military systems analyses in an ill-posed world: … › ndia › ...1/32 paul h. deitz, ph.d. us...

1/32

Paul H. Deitz, Ph.D. US Army Materiel Systems

Analysis ActivityAberdeen Proving Ground, MD

paul.h.deitz.civ@mail.mil443-421-0039

The National Defense Industrial Association Annual National

Test & Evaluation ConferenceThe Westin Washington

July 22-23, 2014

17JULY2014/15:15

Military Systems Analyses in an Ill-Posed World: Defining the Problem

DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited.

2/32

Outline

The intersection of “Ill-Posed Problems” and military systems analyses

Some key problem examples

An abstraction consisting of task networks which redress the stated shortcomings

Elements of the task network abstraction

Follow on presentations by Britt Bray (Engility), instantiating such a structure, and Chris Wilcox (ATEC/AEC), showing how task networks can inform Developmental and Operational Testing

3/32

Introduction [1/3]

What are Ill-Posed Problems?• Problems which are under-specified, under-determined, or

under-constrained

Such problems typically arise when attacked with ad hoc methods

Some exemplars (illustrative vice exhaustive) of requisite context specification in military analysis:• Key mission linkages across levels of war• Specification of relevant metrics• Accommodation of dynamic entities• Performance and effectiveness characterization of System-

of-Systems (SoSes)• Integration of materiel and human factors

4/32

Need for overarching analytic structure/framework:

Introduction [2/3]

• If not provable, should be plausible• Captures complete mission context• Supports formal language

• Not ad hoc • Not bottom up• Not simply a recipe

5/32

How do warfighters/operators describe missions?• They use the Military Decision-Making Process [MDMP]

• One of the outputs of the MDMP is a kind of Gantt structure – a time- and dependency-based ordering of tasks to be performed; available resources (effectors) with the capability to achieve desired task outcomes are assigned to tasks

• To avoid potential confusion over meaning, authoritative, doctrinally-based task language is now codified by formal Joint- and Service-Task Lists (w conditions and standards)

Introduction [3/3]

6/32

Example: Key mission linkages across levels of war

7/32

STRATEGIC THEATER

OPERATIONAL

TACTICAL

ConcentrateForces inTheater ofOperations

OP 1.2.3

Concentrate TacticalForces

BTT 1.2.1

Foster Allianceand RegionalRelations and

SecurityArrangements

ST 8.1

Cooperate Withand Support

NongovernmentalOrganizations

(NGOs) in TheaterST 8.2.11

Cooperate Withand Support

Private VoluntaryOrganizations

(PVOs) in TheaterST 8.2.12

CoordinateHumanitarian andCivic Assistance

ProgramsST 8.2.4

MOUT Mission Decomposition‡

Enemy team fires on disabled

vehicle fromchurch tower

U-ART 1.2.2.3.3.1

Gunner of disabledvehicle returns fire

on church towerw/ 25mm auto-gun

U-ART 1.2.6.1.1.2.1.1

Platoon leaderorders gunner

to cease fire onchurch

U-ART 1.2.6.1.1.2.1.2

Platoon leader callscompany commander reports incident and informs commander

of significant damage to church, and several

civilian casualties

Platoon leader receives order from company commander to

break contact with enemy

U-ART 1.2.6.1.1.2.1.4U-ART 4.5.1.1.1

Platoon leader orders

evacuationof casualties

U-ART 1.2.6.1.1.2.1.3

Formulating a Mission

Horizontal (Time) Linkages

STRATEGIC NATIONAL

Top-Down Linkages

‡ Mission build by Dynamics Research Corporation, Nov 2000

Strategic Context

Constrains Materiel Options

Tactical Context Major

Determinant in Material

Survivability

SN 1

Support PeaceOperations

SN 8.1.3

Conduct ForeignHumanitarianAssistance and

Humanitarian andCivic AssistanceSN 8.1.5

Cooperate withand Support

NGO’s and PVO’s

SN 8.1.9

ConductStrategic

Deployment andRedeployment

FosterMultinational

andInteragencyRelations

SN 8

Logistical Support Affects

Mission Robustness

Jump Effect Jeopardizes Campaign

8/32

Mission context‡ and materiel behavior at all levels of war can, and do, effect mission outcomes at all

levels of war!

Key Observation

‡ E.G., accessibility of materiel to combat zone, global politics, logistics posture, threats, cultural/political environment, . . .

9/32

Task networking at a single level of war

10/32

The Beginnings of Task Execution in V/L Analysis

Describing ballistic events at the

Microscale of combat

Proper characterization of effectiveness was an

open question!

11/32

Task Cycle

Tasks

Interactions

Components

Capabilities

Task Cycling via MMF, circa 2001

Time-Forward,

Bottom-Up Execution

Task Cycle

Results of Bottom-Up task execution are assessed using measures of effectiveness derived during Top-Down mission planning Task execution generates interactions. The resulting

effects are compared with the plan, and they drive selection of follow-on tasks to form cycles Execution process is Time-Forward and Bottom-Up

12/32

Developing Relevant Metrics Top Down

4. Tasks, Operations

1. Interactions,Effects

2. Personnel, UnitsComponents, Systems

3. Functions,ServicesTime-Backward,

Top-Down Inferencing

Relevant Metrics by Top-Down Inferencing Define task(s) included in operational plan Infer requisite Services/Capabilities (through mission

analysis and wargaming Identify critical entities (people/material), and Identify critical Interactions to avoid/enhance (through

wargaming and rehearsal)

13/32

Interactions

• Ballistic Effects• Jamming• Damage Repair• Chemical• Resupply• Repair• Laser Damage• Sleep• Directed Energy• Nuclear• Physics of Failure• Logistics Burdens • Reliability• Fair Wear & Tear• Fatigue• Heat Stress• . . .• . . .

