COINS:

1

• COINS
• Counter Insurgency Simulation
• Institute for Revolutionary and Insurgency
  Studies
• (IRIS)

Project Proposal 9 February 2006 Operations
Research
2
Institute for Revolutionary and Insurgency
Studies (IRIS)

• SynthesisPreston Dunlap (project lead, Java,
  website, insurgency framework)Wesley Wood
  (output requirements, simulation, analysis)
• ModelingPierre Jutras (missions, special forces,
  SME)Lisa Keller (rule sets, stochasticity, SME)
• DevelopmentMike Collins (database design,
  VB)Mark Donahey (requirements, MATLAB)
• SponsorDr. Al SweetserPh.D., Operations
  ResearchDirector, Simulation and Analysis
  CenterOffice of the Secretary of Defense
• Faculty AdvisorProf. Andrew Loerch,
  Ph.D.Professor, George Mason University
• IRIS Website
• http//mason.gmu.edu/pdunlap/

3
Objectives

• Develop a simulation to analyze insurgency and
  counterinsurgency campaigns for use in the
  evaluation of methods to achieve objectives in
  Operation Iraqi Freedom
• Analyze the effectiveness of a range of Coalition
  strategies in a scenario against a fixed
  insurgent strategy
• Tasks
• Characterize parameters of insurgency
  counterinsurgency
• Develop simulation environment program
• Develop insurgent control system
• Test evaluate simulation

4
Project Value

• Since the end of major combat operations in Iraq,
  Insurgent groups continue to demonstrate an
  ability to adapt, relocate, regenerate, and
  sustain a campaign of intimidation against Iraqi
  officials, professionals, collaborators with the
  coalition, and religious figures.1
• Since 2 May, 2003
• 2,295 Coalition Casualties2
• Estimated 21,000 25,000 Iraqi Casualties3
• A successful insurgency simulation will be a
  valuable tool for evaluating Coalition
  counterinsurgency strategies in achieving
  Operation Iraqi Freedom (OIF) objectives
• Quell insurgency
• Promote legitimate governance
• Achieve stability
• Maintain popular support (Coalition homefronts
  and Iraq)

Sources 1 Measuring Stability and Security in
Iraq, DoD Report to Congress, October 2005 2
icasualties.org 3 iraqbodycount.org
5
Sponsors Vision for COINS

• Currently no known comprehensive and
  community-accepted tools exist that provide the
  capability to analyze irregular warfare at the
  campaign level
• COINS has the potential to add significant
  capability to analyze insurgencies
• Results of the baseline study could support
  current OIF strategic assessments

• If accepted, could integrate COINS into the
  Simulation and Analysis Centers Irregular
  Warfare and Special Forces Analysis division.
  Potential follow-on work with GMU may occur to
  further develop the model.

6
Technical Approach Summary

• Leverage
• Brian Trains model of the Algerian insurgency of
  the 1950s
• OIF insurgency framework initiated in Dr.
  Loerchs OR652 course
• Perform literature review of insurgent and
  terrorist theories
• Develop a model to aggregately represent
  insurgencies
• Apply the model to the current situation in Iraq
  (OIF)
• Adjust model parameters based on available data
  sources, including
• International Institute for Strategic Studies
  (IISS)
• Globalsecurity.org
• Subject matter experts (OSD, Center for Army
  Analysis, US
  Special Operations Command, MNF-Iraq)
• News reports
• Integrate parameters into a computerized
  simulation representing Iraq campaign
  (Integration and Verification Testing)
• Develop an intelligent insurgency control system
• Analyze baseline scenario
• Develop an insurgent (Red) strategy
• Analyze Coalition (Blue) effectiveness against
  Reds strategy

7
Parameter Characterization
Process

• Model parameters shall be selected to include the
  key strategic features of the OIF
  counterinsurgency effort
• Data will be selected to characterize the
  relationships among the parameters
• Analysis will yield a transfer function for each
  parameter that describes how variation of the
  target parameter effects other parameters
• As parameters are characterized, they will be
  documented for reference by the development team

