Effects-based Operations: An Overviewindianstrategicknowledgeonline.com/web/ebo1.pdf · EBO...
52
1 Effects-based Operations: An Overview Dr. Maris “Buster” McCrabb (757) 508-8735 [email protected]
Effects-based Operations: An Overviewindianstrategicknowledgeonline.com/web/ebo1.pdf · EBO encompasses and supplements both target-based and objectives-based a對pproaches. The goal
The challenge is predicting & assessing what physical actionsproduce the desired behavioral effects over some period of time
EBO …
• is an approach, a way of thinking
• supports all missions from HUMRO to MTW
• utilizes lethal and nonlethal force
• offers an approach that models an “Enemy as a System”
• offers Economy-of-Force
• offers an approach to Effects-based Dynamic ISR Management
EBO …
• is an approach, a way of thinking
• supports all missions from HUMRO to MTW
• utilizes lethal and nonlethal force
• offers an approach that models an “Enemy as a System”
• offers Economy-of-Force
• offers an approach to Effects-based Dynamic ISR Management
"Effects based operations is a methodology for planning, executing and assessing operations to attain the effects required to achieve desired national security objectives.“ AFDD 1
Presenter
Presentation Notes
The definition shown here is directly out of the Air Force Doctrine Document 1, “AF Basic Doctrine” EBO is an approach to planning, executing and assessing military operations with an explicit focus on effects as opposed to targets or even objectives. Many people ask, Isn’t this the way we have always fought wars? Didn’t we always focus on the effects we want to achieve? The answer is yes, commander’s certainly always consider effects when planning and fighting wars. What is lacking and what we are developing are the automated tools to build and assess plans that link objectives to effects (including direct, indirect, physical effects, behavioral effects and the mechanisms through which effects are achieved). And then to link the effects and mechanisms to specific actions which need to be taken. As stated here, EBO is a mindset, a way of thinking in as much as it is a new methodology. It supports all mission types from Humanitarian Relief Operations all the way to Major Theater War. EBO could utilize lethal and non-lethal force such as information warfare. We are exploring the use on information warfare for EBO. EBO offers and requires an approach to modeling the Enemy as A System or more specifically a System-of-Systems. This is done through enemy COG or NEV modeling. We use the Warden COG analysis model and the Barlow NEV model to do this. What’s important for Ebo is to address not only COG analysis but also cross-cog analysis. For example what effects does one COG such as infrastructure have on another COG such as system essentials or leadership. EBO offers economy of force by specifying both dependencies and interactions between various target systems/COGs and mechanism. It also offers an approach to dynamically tasking ISR assets to the effects that the commander is out to achieve.
5
Comparison of Effects-Based, Objectives-Based and Targets-Based Operations
Effects-Based
Target-Based
Objectives-Based(Strategies-To Task)
Target-Based (TBO) • ID enemy entities, destroy them• Focus: physical effects at target level• Looks at 1st and 2nd order effects only• No dynamic assessment• No explicit timing considerations
Objectives-Based (OBO) (Strategies-to-Task)• Strategies at one level become objectives for next• Focus: objectives at every level • Considers linkages between objectives and
strategies to achieve those objectives• No dynamic assessment• No explicit timing considerations
Effects-Based (EBO) • Address causality between actions and effects• Focus: desired effects (physical and behavioral)• Encompass both target and objective-based methods• Models the enemy-as-a-system w/adversary reaction• Considers Direct, Indirect, Complex (synergistic),
To understand EBO it is necessary to understand objectives-based and target-based approaches for planning, assessing and executing military operations. EBO is not intended to replace either of them. Rather it encompasses both of these approaches but adds support for dynamic tasking & assessment, economy of force and timing considerations. Target-based approaches identify the enemy entities or targets and sets out to destroy them. The focus is on the physical effects at the target level only. It has been the traditional—and bloody—approach to warfare for millennia. Objectives-based approaches look at the strategy at one level and turn that strategy (such as the national security level) into objectives at the next lower level (such as the theater or campaign level). The focus here is on objectives to satisfy the higher level strategies. This became a commonly used approach for planning, assessing, and executing warfare at all levels over the past decade in the USAF. With an effects-based operations approach one explicitly examines and models the causes between actions and effects. Both physical and behavioral direct and indirect effects. Effects are the main focus. EBO encompasses and supplements both target-based and objectives-based approaches. The goal is to model the enemy as a system and provide dynamic real-time assessment as opposed to the other approaches where no dynamic assessment is made.
