MCN_ VIRT_ and RST for NPS-CEC Collaboration

Document Sample
MCN_ VIRT_ and RST for NPS-CEC Collaboration Powered By Docstoc
					 Model-based Communication Networks,
Valued Information at the Right Time (VIRT)
      & Rich Semantic Track (RST):
   Filtering Information by Value to Improve
         Collaborative Decision-Making

      Project Report on CEC Collaboration

            Rick Hayes-Roth & Curt Blais
       hayes-roth@nps.edu & clblais@nps.edu

                   August 21, 2008
                                               #1
                Outline

 Overall Vision: Model-based Communication
  Networks, VIRT, and Rich Semantic Track
 CEC/VIRT Project Prior Results and
  Accomplishments
 2008 Statement of Work
 Recent Results
 Where Do We Go from Here?




                                              #2
                Outline

 Overall Vision: Model-based Communication
  Networks, VIRT, and Rich Semantic Track
 CEC/VIRT Project Prior Results and
  Accomplishments
 2008 Statement of Work
 Recent Results
 Where Do We Go from Here?




                                              #3
             Overall Vision: Model-based
            Communication Networks, VIRT
              and Rich Semantic Track
                                                            Common Track Semantics
                                                            State-full Network
                                           Global
                                        Information
                                            Grid

                                                                              Future
  Present                                                                              Present
            Future                                                  Present
                                     Shared World Models   Future
Past                                                                                      Past
              Present

                            Future
                                                             Past



                     Past

                                                                                                 #4
             Overall Vision: Model-based
            Communication Networks, VIRT
              and Rich Semantic Track
                                                            Common Track Semantics
                                                            State-full Network
                                           Global
                                        Information
                                            Grid

                                                                              Future
  Present                                                                              Present
            Future                                                  Present
                                     Shared World Models   Future
Past                                                                                      Past
              Present

                            Future
                                                             Past



                     Past

                                                                                                 #5
                Outline

 Overall Vision: Model-based Communication
  Networks, VIRT, and Rich Semantic Track
 CEC/VIRT Project Prior Results and
  Accomplishments
 2008 Statement of Work
 Recent Results
 Where Do We Go from Here?




                                              #6
    Key Results Previously Reported
   Advanced theory and implementation of Valued Information at the
    Right Time (VIRT) and Rich Semantic Track (RST)
        Demonstrated significant reductions in bandwidth from 2-5 orders of
         magnitude in initial studies to 45-90% in CEC-specific simulated
         message streams
   Developed simulation framework for studying VIRT and RST
   Prepared/published/presented several theses and papers
   Established VIRT as key focus of W2COG
        This work now embraced by DISA and JITC
   Helped PACOM start up CMA JCTD around sharing of rich track
    information; leading design and development of the MIEM
   Advised Joint Track Management (JTM) Architecture Working Group
    and now CNDE (Consolidated Navy Data Enterprise)
   Conducted computational analysis of CEC simulated message
    streams
   Mapped CEP-to-Track-User message elements to abstract Rich
    Semantic Track model
   Defined initial measures of performance for evaluating VIRT
    applications


                                                                               #7
 CEC/VIRT Message Stream Analysis
 Simulated message streams provided by JHU/APL
 Focus on application of VIRT at an intermediary node
  (“VIRT Track User”) between the CEC network and the
  general GIG network
 Value determined by “GIG User” Conditions of Interest
  (COIs) relating to expected position and velocity, track
  identification, and engagement status
 Computed bit traffic reduction of 45-90% in short
  duration, highly dynamic air tracking scenarios
 Demonstrated capability to “tune” bit traffic flows
  based on user-defined thresholds in accuracy of
  estimations


                                                             #8
    CEC Message Stream Analysis

 Can achieve significant reduction in bit traffic from
  attention to user information needs (COIs)
 Demonstrated mechanisms for enabling CEC
  message traffic to be filtered for non-CEC users
 Opportunity for further research into user-specified
  COIs and message processing using more of the
  message content (e.g., certainty and accuracy data)




                                                          #9
                Outline

 Overall Vision: Model-based Communication
  Networks, VIRT, and Rich Semantic Track
 CEC/VIRT Project Prior Results and
  Accomplishments
 2008 Statement of Work
 Recent Results
 Where Do We Go from Here?




