysis
Domain Modeling
14
The Formal Model

• Unambiguous description of model structure
  separated from software implementation
• Useful once users and colleagues understand
  informal model and want more detail
• Used as an aid to detect omissions and
  inconsistencies and resolve ambiguities inherent
  in informal models

15
Problems of Conceptual Modeling

• No hard and fast rules concerning model
  formulation (Uncertainty for both the developer
  and user)
• What does a Conceptual Model look like?
• No magic formulas governing what should be
  included in a model in the form of variables and
  parameters, descriptive relationships and
  constraints, or criterion for judgment of
  effectiveness (Lack of Standardized Processes)
• What should a Conceptual Model contain?
• Nobody solves the problem, rather everybody
  solves the model that they have constructed of
  the problem (Diversity and Inherent
  Incompatibility)
• What technique to use? UML, DoDAF, IDEF0,.

16
Relationship with Base Object Models

• BOMs capture key views of a conceptual model as
• Entity or event types (Static descriptions)
• Patterns of interplay and State Machines (Dynamic
  descriptions)
• Used as a building block for composable
  simulation environments.
• Payoff for Conceptual ModelingAn End State or a
  standardized formal model

17
SCM SG Goals

• Research the use of conceptual modeling in the
  MS or related information technology domains
• Perform exploratory work into the establishment
  of best practices for simulation conceptual
  modeling
• Establish recommendations for persistent
  management of the topic within SISO

18
How to Accomplish Goal 1

• SCM SG will research the use of Conceptual
  Modeling in MS and related information
  technology domains
• Make a data call for readily available empirical
  evidence to create a set of best (most-effective)
  practices
• Data Calls
• Partially successful
• Could always use more data
• Provide SCM SG documents through SISO web site

19
How to Accomplish Goal 1

• Issued email to SCM SG reflector just prior to
  SIW (30 March 05)
• Received Responses from
• Susan Solick
• Don Timian
• Jim Weiner
• Vahid Mojtahed
• Nathalie Harrison
• Tommy Nordqvist
• Averill Law

20
How to Accomplish Goal 2

• SCM SG will perform exploratory work into
  establishment of best practices for Conceptual
  Modeling
• Identify and evaluate the significance of the
  topic to the scope of the SISO and interests of
  its members
• Refine existing Conceptual Modeling terminology
• Draft a taxonomy of concepts
• Document relevant techniques and beneficial
  procedures

21
Plan 2A

• SCM SG will perform exploratory work into
  establishment of best practices for Conceptual
  Modeling
• Identify and evaluate the significance of the
  topic to the scope of the SISO and interests of
  its members
• Refine existing Conceptual Modeling terminology
• Draft a taxonomy of concepts
• Draft Simulation Conceptual Modeling Vision
  Statement
• Introductory Statement
• SCM topics to be addressed
• SCM terminology requiring definition and/or
  refinement
• SCM concept taxonomy

22
Introductory Statement

• A simulation conceptual model is an abstraction
  from either the existing or a notional physical
  world that serves as a frame of reference for
  further simulation development by documenting
  simulation-independent views of important
  entities and their key actions and interactions.
  A simulation conceptual model describes what the
  simulation will represent, the assumptions
  limiting those representations, and other
  capabilities needed to satisfy the stakeholders
  requirements. It bridges between these
  requirements, and simulation design.

23
SCM Topics to be Addressed

• Justification
• - Indication of the value added by a conceptual
  model (business case for CM)
• - Method to keep on track designer and developer
• - Identification of stakeholders
• - Impact of conceptual models upon stakeholders
• - Ability of conceptual model to address
  relevancy of real world components
• Conceptual model only artifact to document
  relevancy of real world upon interactions where
  otherwise not recognized
• Part of KA and simulation / model validation
  process

24
SCM Topics to be Addressed

• Methodology
• - Advance the understanding of a conceptual
  model, their creation and use process, goal to
  increase stakeholders requirements in final
  solution.
• - Variety of formats (UML, animated powerpoint,
  word documents) needed communicate between
  stakeholders
• - Archivable products, persistent for community
• - Common thread of components of well-formed
  conceptual models (tools, critical mass of end
  user content, strategy and manage reusable items,
  training)
• - Reuse of legacy simulators
• - Provision of a roadmap to reverse engineer
  conceptual model for legacy simulations
• - Future vision of utility of conceptual model
• - Reuse of systems/simulation engineering
  processes and products
• - Bottom-up or top-down development of conceptual
  models
• - Rationale for layering of conceptual modeling
  (national, commercial security interests)
• - Boundaries of conceptual model (conceptual,
  design, implementation)

25
SCM Topics to be Addressed

• Relationships
• - Provide support for VVA
• - Applicable to community broader than DoD
  audience (game developers)
• - Relationship to other SISO PDGs (BOM and VVA
  Overlay)
• - Relationship to high level simulation
  development processes (e.g. IEEE 1516.3 FEDEP
  Recommended Practices Guide)

26
SCM Terminology to be Defined

• Model
• Modeling
• Simulation
• Simulation conceptual model
• Purpose(s)
• Requirements
• Identify
• Audiences
• Stakeholders
• Domain space
• Referent
• Representation