6
Definitions
• No doctrinal template in approved Joint doctrine– USAF Doctrine has rudimentary definitions
• Historically and theoretically behavioral most important but hardest– therefore, traditionally military operations focus on physical effects
• Major problems with behavioral effects:– causality between action & effect– observability of effect– uncertainty of intervening variables
7
AFDD 2-1 Definitions
• Direct Effect
• Indirect Effect
• “Effects” described
• Limitations:
– Precise definitions of “effects”
– Effects versus Objectives
– Conflating Mechanism
– Connections between Direct & Indirect Effects
– Complex or Cascading Effects
8
AFDD 2-1 Definitions
• Direct Effect: “Result of actions with no intervening effect or mechanism between act and outcome. Direct effects are usually immediate and easily recognizable.” (AKA 1st order effect)
• Indirect Effect: “Result created through an intermediate effect or mechanism to produce the final outcome,which may be physical or psychological in nature. Indirect effects tend to be delayed,and may be difficult to recognize.” (AKA 2nd, 3rd,…, n-order effects)
9
Effect
• A result from some action
– Assumes a causal connection
– “IF x THEN y” is a statement of (direct) effect
• Point of View dependent
Action
Cause
Action
EffectEffect
AFDD 2-1 EBO CONOPS
10
Mechanism
• Actions Cause Results (Effects)
• Mechanism explains cause
– Rationale explains purpose (Commander’s Intent)
• IF (x; action) THEN (y; result) BECAUSE (a, b, … ; mechanism)
– Uncertainty in action, result and cause suggest a probability- based approach (Bayes)
• Clausewitz: if you defeat an enemy’s fielded force, then a rational leader will submit because his country is undefended.
• Douhet: if you terrorize the population, then a rational leader will submit because the populace will rise up, rebel, and force submission.
Presenter
Presentation Notes
The yellow words defeat and terrorize are effect words. The blue terms fielded force and population are objects of those effects. They are highlighted because they will be found in the models we use to trace and predict effects in an enemy.
11
Mechanism continued
• Slessor: if you interdict infrastructure or supply, then fielded forces will be less combat capable, making them easier to defeat which leads to…an undefended country.
• Schelling: if you threaten things leadership value, then they will submit rather than lose those things.
• Warden: if you isolate leadership, then they are prevented from doing something which would thwart our will. They submit because the imposed paralysis prevents them from doing otherwise.
Presenter
Presentation Notes
The example from RAF Air Marshall Sir John Slessor was chosen because it points out an idea crucial to EBO that will be fleshed out soon: direct actions that ultimately result in indirect effects. That is, the effects on infrastructure and those effects on an enemy’s fielded force which ultimately leads to defeat of the enemy. Professor Thomas C. Schelling, a seminal theorist on deterrence and compellence, is used to point out the ultimate connection to behavioral effects at the strategic level. Conflict starts as an idea in the minds of strategic leadership and it is there that it is resolved. That notion, of course, underlies the contention of USAF Colonel (ret.) John A. Warden, III that leadership is always the central center-of-gravity (COG).
12
Mechanism finished
• In reality these are few, if any, single mechanisms at work.
– Direct and indirect effects combine to form complex and cumulative effects
– Effects cascade for good or ill
• Predicting, then assessing, how physical actions spawn behavioral effects (I.e., results from actions) is the major challenge.
Presenter
Presentation Notes
In a few slides, more complete definition of terms will be offered. Again is the emphasis that EBO is not simple—and Clausewitz reminded us that in war the simplest thing is very hard.
13
Actions & Tasks
• Tasks are work (actions) to be done
– Operationally, tasks are normally assigned to tactical units (squadrons, brigades, CVBGs, etc.)
• EBO theory is agnostic to the originator of action
Mechanism (Cause)
Action (Tasks)
Effect (Result)
14
Direct & Indirect Effects
• Direct Effects result from (caused by) direct actions.
• Indirect Effects result from (caused by) indirect actions.
• Indirect actions may be the result of a (previous) direct effect.
Mechanism (Cause)
Direct Action (Tasks)
Direct Effect (Result)
Mechanism (Cause)
Direct Effect (Result) & Indirect Action
Indirect Effect (Result)
15
Complex & Cumulative Effects
• Complex Effects are a combination of effects, either direct or indirect, at an instant in time & space.
– Mechanism explains how these effect combine
• Cumulative effects are complex effects as viewed over some time interval.