                                              # 10
         SOW Adjusted to $100K Budget

   1.   Develop and analyze strategies for optimizing
         distribution of track data among distributed nodes, with
         participating nodes having a variety of needs for
         precision and timeliness
- 2.    Enhance the initial simulation environment to create
         and analyze a variety of usage scenarios, to support the
         analysis and to improve the demonstration of results
   3.   Improve the methods and tools available for describing
         what information is valued, selecting it from available
         track data, and providing it to interested clients
+4.     Help assure that CEC approaches work harmoniously
         with other DoD initiatives and other DHS agency efforts
         that require air tracking capabilities



                                                                    # 11
                Outline

 Overall Vision: Model-based Communication
  Networks, VIRT, and Rich Semantic Track
 CEC/VIRT Project Prior Results and
  Accomplishments
 2008 Statement of Work
 Recent Results
 Where Do We Go from Here?




                                              # 12
        Demonstration Software

Web-based application to enter VIRT
 conditions of interest

Server-side software checks COIs against
 input data streams (XML) and generates
 alerts to registered clients (sent via e-mail,
 cell phone text message, or with other
 transport mechanisms possible)




                                                  # 13
User Registration and Login

 User Registration   User Login




                                  # 14
Create New COIs or Display/Edit/Subscribe to
Existing COIs (using Cursor-On-Target Data)




                                               # 15
Editing a COI




                # 16
Subscribing to a COI




                       # 17
            Ongoing Development

Adding track stream generator
Adding geographic visualization
Adding predictive event recognizer:
   At some future time t, where t < Tmax, determine if the
    distance between object A and object B is less than (or, the
    alternative, greater than) some threshold distance D, and
    where the probability of this assertion being true is greater
    than 1 – α, where α is a given significance level.
Integrating knowledge base
 representation of Rich Semantic Track
   Using RST as a translation hub among diverse track data models



                                                                     # 18
Progress on Standardizing RST:
          The Vision
   CEC                                    CMA
   Track                                Maritime
 Messages                             Information
                                    Exchange Model
             Rich Semantic Track
             - conceptual hub for
               interchange and
                  automated
                   reasoning


   Joint                               Other
   Track                               Track
Management                          Data Models
Data Model


                                                     # 19
   The Rich Semantic Track Model

Track
  Beliefs
      Identity and Characteristics
      Dynamic State at Time T
      History of states (past “track”)
      Predicted states (future “track”)
  Meta-Information (applicable to each element of belief)
      Evidence
      Inferences
      Error and uncertainty estimates
      Temporal qualifications
      Spatial qualifications

              The top-level conceptual hierarchy for Track.
        The full hierarchy has more than 125 high level concepts.
                                                                    # 20
                    MIEM Objectives
Share actionable maritime intelligence in     Dynamic nature of all quantities
                                                 (potentially any belief/value
 a net-centric way:                              can vary in time)
   Simple/raw data exchange                   Inexact information (“roughly 30
                                                 feet long”)
   Fusion output representation               Relative information (“a mile
   Advanced analytics support                    from the leader of XYZ”)
                                              Conditionals (“all facilities open
Establish unambiguous maritime                   on Thursdays”)
 lexicon that:                                Complex queries (“return last
   Supports communication among data             five ports of call for vessels
                                                 flying Chinese of Korean flags
     providers and consumers                     and within 3 days of US costal
   Embodies broad expertise covering the         waters”)
     extent of the maritime domain            Pedigree (“position was derived
                                                 from AIS message A, ELINT
   Leverages existing models and                 data B, and HUMINT source C,
     knowledge bases                             using inference rule D”)
                                              Belief conditions (“value is
Enable communications from/with future           considered 85% reliable”,
 services and capabilities                       “value was provided by
                                                 source S”, etc.)
   Permit extension to more sophisticated     Behaviors, states & histories
     data without change to existing             (“the events make up an
     systems                                     overall fishing voyage”)
   Import partners’ databases and export to
     them with efficient translators


