Integrating Warfighting Experimentation with Dynamic Modelling A - - PowerPoint PPT Presentation

1

Integrating Warfighting Experimentation with Dynamic Modelling – A Practical Approach Martin Parr ++44 (0) 1252 455749 mcparr@dstl.gov.uk

2

Contents

• Aims
• Background
• Implementation methods
• Example: IADS Theme
• Outline Example: Information and Intelligence Theme
• Evaluation
• Conclusions

3

Aims for the work

• Add value to WFE using dynamic modelling
• De-risk critical areas of the experiment
• Provide a ‘dynamic data repository’
• Aspiration: Assess in advance the benefits that

modelling may provide

£ £

Value Value

4

Cultural issues ….

‘ …it is completely what …??’ ‘This is completely straightforward …’ ‘He said that last week’ ‘Don’t worry – we are close to getting a solution’ ‘This is completely straightforward’ ‘…mmm… are lots of unknowns here’ ‘What is that chap in the corner doing?’ ‘ ’ What the experiment manager thought What the analyst/modeller said

5

Introduction - OA - WFE and NITEworks

Operational Analysis Modelling Historical Analysis Experimentation Static Dynamic Joint War Fighting Experimentation (WFE)

6

NITEworks AP Themes Networks Information People

Indirect Fire Integration Kill Chain 1&2 Medium Weight Capability B MASC Land Tac Picture 05 Naval MCM ISTAR 1&2 Command and Battlespace Management (Land) Medium Weight Capability A Logistics C2 Effects Based Ops Operational Intelligence Support Group Joint Ops Picture Joint Intelligence Picture Classified 1&2

7

Introduction (2) - Analysis Environments

Real Real Real

Analysis of Current Ops

Real Real Real and Simulated

Real (abstract engagem’t) & simulated

Simulated

People

Real Real Simulated Simulated Simulated

Equipment

Real Simulated Simulated Simulated Simulated

Weapon Effects Analysis of Historical Ops Live Simulation Virtual Simulation Wargaming Constructive Simulation or (OA) modelling

Ability to Manipulate and Control

8

Modelling and Experimentation

Time

Experiment Coverage

4 months 3 months

M M – – MEM Paradigm MEM Paradigm

Modelling Experimentation

M M IADS, I2 IADS, I2 Future Logs Future Logs M M – – Vis Vis – – M M – – Vis Vis … …

Breadth

9

Model-Experiment-Model (M-E-M) Paradigm

MODELLING ONGOING & FUTURE THEMES

X X X X X

OUPUTS

Model-Experiment-Model Loop

10

Modelling Criteria

Valid Beneficial

Puzzle What How Problem What How Mess What How

Method Selection

Achievable

Hard Methods Soft Methods

11

IADS Theme (1)

Potential Target

Time

Δt ??

12

IADS Theme (2)

Process Model (Simul8) Entity Level Model (Matlab)

Detection Opportunities End to end process time System bottlenecks Assumptions: Process, workload Potential Target

Time Time Time

13

IADS Theme - Summary

• Results

– Full automatic tracking of all targets within the chosen scenario – Given a less bright target: Link 16 bandwidth assessment showed little additional loading when a manual intervention was needed

14

White or Red? White or Red?

Examples – MISTAR – Vessel Classification

ISTAR Assets

15

Examples – MISTAR

Probe Vessel

16

MISTAR Summary

Achievable

• Initial analysis delivered pre G2

Beneficial

• Conclusions reached

about key system bottlenecks

Valid

• Analyst checks conducted

– enough to allow tools to highlight key bottlenecks

17

Situation Understanding

Operational Landscapes:

• Political
• Economic
• Military
• Social Cultural
• Legal, Ethical & Moral
• Physical
• Technological

Iraqi Problems:

• Insurgency
• Terrorism
• Civil Unrest
• Ethnic Rivalry
• Criminal Activity

Adversary Characteristics

• Asymmetric
• Dynamic
• Uncertain
• Chaos
• Complex
• Novel
• Ambiguous

Enemy Or Friend?