– Effects do have persistence, though that generally varies
Direct Effect (Result) Indirect Effect (Result) Indirect Effect (Result)
Complex/Cumulative Effect (Result)
Mechanism (Cause)
16
Cascading Effects
• Cascading effects are direct, indirect, complex or cumulative effects that ripple through a system– Captures the notions of an acceleration or a
multiplier effect (from economics)– Requires a systems perspective: relationships,
dependencies, or connections between elements• Cascading effects can be vicious or virtuous but
which they are is point of view (POV) dependent– What’s bad for the bad guys is good from our POV
but bad from theirs– In a HUMRO, what’s good for the object of our
attention is good from our POV and theirs
17
Effects-Based Plan Representation
Objective
Indicator
Effect Desired
• direct effect• indirect effect• complex effect• cumulative effect
EBO Terms
Task/Activity
(Isolate the Battlefield)
(Deny Access)
(Mechanism)
18
Indicators
• Are not effects
• Better to be observable then not but “not seen” does not mean “not there”
Win the War
Destroy Will Stop Second Echelon(Isolate the Battlefield)
Prevent RiverCrossing
(Deny Access)
DestroyFuel Res.
Carpet Bomb Drop Leaflets
DestroyBr1
DestroyBr2
DMPI 1 DMPI 2
Acceleration of Straggler Count
River Clear
Units in Bivouac
Traffic Density
Objective/Task/ActivityEffect Desired
Indicator
19
Strategy: Definitions
• The art & science of employing the armed forces…to secure the objectives of national policy by the application of force or the threat of force. [JP 1-02]
• A means1 to accomplish an end.[AFDD 2-1]
• The use of engagements for the object of the war.[Clausewitz]
• The art of distributing & applying military means2 to fulfill the ends of policy. [B.H. Liddell Hart]
• A plan of action designed in order to achieve some end; a purpose together with a system of measures for its accomplishment. [RADM J.C. Wylie]
1tool, implement 2resources
20
Strategy, Objectives & Effects
• Goal, Strategic Aim, End-State: the realm of policy—the decisive results– An end-state is the set of required conditions that achieve the
strategic objectives. [JP 3-0]
• Operational Art: determines where, when & for what purposes [why2]…forces will be employed– Use resources efficiently & effectively to achieve strategic
objectives [JP 3-0]
– Defines the parameters of operations (restraints & constraints) [FMFM 1-1]
• Effects: the result (outcomes) of direct actions (e.g., missions) and “indirect” actions (e.g., functions) [AFDD 2-1]
2rationale
21
Ends, Ways, Means, Risk
• Ends: What military conditions must be produced…to achieve the strategic goal?
• Ways: What sequence of actions is most likely to produce [those] condition(s)?
• Means2: How should resources…be applied to accomplish that sequence of actions?
• Risk: What is the likely cost…in performing that sequence of actions
[JP 3-0]
22
Strategy, CONOPS & COA
• A verbal or graphic statement, in broad outline, of a commander’s assumptions or intent in regard to an operation….It is included primarily for additional clarity of purpose. [JP 1-02]
This is a build slide. The idea is to encapsulate the EBO concept on one slide. This is a real historical example from DELIBERATE FORCE (the air campaign against the Bosnian Serbs during Aug-Sep ’95). BSA = Bosnian Serb Army. BiH = Bosnia-Herzegovina Two other ideas worth mentioning: 1. From every study (e.g., SAB, DSB) done in the 1990's the same conclusion is drawn: US aerospace power capabilities outstrip our C2ISR capabilities. To a great extent, this was re-confirmed in ALLIED FORCE particularly concerning time critical targeting, battle damage assessment, and especially campaign or operational-level assessments. 2. Effects-based operations span the gamut of missions the USAF performs. As General Deptula's briefing ("Expeditionary Aerospace Operations") shows, even humanitarian relief operations (HUMRO) can be planned, conducted, assessed and sustained using the effects-based model. Indeed, doing so better lays out capability and means requirements.
25
Monitor Assess Plan Execute
Entities (Shallow Extraction)
Relationships(Intermediate Extraction)
Discovered & Inferred Knowledge
Events(Intermediate, Deep Extraction)
Level 0
Sub-object Data Assoc. & Estimation
Level 1
Object Refinement
Level 2 Situation
Refinement
Level 3 Impact
Assessment
Level 4 Process
Refinement
JIPB, ONA, PBA & EBO
DefineEnvironment
DescribeEnvironment’s
Effects
EvaluateAdversary
DetermineAdversary
COAsOER
Databases&
Sensors
Effects-based operations is the
‘engine’ that drives IPB, dynamic C2,
and ISR management.
