Semantic Web for the Military User

1
Semantic Web for the Military User

• C2 Applications Breakout
• Report
• 11/14/01

2
C2 Apps Participants

• Mike Dean BBN
• Rob Rasch BCBL-L (Army)
• Mike Rimmer NWDC
• Mark Gorniak AFRL
• Ken Whitebread LMSC/ATL
• Frank White SSC
• Martha Kahn Global Infotek
• Hal Hultgren NWDC
• Paul Neves BBN
• Lee Lacy DRC
• Tom Martin RME

3
Objectives

• Explore Command and Control Applications for DAML
  developments, both long and short term
• What are ways in which we can best explore the
  value of DAML for C2 functions

4
Framework Expeditionary Sensor Grid
5
C2 Applications Outbrief Framework - ESG

• Control of systems/sensors/and networks
• With auto registration, knowledge of
  capabilities, can more intelligently control for
  optimal or appropriate coverage for the mission
  at hand for planning, and adaptation to changing
  conditions.
• Infrastructure Applications
• Semantic interoperability of systems, dynamic
  reconfiguration with understanding at the
  machine level of the implications
• Sensor Applications
• Smart Sensors communicating with one-another
• E.g., unattended acoustic/IR aerial, ground, or
  underwater sensor (one of potentially thousands)
• Fusion, aggregation and reporting

6
Expeditionary Sensor Grid

• Objective
• Many cheap sensors, netted, fused, aggregated and
  managed
• Air, Ground, Maritime, not uniquely Naval, but
  essential for assuring access to the litoral
• Requirements at sensor, network, and system
  control levels
• Sensor Fusion, Sensor Management, Communications
  Management, all issues requiring automotion

7
DAML in Expeditionary Sensor Grid(ESG) for Data
and Information Fusion

• ESG - A sensor grid architecture that integrates
  sensors, networks, decision aids, warriors and
  supporting systems into a highly adaptive
  comprehensive system that operates from the
  surface to space and from sea to land.
• Widely varying heterogeneity, thousands of
  sensors and sources - coalition, legacy and
  emerging systems and sensors
• Fusion will require many more operators than
  practical and automation of fusion will depend on
  software agents
• DAML Contribution
• Referenced metadata regarding sensor
  characteristics, performance and reliability
  (pedigree), and decision-needs context for
  multiple heterogeneous sensors/sources/systems
  enabling high level automated (agent-based)
  fusion.

• FY 02 ESG Enabling Experiments
• Assess DAML markup utility for enhancement of
  fusion
• Explore bandwidth overhead and scalability
  issues.

8
DAML in Sensor Fusion
Semantic Resources
Time Series
Sensor Fusion Processor
Sensor d specification Algorithm description
C2 Fusion
Sensor c specification Algorithm description
Sensor b specification Algorithm description
Sa
Classification Confidence Localization Tracks
Event Data
Sa
Sa
Sb
Sc
Szz
Sd
Se
...
Sa
Heterogeneous Sensors
Event Time/Freq, Intensity, Shape Features,
Sensor Data
Fixed Data Structure
XML
DAML
Sensor Processing Stream using fixed structure,
XML, and DAML
9
DAML in Sensor Fusion

DATA AND INFORMATION FUSION (Client and
Services?)
Semantic Resources

Visualization, Management
Sensor c specification Pedigree Algorithm
description Unit definition Decision Context
High
Sensor b specification Pedigree Algorithm
description Unit definition Decision Context
Impacts
Sensor a specification Pedigree Algorithm
description Unit definition Decision Context
Situation
Level 3
Impact
Situations
Level 2
Situation Refinement
Event Time/Freq, Intensity, Shape Features,
Sensor Data
INFERENCE LEVEL
Object Base
?
Level 4 Resource Refinement
Level 1
Existence of object Classification Confidence Loca
lization Tracks Event Data
DAML
Object Refinement
Objects
?
Level 0
Signal Data Refinement
Low
Szz

Sc
Sd
Se
...
Sb
HUMINT
Fixed Data Structure, XML
Open Source
Time Series
Heterogeneous Sensors and Sources
Sensor Processing Stream using fixed structure,
XML, and DAML
10
By the way

• XML 10x hard code
• DAML 2x XML
• Jini/Java breakpoint

11
Recommendations/Actions
12
Communications

• For ESG, smart agent needed for dynamic
  communications management to do reasoning about
  the network to support the smart agent
  operation, need
• Comms ontology from Operational Level to Tactical
  Level (see next slide)
• Identify boundaries of sensor nets, etc.

13
Ontology Development

• Ontology from Operational Level (CJTF) to
  tactical level (weapons on target) for supporting
  modeling of sensor/ communications/ information
  management needs
• Benefits to ESG
• Experiment design
• System Concept Testing
• Once refined, system design, and construction
• Ties to many many other needs
• Leverage of Doctrine/TTP/Lessons Learned/Training

14
Experimentation (General)

• Experimentation (EEE)
• For both Communications and Fusion applications,
  refine the CoABS Grid DAML interfaces for Utility
  for
• Dynamic Communications
• Sensor Management
• Sensor Fusion

15
DAML For Fusion

• EEE Experimentation
• Explore the Depth of where DAML markup makes the
  most sense
• Explore Jini/Java (I.e., Grid) object translation
  to DAML
• Research
• Assess DAML tools for the multiple layers from
  the physical to the information management
• Bandwidth tradeoffs, etc.
• Modeling DAML/assessing utility in the mobile
  environment

16
Homeland Security

• Will you provide a wish-list of ontologies?
• Can you help with exporting data in DAML
• Ultimate goal can assess introducing DAML at
  lower level in process for ultimate use in agent
  applications
• Nearer term
• Researchers could work with data
• Will help with developing other (ESG) sensor
  related data