Campaign Pan

Information and Intelligence (I2) Theme

18

The process modelled by the I2 Theme

class CCIRM (VKB & VAT) IRM::IRM + Check RFI Requirements() + Combine RFI Details() + Establish IR Management() + Manage Expiry of Collection Task() + Manage Expiry of Received IRs() + Manage Expiryr of Nominated RFIs() + Manage IR Information() + Match Intelligence Products() + Monitor Changes to Information Requirements() + Pass RFI to Tasking() + Publish Information Product for Analysis() + Raise RFI() + Report Intelligence Product() + Request Intelligent Report() + Search for Similar RFI() + Task RFIs() CC::Collection Nomination + Check IR Against Collection Task() + Create Nomination() + Report Nomination() + Track Nomination() CC::Collection Plan + Create Initial Collection Plan() + Establish & Update Intelligence Day() + Maintain Awareness of ISTAR Mission Status() + Manage Collection Information() + Manage Information fo Intelligence Day() + Manage, Maintain and Review Exploitation Status() + Review Collection Nomination() + Update Initial Collection Plan() CC::Collection Mission + Establish Collection Missions() + Input Initial Collection Plan() + Prepare Collection Mission Resourcing() + Re-Initialise Collection Missions() + Resource Collection Missions() + Schedule Collection Missions() + Track Collection Schedule() + Update Collection Mission Resourcing() CC::Collection Task + Disseminate Collection Task Awareness() + Establish Collection Tasks() + Initialise Collection Tasks() + Provide Collection Task Details() + Raise New Collection Task() Information Requester (from Context) Collector::ISTAR Operation + Maintain ISTAR Library Information() + Manage ATO() + Manage Informatioin() + Manage ISTAR Assets() + Manage ISTAR Assets & Missions() + Plan ISTAR Mission() + Process Intelligence() + Re-Plan ISTAR Mission() + Schedule ISTAR Missions() Collector::ISTAR Task + Execute ISTAR Operational Tasks() + Re-Initialise ISTAR Operations() + Register ISTAR Dynamic Tasks() + Register ISTAR Operational Tasks() + Resource ISTAR Operational Tasks() Targeteer (from Context) IP::Intelligence Processing + Acknowledge RFI Amendments() + Alert Analysts of New Intel Input() + Allocate Task to Analyst() + Collate Outputs() + Decompose & Allocate IRs() + Format & Publish Data() + Identify Areas of Research() + Manage Intelligence Library() + Produce Daily Brief() + Receive Informal Request() + Review Analyst Output() + Review RFI Request() + Support Intelligence Processing() IP::Intelligence Analyst + Allocate Task to Intelligence Analyst() + Assess & Assimilate Information() + Establish Intelligence Analysts() + Identify Outstanding Intelligence() + Manage Analyst Information() + Prepare Analyst Resource Brief() + Raise Request for RFI() + Receive Tasking() + Search Data Stores() IP::Intelligence Task + Establish Intelligence Tasks() + Manage Intelligence Task Information() + Prepare Intelligence Task Brief() + Register Intelligence Task() + Report Intelligence Research Tasks() Customer (from Context) Commander (from Context) SAM::Sensor Asset Management + Monitor Collection Assets() + Report Collection Asset Availability() SE::Intelligence Processing Support + Assess Collection Task Results() + Establish Support Elements() + Manage Requests for Intelligence Information() + Manage Support Element Information() + Search Extant Data() SE::Support Element Collection Task + Collect Support Element Information() + Monitor Support Element Collection() + Register Support Element Collection Task() + Report Collected Information to Support Element() VAT::Analyse Information + Assess & Assimilate Knowledge() + Brief Analysis Team() + Collate & Report Intelligence() + Manage Analysis Team Information() + Prepare Collection Request() + Register Analysis Team() + Release Analysis Team() + Search Knowledge Base() Common support to ALL instantiations of Intelligence Processing provides information to 1..* 1..* requires collection & exploitation from queries 1..* provides intelligence to supples target information to 1..* supplies exploited product to provides collected information to provides information to 1..* 1..* tasks requests information from 0..* provides intelligence to tasks provides intelligence to asks question of provides intelligence to uses interrupts directs informs contracts collection & exploitation to informs informs 1..* asks question of 0..* carries

• ut

carries

• ut

Intelligence Requirements Management Collection Sensor Management Support Element Analysis Team Collection Coordination Intelligence Processing

19

Instantiating the I2 Model at Organisational Levels

• ut

• ut

• ut

• ut

• ut

• ut

JTF HQ Bde HQ BG HQ

Stimulation: 1. Research Tasks (background) 2. RFIs

20

Results (1) – Resource Usage over time

1 2 3 4 5 6 7 8 9 5150 10300 15450 20600 25750 30900 36050 41200 46350 51500 56650 61800 66950 72100 77250 82400 87550 92700 97850 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 AnalysisCoordinator Bde IntelligenceAnalyst Bde 1

Simulation Time Resource Level

1 2 3 4 5 6 7 8 9 5150 10300 15450 20600 25750 30900 36050 41200 46350 51500 56650 61800 66950 72100 77250 82400 87550 92700 97850 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 1E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 2E+05 AnalysisCoordinator Bde IntelligenceAnalyst Bde 1

Simulation Time Resource Level

21

I2 Summary

Achievable

• Initial analysis completed by G2.0

Beneficial

• Informed understanding of

the problem space

• Informed architecture
• Informed CONOPS

Valid

• Some validation applied to

the model

• Further V&V activity to be

applied after the experiment

22

Theme Modelling Evaluation (1)

• Add value to a WFE using dynamic modelling

– IADS – scenario assessment – IADS – bandwidth assessment – I2 – utility of new ways of working

• De-risk critical areas of the experiment

– IADS – key architecture questions exposed early in lifecycle – IADS – key scenario issue understood early on

• Provide a ‘dynamic data repository’

– Dynamic models shared with broader OA community – problems in commonality of tools

23

Theme Modelling Evaluation (2)

• Aspiration: Assess in advance what benefit modelling may

provide Effect of new ways of working Dynamic effects on system loading ‘integrated understanding’ of problem space, dynamic assessment of system loading I2 Scenario Bandwidth not major problem Equipment vrs process Link 16 bandwidth analysis, assessment of Δt for Human in- the-loop IADS Key outputs (Post Activity) Anticipated purpose (Pre- Modelling) Theme

24

Theme Modelling Evaluation (3)

• What not to model

– Very simple problems – few parameters, limited problem space – Problems with insufficient data – Modelling humans

• When to proceed with caution

– Where there is insufficient knowledge of problem space – modelling can become an activity in its own right rather than a support to WFE

25

Conclusions

• Pre experiment modelling must be done rapidly
• Essential components for rapid development of

dynamic models:

– Relevant Model Repository – Pool of skilled analysts/developers – Regular review

• Can we realistically assess benefit before the

modelling takes place?

– No – but we know about some of the things we should not do

26

Questions