Supports JP 3-56.1 Stage I: Operational Environment Research
A recurring theme is the essential reliance EBO places on intelligence preparation of the battlespace (IPB). IPB is a rigorous analytical methodology that is focused on providing predictive intelligence to warfighters at the right time for use in planning and executing operations. The process provides a structured means to gain, correlate, and exploit information at all levels of war, supporting decision makers from the JFC down to a tactical entity. While the process is sequential, it is also continuous and cyclical. The four steps are define the environment, describe the battlespace’s effects, evaluate the adversary, and determine adversary course-of-actions (COA). Looking at the battlespace’s effects is a necessary step in determining whether the commander’s desired effect is attainable or not. Evaluating the enemy and estimating potential enemy COA might reveal the capacity and/or intention of employing a class of weapons that the commander might consider or such importance as to designate them as a TST/TCT. IPB might also reveal what isn’t known and hence the potential for targets to arise (emerge) during the course of operations. The Fusion/EBO framework in the background emphasizes how basic a capability they provide for IPB. Each one of the IPB elements depends upon information. Fusion engines provide such information from disparate data sources. Essential elements of enemy information include: adversarial intelligence collection capabilities, presence and intentions, critical information, and probable adversary knowledge. Finally, the slides shows an understanding of ONA and PBA based upon the shown sources. ONA & PBA share the core IPB processes. ONA and EBO share COA option development. PBA shares ISR planning with EBO.
26
Working Together
JBI
FusionEBO
EBO RequestsActionable Information
JBI Finds What’sKnown and TasksGeneration of New
Information
Fusion Engines & Fuselets Generate NewInformation & Pass to EBO
While the IPB process is sequential, it is also continuous and cyclical.
Presenter
Presentation Notes
This is a build slide. The sequence gives one manner in which the tools might work together in a generic way. The term “actionable information” is from Dr. Bob Herman’s “Fusion Development” brief. Not 100% sure what constitutes “actionable” but at least it must be context sensitive and presentable along with all its other attributes (e.g., timely, accurate, tailorable, etc.). Without elaboration, the EBO requests may be in the form of queries for archived or past information, or a subscription for future information. Likewise Fusion tools can create information regardless of tasking and publish that via the JBI. “Fusion engines” and “fuselets” are used to differentiate between “heavy” and “light” tools.
27
Combat
Campaign
2. Operations &AnticipatedResponses
Plan
Predicted Effects
Observed EffectsCombat
Campaign
3. ActualOperations(Target Set /
Target / DMPILevel)
4. ActualResponses
(COG/Target SystemLevel)
Assess1. DesiredBehavior
(COG/TS Level)
EBO Macro Model
28
Semantic Network
Approach to Modeling EBO
Size = Importance of NEV to National LeadershipThickness = Importance of Connection to other NEV
We have leveraged a number of previously build operational models to help us generate a conceptual framework to build off of for this EBO program. The goal of the EBO model is to provide a framework that helps identify and predict how actions taken by our forces will lead to the direct and indirect effects we desire to defeat our enemy or perform our other missions. The EBO model leverages existing models used for planning, execution and analysis. It augments them to 1) support dynamic tasking across planning, execution and assessment; 2) explicitly incorporates a model of the enemy-as -a-system and enemy reactions and 3) supports economy of force through the specification and analysis of interconnections between target systems/COGS to determine indirect effects. This chart shows the models were are adopting. These four models are being combined to form an EBO meta- model. The first model is a slightly modified version of the JP 3-56.1 (C2 for Joint Aerospace Operations) Campaign Planning Model. It was augmented to include the specification of rationale and incorporates feedback loops for both combat and operational assessment in support of dynamic tasking. The second model leveraged is Warden’s “enemy as a system model”. This model is used for target set analysis based on the enemy’s vital center. It models the enemy as five concentric circles. The innermost circle is leadership and as one moves outward you encounter organic essentials “stuff that goes through the pipes”: power generation, industrial production, agriculture, infrastructure (the pipes: roads, railways, canals), population, and fielded forces. Each ring is itself composed of the same 5 circles (decomposition). The main set of rings constitute the enemy’s COG. The sub rings are called “target systems” and are further broken down into “target sets” which are used to identify specific targets. A key strength of the Warden model is the interdependencies. The 3rd model is Jason Barlow’s National Elements of Values model. This model is used to target those entities that most directly support the enemy’s ability or will to continue w/ it’s current behavior. This model weighs both the importance of the NEV and the link between the NEVs. The last model shown is McCrabb’s “Enemy Reaction” model. This model is used to identify how the enemy might react to an attack. The model breaks reaction into three cases: how the enemy should react, how the enemy can react, and how the enemy could react. All based on known and perceived capabilities.