                                                                                   # 21
 Levels of Value Added Information
 Level                  Type                        Example                    Value added

1 (lowest)    Sensor system reports         AIS (Automatic                Reduced development
                                            Information System)           costs for consumers

     2        Caveats & simple meta-        Sensor type, classification   Implicit quality
              data                                                        assessment

     3        Fused data & inferred         Position, crew                Synergistic improvement
              beliefs                                                     in SA
     4        Degree of belief &            Evidence, quality             Explicit information
              pedigree                                                    about quality
     5        Multiple alternatives &       Ambiguity, uncertainty        Explicit assertions of
              analysis                                                    certainty
     6        History, behavior & future    Voyages &                     Enables basic predictive
              projections                   predicted courses             analysis
     7        “Of interest” conditions &    Suspicious cargo on board     Increased analytical
              watch lists                                                 efficiency
     8        Threats & anomalies           Dangerous undeclared          Increased pre-emptive
                                            cargo                         threat reduction

9 (highest)   Case files for key entities   Histories, highlights,        Enables in-depth
                                            comprehensive details         predictive analysis



                                                                                                     # 22
    MIEM is a Nascent Standard

 CMA JCTD transfers it to MDA COI in Oct.
 Navy and USCG committed to MIEM
 NIEM (DHS) also committed to MIEM
 RST concepts should propagate more widely




                                              # 23
                Primary Object Types


• Vessels - Characteristics, capabilities, dynamic state, and relationships

• Persons - Identification, description, whereabouts, relationships

• Cargo - Shipments, equipment, manifest, and goods

• Facilities - Ports, organizations, and governments

• Events - Relates entities with associated causes and effects

• Threats - Capability, opportunity, level, threatening entity, and target

• Of Interest Lists - Heterogeneous lists of MIEM objects

                                                                             # 24
                         Vessel Model Details
                                             Extended Base
                                          • Base Metadata
                                          • Extended Metadata

                                                                                    Support Types
                                                   Vessel                           • Voyage Type
                                                                                    • Track Type
                                                                                    • Kinematics Type
                                                                                    • Boarding Type




   Identifiers   Characteristics   Documentation             Movement            State       Affiliations
   • Name                          • ISSC                • Movement Segments   • Equipment    • Persons
   • Call Sign                     • NOA                 • Ports of Call       • Cargo        • Cargo
   • MMSI                          • Safety Cert         • Voyages             • Events       • Facilities
   • IMO




Capabilities          Physical       Miscellaneous
• Range             • Size            • Home Port
• Speed             • Structure       • Classification
• Cargo             • Design

                                                                                                             # 25
          Typical Vessel Relationships
                                                           Vessel


                                               1                   1   1
                              has-a                 has-a              has-a
                    0..*                            0..*                            0..*

                  Voyage                   Ports Of Call                    Persons On Board
                                                                           • Passenger Reference
           • Number
                                                                           • Crew Member Reference
           • Origin/Destination
           • Type                                           1
           • Use Type             related-to               has-a                           on-
                                                            0..*                           board
                                                   Port Of Call                     Person
                                          • Time of Arrival
• Embedded (“has-a”)                      • Time of Departure                • Name
                                                                             • Citizenship
                                          • Port Identifier
• Associations (“on board”)
           - Strong, explicit relationships
           - Defined Association Types
• Affiliations (“related-to”)
         - Weak relationships between entities
         - ID/IDREFS references
                                                                                                     # 26
                 Person Model Details

                                    Extended Base
                                  • Base Metadata
                                  • Extended Metadata

                                                                        Support Types
                                        Person                          • Handedness
                                                                        • Gender
                                                                        • Associations




                  Physical
 Identifiers                          Details            Whereabouts   Affiliations
                Characteristics
• Name          • Height           • Birth              • Work         • Family
• Citizenship   • Weight           • Death              • Current      • Organization
• SSN           • Color            • Biometrics         • Residence    • Employment
                • Gender           • Events             • Temporary
                • Marks