Dynamic Geometry/Material [1/3]

Many interaction mechanisms can induce change (damage/fix) in entities (people/material).

14/32

Component Networks

Dynamic Geometry/Material [2/3]

15/32

Interactions

A sequence of interactions is tracked at Level 2 At any given time, the

component state space is mapped to Level 3 to assess capability Then mapped to Level 4 to

assess mission effectiveness

Dynamic Geometry/Material [3/3]

Sequential interactions must integrate at Level 2. Then evaluated for:

Performance and Effectiveness

16/32

Adding: a] Context for Single Level of War, and b] an OPFOR

An Opposing Force was added

Task Path Variants

17/32

6. Context, Environment (Military, Civil, Physical, etc.)

7. OPFOR Why = Purpose, Mission7. OWNFOR Why = Purpose, Mission

The MMF Diagram

11 Fundamental Elements: 7 Levels, 4 Operators

5. Index: Location& Time

O1,2O1,2O2,3

O3,4 O3,4

OWNFOR OPFOR

O2,3

O4,1

2. Personnel, UnitsComponents, Systems

1. Interactions,Effects

3. Functions,Services

7. Mission

4. Tasks, Operations

3. Functions,Services

7. Mission

4. Tasks, Operations

2. Personnel, UnitsComponents, Systems

O4,1

All of the underlying layers are about Context!

18/32

Task Prosecution and DT/OT

In Developmental Testing:• Identify expected interaction mechanisms based on mission

profile • Confirm ability to model/simulate via test

In Operational Testing and M&S:• Vary the time sequence and frequency of interaction effects

based on mission profile

Interaction Mechanisms

All Interaction Mechanisms must focus on component state change

Aggregation of multiple mechanisms (over time) must take place at the platform component level

19/32

Task Network Abstractions In the Macrospace

Combat Abstraction by Level-of-War

Combat Abstraction by MDMP/MMF

Combat Represented by Physical Layers

20/32

Single Platform

Microspace

Platform Physical Networks

Task Abstractions For a Single Platform

Macrospace

21/32

Task Abstractions for Multiple Platforms

Macrospace

● ● ●

Platform 1

Platform 2

Platform p

SoS

Microspace

Multiple Platforms Multiple Platform Physical Networks

Collective Tasks

Key Construct!

Col

lect

ive

Task

s

22/32

TIME

OWNFOR Material Connects to Task Network

23/32

TIME

OPFOR Material Connects to Task Network

24/32

OWNFOR and OPFOR Interact over Time

Networks

Red on Blue Task

Interactions

Blue on Red Task

Interactions

25/32

The connection between Materieland Human Performance

26/32

1980 1990 2000 2010

1980 1990 2000 2010

Origins/Applications of Military Task Formalism

Joint/ Service

Task Lists‡

Human Factors

Community

Ballistic Vulnerability Community

Microtasks

Macrotasks

Microtasks & Macrotasks

‡ In support of the MDMP

IMPRINTHARDMAN

Blueprint of the

Battlefield

MMFV/L Taxonomy

27/32

Task Networks Infuse All Structures [1/2]

1980 1990 2000 2010

1980 1990 2000 2010

IMPRINT

MMFV/L Taxonomy

Joint/ Service

Task Lists‡

Microtasks

Macrotasks

Microtasks & Macrotasks

‡ In support of the MDMP

HARDMAN

Blueprint of the

Battlefield

MMF

Joint/ Service

Task Lists‡

MMF

28/32

Task Networks Infuse All Structures [2/2]

1980 1990 2000 2010

1980 1990 2000 2010

IMPRINT

MMFV/L Taxonomy

Joint/ Service

Task Lists

Microtasks

Macrotasks

Microtasks & Macrotasks

HARDMAN

Blueprint of the

Battlefield

MMF §

IMPRINT §

§Mission Centered Human System Analysis, Diane Kuhl Mitchell [Army Research Laboratory/ HRED] and Gene Brennan [Alion Science and Technology], 19 March 2008.

29/32

Summary [1/3]

Many ORSA problems are Ill-Posed

Using the MDMP:• Task networks [OWNFOR and OPFOR] linked by level-of-war/time

• Interactions may occur between all platforms; need to accommodate dynamic component geometry.

• Many possible networks, e.g. digital, mechanical, electrical, hydraulic, optical/visual, acoustic, . . .

Mission success rolls up from task performance at DT/atomic level to task results at OT/collective levels

30/32

Summary [2/3]

Top-down inferencing of task cycles provides the relevant measures of success, capability and material properties.

Task interactions, both destructive and constructive, occur between both friendly and opposing forces.

The methods employed by the Human Dimension community [IMPRINT] are fully integrable into the MMF construct.

The fate of one platform can potentially affect any other.

31/32

Summary [3/3]

The effect of platform change depends on mission context. Criticality of platform damage/dysfunction is determined by closest logical proximity.

The MDMP, though ubiquitous among Warfighters, is seldom used by the ORSA Community (e.g. in DT, LFT&E, OT, Mission-Based T&E) or to provide links to Human Factors analysis.

Both the Macro and Micro worlds would be far richer were their task network analyses integrated and linked!

32/32

END