30
Campaign Model
StrategyObjectives
Resources
COG Analysis
Mechanism
StrategyObjectives
Resources
TGT Analysis
Mechanism
StrategyObjectives
Resources
TGT Analysis
Mechanism
COA
COA
COA
COA
COA
Campaign/Objectives
Force/Tasks
Mission/Activities
Combat Assessment
Campaign Assessment
Modified from JP 3-56.1
Stage I: OER (IPB)
Presenter
Presentation Notes
The next layer of detail is based upon this Campaign Model found in Joint doctrine. The two modifications are making mechanism—the explanation of cause—explicit and separating campaign assessment from combat assessment. Note this model extends from the campaign level—the domain of Joint Force Commanders and component commanders—down to the tactical level. Note also this model applies to all instruments of military power: land, sea, aerospace, special forces, or information operations.
Those familiar with John Warden’s model recognize some of its limitations. It is included here to show one option for the COG analysis part of the concept.
33
AgentAgent
Internal Model&
Library
Internal Model&
Library
Sensory Boundary
Adaptation Space
Input Process Output Outcome
InterveningVariables
McCrabb’s Agent Adaptation Space (Stage I-V)
34
The (Almost) Complete Puzzle (Stage I-V)
Commander’s IntentObjectives/
Desired Effects(Stage II)
IPB Tools& Processes
EBO/DTTT&Ps
Enemy COA vs.Friendly COA
(Stage IV)COA Selected
Branch
Branch
Stage VWargaming
Presenter
Presentation Notes
The “Almost” reflects the imprecision in these concepts individually that will not get any better when they are put together. Everything starts from Commander’s Intent (CI): end state, purpose, method and risk. CI informs, directs, guides and feeds upon the processes, products and tools of EBO and IPB. The common elements between EBO and IPB are: they are heavily involved in prediction, they rely on the production, exploitation and dissemination of information to include tasking ISR assets to gain the required data, they focus on effects not actions, and they ultimately must deal with targets though such a designation has a considerably wider dimension than merely “points in three-dimensional space against which force will be applied.” Many targets acquire special designations, generally derived from CI. Among those are decisive points, HVT (high value targets), HPT (high payoff targets), TUT (targets under trees), and TST (time sensitive targets).
In Firing for Effect then Colonel David A. Deptula outlined a framework for effects-based planning that is adopted here with some (believed) small, adaptation. IPB and EBO conduct COG and TSA from which specific targets are identified within each system. This remains the same in Deptula’s work. The small change adapted here is that in parallel with that analysis, EBO tools are turning objectives and overall desired effects (for example, disrupted TBM R&D) into more specific effects. This allows those charged with determining specific targets and weaponeering options to have guidance on which target/weapon pairing is most likely to produce the direct effects sought. Since indirect effects can only arise from some previous direct action, this small modification does appear crucial.
36
Cause (Mechanism) Events
Tasks
Conditions (End-States)
Indicators
Results (Effects)
Actions (Direct and/or Indirect)
=
=
Strategy (Ways, How): A Plan of Actions (Tasks) that employ resources (means, with)
To accomplish Ends (What)
What
Why
How + With, Who, Where, When
= COA
Effects & Operational Art
Presenter
Presentation Notes
“Operational art translates the JFC’s strategy into operational design, and, ultimately, tactical action, by integrating the key activities of all levels of war.” (JP 1-02, Dictionary of Military and Associated Terms) Operational art requires commanders to answer: what conditions must be produced to achieve the strategic goal (an end-state is the set of required conditions that achieve the strategic objective); what sequence of actions will most likely produce those conditions (ways; a COA is a plan of action that will accomplish a mission); what resources (means) are needed to accomplish the sequence of actions; and what are the likely costs or risks. (JP 3-0, Doctrine for Joint Operations) Joint doctrine uses “why” to explain rationale or purpose. EBO also uses “why” for mechanism (e.g., “Why this COA will achieve the required end-state.”). Results equate to conditions (what is wanted) and actions equal tasks performed by tactical units (I.e., how the results/conditions/end-state will be achieved). Indicators can exist at each point. Note that those at the action, tasks, and events level are normally associated with combat assessment while those at the mechanism, effects, and conditions levels are normally those required for campaign assessment.