                                                                                         # 27
                           Facility Model Details
                               Extended Base
                            • Base Metadata
                            • Extended Metadata


                                   Facility                                                   Port




                Physical
Identifiers                         State         Affiliations   Documentation
              Characteristics
• Name         • Location          • Cargo    • Contractors      • Certifications
• BE Number    • Accessibility                • Organization
• Type         • Sub-Facility                 • Government
               • Parent-Facility              • Staff

                                                                         Physical
                                                                                             Identifiers    State
                                                                       Characteristics
                                                                      • Depth                • Port Name   • Vessels
                    Support Types                                     • Max Vessels          • Code
                   • Port Associations                                • Number Docks         • Type
                   • COTP Region                                      • Cargo Capabilities




                                                                                                                 # 28
                           Cargo Model Details
                                               Extended Base
                                            • Base Metadata
                                            • Extended Metadata




                           Shipment                                       Equipment




  Identifiers      Characteristics      Affiliations        Status
• Bill Of Lading   • Weights           • Equipment         • HazMat
• Booking Number   • Measures          • Goods Items       • Status
• Identifier       • Declared Values   • Involved Party    • Biometrics
                   • Route                                 • Events



                                           Identifiers       Characteristics        Affiliations    Status
         Support Types
                                            • Number       • Security Devices       • Owner        • HazMat
         • Goods Item                       • Identifier   • Weights                • Shipment     • Empty
         • Manifest                         • Type         • Measures               • Vessel       • Events
         • Associations                                    • Temperature Controls   • Facility



                                                                                                              # 29
         Abstract Types:
Threats, Of Interest Lists & Events

                         Extended Base
                       • Base Metadata
                       • Extended Metadata




      Threat              Of Interest List         Event
• Capability             • Name              • Name
• Intent                 • Publisher         • Start/End Time
• Description            • POC               • Description
• Level                  • Type              • Category
• Opportunity            • Items             • Type
• Threatening Entity                         • Location
• Target of Threat                           • Affiliated Entities



                                                  Incident
                                             • Severity
                                             • Casualty Details




                                                                     # 30
Base Types with Metadata
• All beliefs carry Metadata
• Simple beliefs carry Basic Metadata
     • MIEM Support Types carry basic Metadata
• Complex beliefs carry Extended Metadata
     • MIEM Primary Types carry extended metadata

  Basic Metadata                Extended Metadata
• Affiliations                  • Information Source
• Comments                      • Analysis
• Validity Time                 • Anomaly
• Confidence                    • Data Rights
• Completeness                  • Pedigree
                                • Vulnerability

   AddressType
     NameType                      VesselType
        POCType                      EventType
                                      PersonType




                                                       # 31
       How do we Use the MIEM
        to Describe Situations?


An Illustration of Vessel State and Entity Relationships
As of February 2008, the ship was sold to an Iranian company, IC2,
and was reflagged as a Panamanian. It sailed from Portland, ME to
Abu Dhabi where it had some new equipment EQ1, EQ2 added to it
by organization ORG3. Then it made a new voyage to South Africa,
with stops at Djibouti and Dar es Salaam before arriving at Cape
Town with a filed crew and passenger manifest. We have good track
observations on the first leg of this voyage only.




                                                                     # 32
                      Example Vessel State and
                        Event Relationships
    Part-of relationships                                                Vessel State
        (Embedded)                               Voyage Details
    Explicit relationships
      (Associations)                Track
     Weak relationships             Movement Details                          Persons
       (Affiliations)                                                         On Board
                                                              Ports Of Call
                                                                              Passenger
                                   Port of      Port of   Port of   Port of
                                    Call         Call      Call      Call     Crew
Ownership      Flag    Equipment


                                             Port Arrivals(4)                    Person 001

                                                                                 Person 002
 Sold       Re-flag    Equipment               Port Departures(4)
                        Change                                                  Person n
                                                     Voyage Start/Stop/End
                                                     Boarding
                                                                               Voyage
Event Details                                                                  Manifest