Identify Required Parameters
Document Parameter Characteristics
Select Data Source(s)
Conduct Analysis
8
Parameter Characterization
Insurgents
Coalition

9
Data Sources

• Requirement
• Model parameters shall be traceable to one or
  more sources of evidence that support the
  parameter value
• Challenge
• Many sources of information about operations in
  Iraq are classified.
• Many specific parameters may never have been
  directly studied
• Values of model parameters will be inferred from
  open data sources
• International Institute for Strategic Studies
  publications
• Online archives such as www.globalsecurity.org
• Department of Defense Reports and Announcements
• News Reports
• External and internal Subject Matter Experts
  (SMEs)
• Studies of previous insurgencies

10
Parameter Characterization

• Example Parameter Intelligence Mission
• What we need to learn
• What is required to conduct it?
• Cost, units, environment?
• What are the possible outcomes?
• Head off ambush?
• Discover arms cache?
• Learn nothing?
• What other parameters influence the outcome?
• Local support, resources committed, presence of
  insurgency?
• How often is each outcome expected given other
  parameters?
• How do we value these outcomes in terms of other
  parameters?

(a)
(o)
(b)
P(o)
U(o)
11
Parameter Characterization
Other examples of parameters we will need to
characterize

• States
• Communication/coordination capabilities of
  insurgent and coalition forces
• Population effects (urban, rural, remote)
• Effects of popular support
• Effects of terrorized population
• Resources
• Appropriate selection and valuation of resources
• Limitations on resource mobility
• Duration of value

• Units
• Best representation of insurgent and coalition
  units
• Cost of deployment and maintenance
• Mobility limitations
• Operational capabilities
• Missions
• Representative set of insurgent and coalition
  actions
• Social, political, economic, and military effects
• Available reactions

12
Simulation Development

• Simulation development will be conducted in
  parallel with parameter specification
• Each epoch of the simulation shall represent one
  time step that includes preparation, operations,
  recovery, and evaluation for both factions
• Phases of each epoch are adapted from Brian
  Trains model of Algerian Insurgency
• The simulation shall limit information available
  to each faction about their opponents activities
  based on results of parameter specification

13
Phase 1 Random Events

• Determines if a random event will affect each
  factions Political Support Level (PSL), force
  structure, or objectives.
• Examples
• Danish press releases cartoons depicting the
  prophet Muhammad (PSL drops)
• Tsunami occurs in the far east requiring military
  forces (reduce number of forces)
• Conflict erupts with Syria or Iran
• Determined by a stochastic process.

14
Phase 2 Reinforcement

• Both factions expend resources to build forces
  for placement within the simulation.
• Insurgent resources represent time spent planning
  or training, expenditure of ammunition,
  explosives, etc.
• Resources are derived from various sources
  (foreign governments, terrorists, etc.)
• Coalition resources are derived from the
  political will and commitment of the Coalition
  homefront populations (tied to PSL)
• Coalition player reinforces troops first (allows
  Insurgents to respond to
  Coalition developments)

15
Phase 3 Deployment

• Factions prepare units for operations
• Coalition
• Mobile units may be placed on patrol in this
  phase and/or moved to other territories
• Insurgents
• Units may be mobilized in preparation for later
  missions (become more vulnerable to detection)

16
Phase 4 Operations

• Factions begin conducting missions
• Insurgency always hold the initiative by either
  conducting missions at will or awaiting the
  outcome of coalition missions
• Operation stage ends when each faction decides
  that no further missions are necessary (or
  possible)

Tentative Mission Sets
17
Phase 5 Turn Interphase

• Control/Depreciation Segment
• Territorial control is determined
• Depreciation of unused resources is calculated
• Recovery Segment
• Operational capability of neutralized units is
  restored via a stochastic process
• Terrorized regions may be restored to normal
  social conditions
• Redeployment Segment
• Units are reorganized and prepared for next epoch
• Final PSL Adjustment
• Political support and resource levels are
  recalculated to incorporate results of current
  epoch

If either faction has no political support,
simulation ends.
18
Simulation Performance Metrics
Coalition Scoreboard (Region X)
Insurgent Scoreboard (Region X)
Performance Metrics recorded after each round
19
Effectiveness Metrics
Effectiveness metrics are processed at the end of
the game