37
Cause (Mechanism) Events
Tasks
Conditions (End-States)
Indicators(MeasuresOf Success)
Results (Effects)
Actions (Direct and/or Indirect)
=
=
What
Why
How + With, Who, Where, When
= COA
Planning (Stage V)
Presenter
Presentation Notes
Prediction (estimation or forecasting in a dictionary sense) is a planning task. It depends on IPB, especially for the prediction of mechanism. There are several tough problems on this slide. On one side, is the problem of determining what set of actions will achieve the desired effect. This is complicated by many things but especially the presence of uncertainty and multiple mechanisms when none are sufficient. On the other side is determining what set of conditions will produce the strategic goal. On both sides, these challenges multiply when one moves from the purely physical realm to the behavioral realm. As Joint doctrine points out, there are two general means for obtaining objectives by force: an imposed settlement and a negotiated settlement. Both are behaviorally-based. This extends to the operational level where the goal is often expressed in terms of getting enemy commanders to move from a mindset of gaining or maintaining gains to one of reducing or postponing losses. AFDD 2-1 uses measures of success and indicators interchangeably. This may be confusing when viewed in light of the definitions of standards and conditions found in Joint doctrine. There measures provide a means to describe how well the standard must be (the criterion) performed under specified conditions. As used here, indicators, by themselves, do not have the qualitative attribute “how well” in them. Better to think of an indicator as “what is measured” and leave the Joint definitions of standards and conditions alone.
38
Direct Effects Indirect Effects
Direct Actions Previous Effects
Cause
Complex Effects Cascading Effects
Cause
Other & Previous Actions
Effect Achieved
Effect Desired?Yes
No
Problem? Opportunity?
Re-Plan
Execution & Assessment
Indicators
Presenter
Presentation Notes
To reduce clutter, only the lines for indicators of complex and cascading effects are shown. The main challenge shown on this slide is determining how direct and indirect effects combine to form complex effects. By comparison, observing cascading effects is relatively simple. Another key point on this slide is not all occasions when the effect achieved does not match the effect desired is a “bad” thing. Based upon the uncertainty and possible lack of a sufficient cause mentioned previously, having plans work out better than anticipated is a real possibility, especially if the plan was based on a high degree of risk aversion. Adversaries are as prone to making mistakes as we are.
39
Adversarial Models
Display Solutions
Model Components,
Links, Processing to Validate
Improved Information
Representations & Multimodal
Displays
Improved Information
Representations & Multimodal
Displays
Selection of Application Area
Selection of Application Area
IntegrationIntegration
Requirements for Optimal Design & Employment of DF-based DAs
Requirements for Optimal Design & Employment of DF-based DAs
Prototype Development &
Tests
Prototype Development &
Tests
Characteristics of Improved Human
Performance
Characteristics of Improved Human
Performance
Dependencies & Vulnerabilities of
Human Aiding System Potential
Countermeasures
Dependencies & Vulnerabilities of
Human Aiding SystemPotential
Countermeasures
Characterization of Application
Environment
Characterization of Application
Environment
Research Specifications
Research Specifications
Necessary Characteristics for
Investigations
Dependencies & Vulnerabilities
Needing Support
Candidate DisplaysCharacteristics of
the Systems Improved Understanding of
Cultural Differences &
Effects
Improved Understanding of
Cultural Differences &
Effects
Applications of Human-in-the-Loop
Investigations
Applications of Human-in-the-Loop
Investigations
Vulnerable Process
Components
MOPs
Theoretical Understanding to Support Displays
Characterizations of Human
Dependencies & Vulnerabilities
Advanced Models of Complex, Dynamic Decision Making in
AADM/IO Environment
Advanced Models of Complex, Dynamic Decision Making in
AADM/IO Environment
Measures of Validation
Candidate DisplaysMOPs
Enhanced Operational Understanding to Improve Subsequent Design
Presenter
Presentation Notes
Initiatives in Aided, Adversarial Decision Making (1) Advances in understanding and modeling complex dynamic decision making (2) Advances in understanding of dependencies/vulnerabilities of Human and Systems to IO (3) Applications of Human-in-the Loop investigations (4) Improved techniques for information display (5) Understanding of cultural differences and their effects
40
Effects
Physicaldamage
(Enemy’s Available Military Worth)
Resolve ofEnemy’s
population
Enemy’s Will(Leadership)
Enemy MilitaryCR’s, CV’s & CC’s
EnemyPoliticalResolve
EnemyMilitaryWill
Nationalinfrastructure
EnemyCoalition
Unity
Enemy Regional andInternational
Support
Limited Occurrences ofCollateral Damage
Compoundingenvironmental factors
Each vector needs aconsistent scale or
traceable relationship
Big Picture
• EBO ATD seeks an integrated set of strategy, wargaming, COG analysis, and campaign assessment tools.
