Lessons Learned from C2IEDM Mappings within XBML

1
Lessons Learned from C2IEDM Mappings within XBML

• Charles Turnitsa
• Sai Kovvuri
• Andreas Tolk, PhD

• Liam DiMasi
• William Sudnikovich
• Verlynda Dobbs, PhD

2
Introduction and Overview

• Data Mapping exercise is being done to support
  the ongoing efforts of the XBML project for DMSO
• A reusable data translation component will be
  derived from the results of this exercise
• A recorded and reportable Methodology was used
  during the exercise that in itself can be of some
  value to others attempting a similar task

3
BML What is it?

• BML Battle Management Language
• Unambiguous language used to command and control
  forces and equipment conducting military
  operations
• To accurately describe BML requires 3 views
• Doctrine Must be aligned to existing doctrine
• Representation Must model concepts that can be
  interpreted and processed by existing systems
• Protocol Must specify the underlying protocols
  for transferring BML information between
  participating systems

4
C2IEDM What is it?

• C2IEDM Command and Control Information Exchange
  Data Model
• Designed to describe all objects and situations
  in the battlesphere, and to enable the
  communication of those descriptions between C2
  systems
• Robust data model, designed through collaboration
  and consensus of 24 nations over 2 decades
  presented by MIP Multilateral Interoperability
  Programme http//www.mip-site.org
• Highly normalized, hierarchical, and designed to
  be extensible

5
BML Prototype (1)

• BML Prototype is currently comprised of four
  components
• Combined Arms Planning Execution System (CAPES)
• Multi Source Database, based on JCDB
• OneSAF Testbed
• BML Graphical User Interface

6
BML Prototype (2)
7
BML to C2IEDM Why?

• BML targets data interoperability between C2
  systems, in particular planning systems, other C2
  systems, simulation systems, and robotic forces
• C2IEDM deals naturally with all of the C2 data
  requirements needed to make BML work
• Given the robust, extensible nature of the
  C2IEDM, and that it is accepted by our coalition
  partners, it seems logical to employ it as a data
  model for BML

8
Methodology (1)

• A Data Model has three aspects
• Property values allowed values for specifying
  characteristics
• Propertied Concepts a collection of property
  values to describe an entity in the knowledge
  domain of the model
• Associated Concepts a collection of propertied
  concepts to describe a semantic entity
• To map between data models, these three aspects
  must be addressed in a methodical fashion
• Top-down approach begins with Associated Concepts
  (meaningful subsets of data)
• Bottom-up approach begins with the mapping of
  Properties

9
Methodology (2)

• Top-Down Method
• Start with the Associated Concepts and maps them
  to each other
• Identify missing Properties by applying a
  bottom-up consistency check in the end
• Three main steps were taken in following this
  Top-Down Method
• Conceptual Mapping
• Logical Mapping
• Physical Mapping

10
Implementing our Methodology

• Conceptual Mapping
• Understanding the Associated Concepts aspect of
  each model, and matching them up
• Logical Mapping
• Properties and Propertied Concepts that make up
  the matched Associated Concepts are now mapped,
  showing how what is necessary in the target model
  is available from the source model
• Physical Mapping
• There finally exists the need to provide a
  mechanism whereby data stored within a database
  encapsulating the source data model can be
  transported to the target data model

11
Concepts and Relationships

• As can be seen from our method, the key portions
  to these three steps are based in understanding
  the concepts encapsulated within the two data
  models, and also the relationships between them
• Understand BML Concepts
• BML as realized in the MSDB
• Understand C2IEDM Concepts
• As described in the MIP documentation

12
BML Concepts Definition (1)

• Key at showing the 5 Ws Who What When Where
  Why
• BML as it is currently employed has some 22
  different conceptual areas
• These conceptual areas allow BML to accomplish
  its stated goals of (1) unambiguous command and
  control of forces and equipment conducting
  military operations, and (2) provide for
  situational awareness and a shared, common
  operational picture
• As you can imagine, it is of primary importance
  to not only gain understanding of a models
  conceptual areas, but also how they are used
  within that model

13
BML Concepts Definition (2)

• Example of the BML concept, TASK_ORG

14
C2IEDM Concepts Definition (1)

• Robust model 176 information categories, 1500
  elements, flexible by having built in mechanisms
  for extension
• Describes the battle sphere through four key
  aspects
• Objects, and their properties
• Situations, past, present, and future
• Activities, also past, present, and future
• Containers, for grouping data into packages
• Through the four aspects, all concepts related to
  the Battle Sphere can be shown, in a formal,
  hierarchical, and relational manner

15
C2IEDM Concepts Definition (2)

• All objects (organizations, persons, equipment,
  facilities, features, weather, and boundaries)
  are described in a two-tier structure
• Each class of object is an OBJECT_TYPE, and each
  instance of an object is an OBJECT
  understanding this sort of relationship is key to
  understanding the concepts of the C2IEDM
• There are a number of Entities in the C2IEDM that
  can exist independent that is to say that they
  have, with their related properties and
  propertied concepts, a self contained definition,
  we refer to these as Entities and Entity
  Structures

16
C2IEDM Concepts Definition (3)

• Example of C2IEDM Concept, Task Org

17
Logical Relationships (1)

• Gained an understanding of each of the conceptual
  areas of BML
• Identified which of the many entities, or entity
  structure trees in the C2IEDM can best be used to
  represent these areas
• Mapped out how concept areas could be defined
  using the Entities and Entity relationships from
  C2IEDM

18
Logical Relationships (2)

• The guiding principle must be to reuse as much of
  the previously incorporated information as
  possible
• Creating new concepts comprising data elements
  that already exist somewhere else creates
  ambiguity
• Necessary to check the consistency of the
  solutions bottom-up after having reached a
  consensus in the mapping method based on a
  top-down strategy

19
Logical Relationships (3)

• Showing the relationships between TASK_ORG in
  MSDB/BML, and TASK_ORG as envisioned through
  C2IEDM

20
The Future

• We are currently working on an effort to finalize
  the physical mapping step of the process
• We are investigating the use of XSLT (Extensible
  Stylesheet Language Translation)
• Creates relationships between two data models
  described in XML
• Once the relationships are created, we can devise
  a mechanism that will automate the transfer of
  data from one model to another
• This will serve as a real time tool that will
  allow communication between systems that were
  once only understanding BML or C2IEDM

21
Lessons Learned

• Conceptual Understanding
• The most time intensive portion of the study
• Despite the fact that BML and C2IEDM have, as a
  goal, describing the same knowledge domain, there
  exists differences at all three layers
  properties, propertied concepts, and associated
  concepts
• Abstraction Levels
• As the BML is realized within MSDB, there was no
  formal model used in its instantiation
• However, the fact that is instantiated within a
  physical database, and that its concepts were
  well understood, made it an excellent starting
  point for understanding concepts
• Still, there exist differences in abstraction as
  to how the two models, BML (as seen through
  MSDB) and C2IEDM address different concepts
  these were worked out through rigorous
  understanding of the BML source, and knowledge
  and study of the C2IEDM target

22
Acknowledgements

• The work described here, and in the accompanying
  paper is a team effort of the XBML group. All
  individuals of that group, as well as some
  additional experts contributed greatly to the
  success of the research. These include
• Francisco Loaiza, Gene Simaitis and Steven Wartik
  (IDA)
• James Maguira (VMASC)
• Michael Hieb (Alion)
• Mark Pullen (GMU)