• 1. Time to victory/failure ( of rounds)
• 2. Missions which resulted in largest gain/loss
  in Political Support
• (Used to make strategy conclusions)
• 3. Total number of units lost
• 4. Min/Max Political Support Level

20
Insurgent Control Requirements

• An automated insurgent control system (ICS) shall
  be developed to plan and execute insurgent
  operations in the simulation
• The ICS will allow simulations to operate
  consistently and rapidly, requiring only one
  human in the loop
• The ICS shall conform to the following
  requirements
• The ICS shall simulate a decentralized control
  strategy
• The ICS shall favor soft targets and targets of
  opportunity
• The ICS shall generate strategies based on
  imperfect information about Coalition forces and
  resources
• The ICS shall incorporate other characteristics
  consistent with the OIF insurgent force as
  identified by the research team

21
System Requirements

• The COINS Simulation System shall conform to the
  following system wide requirements
• The simulation shall be programmed in an object
  oriented language to facilitate teamed
  development and incorporation of simulation
  parameters
• The simulation system shall maintain an average
  simulation rate gt 0.5 epoch/sec on a PC with lt
  2.0 GHz processor and lt256 MB RAM
• The simulation system shall operate on a Windows
  XP operating system
• The simulation system shall include a user
  interface sub-system for control during
  simulations and data collection
• The simulation shall have two human control
  modes 1) human Coalition control and 2) human
  Coalition and insurgent control
• The system shall store simulation performance
  data to a plain text file for analysis and
  archiving

22
Verification Validation

• Following integration of the model parameters
  into the simulation architecture verification and
  validation will begin
• Verification
• Model parameter documentation will be reviewed
  and compared to coded implementation
• A test case will be run to specifically
  demonstrate the proper function of each model
  parameter
• Validation
• Since no comparable strategic insurgency
  simulation system is known to exist for
  comparison, the validation process will rely on
  the feedback of Subject Matter Experts, the team
  sponsor and advisor

23
Strategy Analysis Scenario

• Following verification and validation, an
  analysis of Coalition strategies will be
  conducted in a test scenario
• The scenario shall be designed to resemble the
  OIF operational environment
• The Red Team or insurgent force shall utilize a
  fixed strategy
• Various Coalition strategies will be simulated
  against the insurgent strategy
• Performance of the coalition strategies will be
  analyzed based on the metrics and record
  presented by the simulation system

24
Projected Schedule
Light Yellow stretch goals
1. Define system architecture 2. Develop
simulation
1. Literature review 2. Define rule-sets
Stage I 5 Feb 15 Mar
1. Combine algorithms to form basic model 2.
Build rule-sets into model
Stage II 16 Mar - 23 Mar
1. Run base case 2. Test and evaluate model 3.
Make modifications to model 4. Re-run modified
model
Stage III 24 Mar - 4 Apr
1. Perform initial VV
1. Present results 2. Build final report
1. Perform extensive VV (Validation
Verification)
Stage IV 5 Apr - 1 May
25
Risk Analysis
RISK
MITIGATION STRATEGY

• Quantified social, economic, political, and
  military parameters may not reflect actual
  processes with fidelity
• Important parameters may be left out or
  unimportant parameters may be given undue weight
• Simulated insurgent strategies may not resemble
  real insurgent strategies
• Adequate data for parameter assessment may not be
  available or exist
• Integrated parameters may not represent actual
  events at a system level
• Functional performance of simulation may fail to
  meet specifications of theoretical model
• Simulation development implementation and/or
  testing may exceed schedule

• Rely on multiple sources of input for parameter
  decisions to ensure best representation
• Maintain open software architecture to facilitate
  parameter modifications
• Identify available data correlates to make
  assessment.
• Use sensitivity analysis to determine risk of
  assumptions to overall system performance
• Maintain communication with parameter
  specification team to recognize potential design
  trades early in development
• Comprehensive documentation of development
  process

26
Project Labor Projections
27
Final Deliverables Cost
28
Thank you.