• Doctrinally sound; not dogmatically rigid
• Support EBO planning & assessment plus objectives-based or targets-based assessment
• Major Theater War => Humanitarian Relief; lethal & nonlethal; kinetic & non- kinetic (e.g., Info Ops)
This is our long range vision for implementing EBO through a dynamic tasking process. The process spans the phases of planning, execution and assessment. Note that assessment occurs during both planning and execution. The chart contains a number of tools and functions to make-up the process. The chart is color coded. Those blocks colored green are receiving the highest priority in this DTO with respect to funding and priority. Those blocks colored yellow are receiving funding for development to a lesser degree within the DTO. For the green blocks we are building prototypes and for the yellow-concept demonstrations. The blue block is an operational capability that we are leveraging and the grey blocks are necessary to meet our vision but at this point are not funded. We are however looking to leverage some ongoing work from various places to fill these blocks (e.g. LOCIS for AST/APT). The flow – The starting point is commander’s intent which includes the commander’s desired end state, method, purpose and risk. EIPB is Enhanced Intelligence Preparation of the Battlespace. This would feed the strategy development tool in order get the most up-to-date evaluation of the adversary. The SDT then decomposes the objectives of the commander’s intent into effects, mechanisms and actions that need to be taken to achieve the effects. This is done through COG analysis and strategy and mission templates. The result is a number of blue course of action options. Blue COAs are then assessed using CAT to determine and optimize the probability of successfully achieving commander’s intent. Once optimized and refined using CAT, blue COAs are sent to a wargaming tool where blue vs. red COA wargaming results in further blue COA refinement. The SDT then send required targetable actions to a targeting tool. Together with weapon asset sourcing and pairing capabilities, the targeting tool produces weapon-target pairing information such as the JIPTL. Mission and instruction data are combined into a continuously generated DAEO which schedules attack missions. This would replace today’s batch-oriented ATO generation process. The DAEO kicks off tactical level activities, that is the classic mission and engagement level actions. When execution commences, combat assessment information is sent back in the form of fused evidence to the CAT. Cat then conducts a reassessment of the campaign and dynamically replans the strategy if required. Difference between BDA (decision to re-attack) and CA (decision to restrategize) Summarize again exactly what part of this is being done under the DTO
42
EBO Key Products
• Strategy Development Tool
– produces blue COAs
– tightly integrates effects, center of gravity/target system analysis & strategy identification
– cross center of gravity interactions
– strategy & mission templates
• Campaign Assessment Tool
– predicts the probability of achieving
Commander’s intent for a blue COA
– model plan’s cause/effects relations
for a given campaign over time
– tradeoff analysis/drill down capability
Campaign Assessment Tool
Objectives Determination andCOG & Target Systems AnalysisObjectives Determination andCOG & Target Systems Analysis
Strategy Development Tool
Blue COA Probability of BlueCOA success
Presenter
Presentation Notes
As stated earlier, our immediate emphasis is on what we view as the two key elements of our EBO Dynamic Tasking Toolkit, a Strategy Development Tool (SDT) and Campaign Assessment Tool (CAT). SDT generates Course of Action options based on Commanders Intent. It does this by focusing on the effects required to achieve Commander’s Intent. It relates objectives to desired effects including indirect and cumulative effects. Center of Gravity analysis, Target System analysis and strategy identification are used to identify actions necessary to achieve the desired effects. The SDT also links COG analysis to wargaming. SDT will include both strategy and mission templates. These are plan skeletons that can be retrieved and reused for rapid plan development. Cross center of gravity analysis is an important attribute of the Strategy Development Tool. COGs today are generally analyzed in isolation. AFRL is developing means to address the impact of one COG such as transportation on another such as system essentials for example. We need to be able to answer the question, “How does a given COG influence other COGs within an enemy’s country IRS and IPB are drivers behind the Strategy Development Tool. Understanding the enemy & the battlespace are essential to predict how to the enemy should, could, or might react to our actions. CAT takes COAs and provides the likelihood (probability-based) of achieving the desired effects for each COA. The CAT tool models a plan’s cause and effects using causal modeling techniques (Bayes Nets). The CAT model is currently being enhanced to incorporate temporal representation. Indicators of effects will eventually feed CAT so that assessment could take place both during the COA planning phase and also dynamically as execution occurs.