                                                                                              # 33
               Example Vessel State and
               Event Relationships (XML)
    The Port with id “PORT0001” is defined as being the port of
 The person with id “PERSON0001” is defined as having the name
      Relate entities to the underlying Events that cause properties.
“Portland, ME” in the country “USA” with a defined set ofthem:
 “John Doe” and has an affiliation to a vessel with id “VESSEL001”
  Change of Ownership causes the state of the “owner” to change
    Capture simple concepts simply – with id “PORT0001”
         which he boarded at the port vessel name is “MV1”




      Describe complex relationships between many entities:

Persons On Board include a crew member with id “PERSON0001”
who has the crew role the “Captain”. He embarked at the Port with
                         id “PORT0001”




                                                                        # 34
   Final Technical Observations

Powerful language for expressing actionable
 intelligence documents


Extensible – Model can be extended to
 produce application specific schemas




                                              # 35
                Status

Beta test version released February 8,
 2008
Incremental Design review held June 10-
 11, 2008
Version 1.0 product release scheduled
 September 19, 2008




                                          # 36
           Beta Testers

CMA JCTD -Cargo
MASTER JCTD
MDA DS COI
MAGNET/MIFCPAC
CMA – Singapore
Seahawk
NSA - RTRG
NAVAIR - Tampa Bay Maritime Domain Awareness
 System (MDAS)
TTCP – Maritime AG 8  International MDA
SPAWAR Charleston - MDA Non Classified Enclave


                                                 # 37
           Transition to MDA-NIEM

                    Planned FY09
 Transition the MIEM to
  the DHS/DOJ National
                                   MDA – Data
  Information Exchange             Mgt Group
                                                Maritime
                                                Domain
  Model (NIEM) as a new                          Owner
  Maritime Domain
 Transition the MDA DS
  COI DMWG to an MDA
  Data Management
  Group
    The MDA-DMG becomes
     the Maritime Domain
     owner




                                                           # 38
            MIEM Conclusion
 MIEM provides a language for expressing
  actionable intelligence
  – Rich semantics for pragmatics of “tracking”
  – Supports higher-levels of analysis
  – Directly supports resource cueing & interdiction
 MIEM’s focus on document sharing supports
  a vital vision of collaborative intelligence
 The NIEM will incorporate the MIEM directly &
  as a guide for higher-levels of Core value
 MIEM elements and approach should benefit
  many intelligence suppliers and consumers


                                                       # 39
                Outline

 Overall Vision: Model-based Communication
  Networks, VIRT, and Rich Semantic Track
 CEC/VIRT Project Prior Results and
  Accomplishments
 2008 Statement of Work
 Recent Results
 Where Do We Go from Here?




                                              # 40
  Observations about our Collaboration

 CEC & IWS have provided generous support
 NPS is good at several things
      Academic work by professors
      Theses by students, but opportunistically
      Applied research if predictable and sustainable
      Opportunistic cross-pollination
      Review and collaboration of others’ plans &
       technical approaches




                                                         # 41
    Prospects for Further Collaboration

 CEC & IWS might want to apply some of our
  findings
    We can help
    Perhaps there’s some important sustainable
     applied research NPS could staff to support
 CEC & IWS might want to continue funding
  NPS to do what it does well
      Academic work by professors
      Theses by students, but opportunistically
      Opportunistic cross-pollination
      Review and collaboration of others’ plans &
       technical approaches
 Or, we could declare “success” and stop
                                                     # 42
                         Conclusion
 We have laid the groundwork for much more intelligent,
  efficient, and effective collaboration networks
   –   Model-based
   –   Bits flow by value
   –   Individual operators establish conditions of interest
   –   Rich semantic tracks underlie models of tracked entities
   –   Standardized forms of RST will power much improved
       information sharing throughout defense & law enforcement

 We have had an extremely productive collaboration
 We’ve learned that $100K is below critical mass for
  staffing and conducting applied research



                                                                  # 43

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:2/24/2010
language:English
pages:43