• Abstraction
• Capture
• Describe
• Assumptions
• Limitations
• Resolution
• Fidelity
• Consistency
• Composability
• Interoperability
• Pattern(s)
• Layering
• Validation of conceptual model
• Well-formed CM
• Credibility

27
SCM Taxonomy of Concepts

• Stakeholder perspectives
• Requirements artifact vs design artifact
• Formal vs informal
• Requirements definition and traceability
• Methodology of establishing credibility
• Configuration management of CM (timestamping)
• Evolutionary nature of CM (changing uses for CM
  over time)
• Support to ancillary documentation

28
SCM Taxonomy of Concepts

• Types (formats capture and presentation) of CMs
• Range of conceptual model uses
• Purposes of CMs
• Scope of CM, e.g. federation vs federate (stand
  alone or associated execution)
• Common frame of reference for CM
• Fidelity resolution (level of detail / level of
  abstraction / decomposition of MS domain space)
• Attributes or characteristics of a CM
• Communication pathways internal and external
  (technical and management)
• Coherence (consistency) between CM views
• Assumption of composability for CM
• Independence from implementation, application
  specificity

29
Plan 2B

• SCM SG will perform exploratory work into
  establishment of best practices for Conceptual
  Modeling
• Document relevant techniques and beneficial
  procedures
• Guide to Conceptual Modeling references
• Listing of SIW Papers
• Bibliography of Books
• Pointers to Online Resources

30
How to Accomplish Goal 3

• Consolidate the findings of the topical
  committees and establish recommendations for
  persistent management of the topic within SISO
• Final Report to SAC
• Circulate via reflector for 30 days
• Resolve outstanding comments
• Conduct 5 day voting period
• When successful forward to SAC
• The Way Forward
• Product Nomination for SCM PDG
• Standing Study Group (SSG)
• Wait until competing PDG activities finished (HLA
  Evolved, BOM)

31
Long Term Goals

• Formulate a CM Recommended Practices Guide as a
  balloted community product as an extension of the
  FEDEP
• Provide input for DMSO VVA Recommended Practices
  Guide web site
• Provide a sound foundation for related SISO
  development groups

32
Open Discussion Topics

• Reaffirm the need
• Strategy
• Create Vision Statement leading to Strawman
• Long term process, a lot of work to be done
• Create Accessible Listing of Conceptual
  References
• Short term product, enhance quality and utility
  over time
• Revision to Vision Statement

33
Introductory Statement

• A simulation conceptual model is an abstraction
  from either the existing or a notional physical
  world that serves as a frame of reference for
  further simulation development by documenting
  simulation and implementation independent views
  of important entities, their relationships, key
  actions and interactions. A simulation
  conceptual model describes what the simulation
  will represent, the assumptions limiting those
  representations, and other capabilities needed to
  satisfy the stakeholders requirements. It
  bridges between these requirements, and
  simulation design.
• Vahid Mojtahed

34
Introductory Statement

• A simulation conceptual model is an abstraction
  from either the existing or a notional physical
  world that serves as a frame of reference for
  further simulation development by documenting
  simulation and implementation independent views
  of important entities, their relationships, key
  actions, interactions and properties. A
  simulation conceptual model describes what the
  simulation will represent, the assumptions
  limiting those representations, and other
  capabilities needed to satisfy the stakeholders
  requirements. It bridges between these
  requirements, and simulation design.
• Nathalie Harrison

35
Additional Vision Statement Comments

• Scenario Dependence
• Our work shows that without a "scenario" you can
  never know the domain in focus and therefore have
  no chance to have a practical approach to
  actually do conceptual modeling. All conceptual
  modeling begin with some sort of focused
  context. So our understanding is however a
  conceptual model can be both simulation-independen
  t and implementation-independent, they are
  scenario-dependent, those important entities,
  relationships, actions and interactions has their
  means in the context of the scenario. - Vahid
  Mojtahed
• Regarding the "scenario dependence", I would
  rather say that it is "community-of-interest
  (COI) dependent" since "scenario-dependence" may
  prevent reusability. It is probably what Vahid
  intended to mean. - Nathalie Harrison

36
Additional Vision Statement Comments

• Readability
• And regarding the last sentence in the
  Introductory Statement It bridges between these
  requirements, and simulation design." There
  should perhaps be some kind of addition that the
  conceptual model, in order to bridge between the
  requirements and simulation design, also must be
  user readable. - Vahid Mojtahed
• Vahid also added that the conceptual model must
  be user-readable. I would also add that its
  should be "human and machine readable". If we
  want to achieve the Model Driven Architecture
  (MDA) philosophy in the future that says that
  concepts must be linked automatically to the
  implementation. - Nathalie Harrison

37
History
2004
1998
2002
2000
2003
2001
1996
2010
1999
2005

BOMPDG
ReferenceFOM Study Group
BOM Study Group
HLA 1516Evolve
FEDEP PDG
Rapidly composableand scalable MS
RPR FOM
BOM PN
BOM Methodology Strawman
FOMpiece parts (FEDEP, OMT)
SRML
HLA1516
CMSE

• Tools
• Components
• Repositories

As long as things are created from scratch,
growth can be at most linear
38
Summary

• The SCM has established a foundation for a
  Product Development Group
• Previous product development experience has shown
  that this will be a lengthy process
• Several challenges remain
• To continue moving forward
• To increase visibility of activities within MS
  community
• Secure support for SCM PDG (, , , CAD)