43
Strategy Development Tool Overview
File Edit Tools Templates Missions Show Help
Legend
Objective
Effect
EffectIndicator
Task
Task/Activity
Mechanism
(Isolate the Battlefield)
(Deny Access)
Presenter
Presentation Notes
To further illustrate the Strategy Development Tool, this is a mock-up of what the front end GUI will most likely look like. The left hand side shows a tree view of the plan hierarchy. This decomposes plans starting with campaign plan, down to component plans, further down to objectives, then down to the effects needed to achieve the objectives including such things as direct, indirect and cumulative effects. The middle view here is used to graphically depict the causal relationship between effects, tasks and actions required to achieve effects and mechanisms. I mentioned mechanisms a few times. The are the explanation of causality between action and effects. For example if we took action to bomb a transformer site and the action resulted in the enemy acceding to our demands because they were without power, then the mechanism was that we disrupted their daily way of living to cause the effect we desired to achieve. The far right hand side shows an effect editor where effects are authored and edited. An underlying ontology is used to guide the constraints for defining effects.. This is also where COG analysis is initiated. The bottom screen shows a gant view. This shows the sequence of events over time. which will be used to eventually interface to a scheduler
44
Campaign Assessment Tool Overview
Build A Causal Model Compile to aBayes Net Analyze Results
(Isolate the Battlefield)
(Deny Access)
Presenter
Presentation Notes
To further illustrate the CAT, this shows a high level overview of what a user does and sees when using CAT. Through a GUI, the user first builds a causal model. Inputs specified include such attributes as effects desired (in this example we’ve built a plan from the red perspective) the objective or ultimate desired effect is to resupply POL, the object on which the effect is to be obtained is another input (in this example we want to resupply POL for the XVIII Airborne Corps) so the object of this effect would be the XVIII AC. Actions taken to achieve effects are also modeled. For example to achieve the effect of operating a road transportation system we need to maintain Routes 49 and 69 and the enemy needs to maintain a truck fleet and local air superiority so we don’t bomb the bridges. Indicators of effect are also specified. The example of an indicator shown here is “observe traffic flow”. What we call “mechanism” explains the cause and effect relationship. This is specified through the arrows between nodes. Also included as input are the persistence and delays associated with effects and the underlying conditional probabilities of the various events occurring. After the model is built using the GUI, the user then presses a button and a Bayes Net is automatically built. Bayes nets are complex diagrams that organize knowledge of a given area by mapping out cause and effect relationships among key variables and encodes them with numbers that represent the extent to which one variable is likely to effect another. Users then analyze results by studying probability profiles to see what the probability of achieving commander’s intent or the ultimate desired effect is over a period of time. The horizontal axis is time and the vertical axis is probability. Other graphical outputs are available to drill down to look at why the probability values are what they are and to conduct tradeoffs to impact overall probability. The idea is to modify the plan so that the success probability is maximized.
Red Decides to Terminate Hostilities
Red Decides to negotiate
Red Decides to use WMD
Probability ProfilesGMUGeorge Mason University
Blue Actions
Lead To
Influence Net
EXECUTABLE MODEL
COAs(actions/ times)
Indicator
IndicatorRed
Decisions
EBO Approach to Homeland DefenseEBO Approach to Homeland Defense
Military Action
DiplomacyFinancial
46
SDT: ATD vs. Objective System
Anticipated Capabilities• Limited coverage of
strategies, JFC/JFACC missions; Limited IW
• Limited COG/TS analysis; no Mission Analysis/Situation Development
• No COA analysis/comparison capability
• No JAOP development support past COA option development
• No CS (e.g., Logistics, Mobility), TACS, or Space considerations/Risk Analysis
Required Capabilities• Covers all strategy options• Covers all JFC missions (e.g.,
HUMRO, PK/PE, etc.)• Covers all JFACC missions
(e.g., CA, SA, CL, etc.)• Incorporates lethal/nonlethal,
kinetic/non-kinetic applications of force
• Supports Joint Air Estimate Process through JAOP development (JP 3-30)
