1 Evaluating a Complex System of Systems Using State Modeling and Simulation National Defense Industrial Association Systems Engineering Conference San.

1

Evaluating a Complex System of Systems Using State Modeling and Simulation

National Defense Industrial AssociationSystems Engineering Conference

San Diego, CaliforniaOctober 20-23, 2003

Dennis J. Anderson*, James E. Campbell, and Leon D. Chapman

Sandia National LaboratoriesP.O. Box 5800

Albuquerque, NM 87185-1176

*(505) 845-9837, [email protected]

Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company,for the United States Department of Energy under contract DE-AC04-94AL85000.

2

Need for System of Systems (SoS) Evaluation

• Evaluating design concepts for complex systems of systems is required for Army transformation and envisioned military systems like

– Future Combat Systems (FCS)

– Objective Force Warrior (OFW)

• From conceptual design to production, SoS analysis will be critical to achieving individual system, and SoS, performance objectives

3

Problem

• Systems of systems characterized by complex combinations and interdependencies of technologies, operations, tactics, and procedures

• Evaluation of a SoS presents unprecedented challenges in– Exploration and analysis of multidimensional trade spaces

– Predict performance across multitude of design and technology options

– Performance characterized by several measures of effectiveness (MOEs)

– Improve and optimize mission effectiveness across wide parameter spaces

• Analyzing performance of several design options of a complex SoS across external parameters and multiple MOEs can generate a massive number of trade space combinations to be assessed, presenting extreme computational issues

4

DARPA IDEAS Future Combat System (FCS) Project Focused on Analysis of Multiple MOES across Large Trade Spaces

Effect

Sense

Communicate

Move

Protect

Command & Control

Functional View

5

FiresUnit Cell

MVRUnit Cell

MVRUnit Cell

FiresUnit Cell

MVRUnit CellFires

Unit Cell

MVRUnit Cell

FiresUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit CellMVR

Unit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

MVRUnit Cell

FiresUnit CellFires

Unit CellFiresUnit Cell

MF LOS/BLOS

C2

RSTA

RSTA

Multi-functionalRobotic Vehicle

MF BLOS/NLOS

MF LOS/BLOS

INF Carrier

MF RoboticVehicle/Sensor

•RSTA Vehicles with UAV controls all organic sensors

•C2 Vehicle command and control unit cell and link to Unit of Action

•Multi-functional (MF) Vehicles Able to fire LOS, BLOS, NLOS

•Infantry Carrier Vehicles for dismounted action and protection

•Multi-functional Robotic Vehicles unmanned ground sensor, unmanned Net Fires (BLOS/NLOS)

Notional FCS Maneuver Unit Cell

Colonel Peter Corpac, April 3, 2001 Deputy Director, Depth and Simultaneous Attack Battle Lab

6

FCS Reliability Analysis Results

7

FCS Spare Parts Optimization

• Minimal logistics footprint required for FCS

• Optimal spare parts determined to minimize downtime for set cost of inventory– Cost in terms of both $ and space

50

60

70

80

90

100

110

0 500 1000 1500 2000 2500 3000 3500 4000

Volume of Inventory (cu ft)

Do

wn

Tim

e (h

ou

rs)

50

60

70

80

90

100

110

$0 $200,000 $400,000 $600,000 $800,000 $1,000,000

Cost of Inventory ($)

Do

wn

Tim

e (h

ou

rs)

8

Internal Investment in System of Systems (SoS) R&D

• Nearly $1M investment in FY03-FY04– Extending SoS methodology– Extending existing tools

• R&D focusing on SoS challenges– Multiple MOEs– Multiple system states – Optimization of multiple MOEs across massive

trade spaces– Large number of systems (UA ~700 platforms)– Massive redundancy– Efficient analysis of multiple scenarios

9

Current Platform, FoS, & SoSModeling Approach

10

AG (NLOS-C) & Comp-C2 Models

11

AG (NLOS-C) & Comp-C2 Example Results

12

Optimization Input

13

Optimization Objectives

14

Optimization Results

15

Summary Optimization Results

16

Time Simulation Software Object

• Developing simulation tool for modeling large number of platforms

• Each platform is an individual object– Object is a collection of elements such as:

• Subsystems• Components• Failure Modes• External Condition states …

– Object can have multiple functions:• Mobility• Communications• Sensing• Firepower …

– Object provides:• Real-time status of any MOE• Probability of maintaining MOE to end of mission• Most likely problem areas• Simulation statistics …

– Object is a state model

17

Battalion Structure

18

System Elements – Repair in State

19

System Elements – Repair at Location

20

C2V C4 Function Redundancy

21

Spares

22

External Conditions

23

Ground Vehicle Scenario

24

Air Vehicle Scenario

25

Simulation Time-Step Output

26

Mission Required Vehicle Probabilities

27

SoS Methodology• SoS assessment methodology based on:

– Previous FCS SoS assessment programs for DARPA and JVB– Internal SoS modeling and analysis research program– Extension of Sandia suite of RAM modeling, analysis, and optimization tools– Continued development of state modeling tool

• Models multiple MOEs• Supports optimization across multiple platforms and multiple MOEs• Generates time simulation software object

– Each platform is a state model object– Each state model object provides

• Real-time status of any MOE• Probability of maintaining MOE to end of mission• Most likely problem areas• Simulation statistics• Handling of on-board spares

– Development of time simulation tool for modeling large number of platforms• Incorporates state model objects into time-simulation environment• Creates and duplicates multiple platform types• Describes MOE/functional areas for each platform type• Scales up to large number of systems• Describes scenario conditions

• Goal is to develop SoS Modeling and analysis suite that integrates state modeling with Sandia RAM toolset and time simulation

28

Next Generation Analysis Suite

Fault Tree EditorMultiple ModelsMultiple MOEs

Data Library EditorManage Data for Fault Trees, State Models,

And Simulation

Results ViewerView Statistical Results

From Fault Tree or State Model Analysis

OptimizationOptimize Spares Inventories

Optimize Multiple MOEs And Multiple Platforms

State Modeling ToolSingle Model

Multiple MOEs

SimulationMultiple Platforms

Multiple MOEsExport Models

To Simulation

Export Models

To Simulation

29

Backup

30

Modeling & Simulation

Design for Reliability / Maintainability

Optimization/Genetic Programming

Prognostics & Health Management

Automated Assembly/Disassembly

Supply Chain Management

Spares Inventory Optimization

Technical Risk Management

Sensitivity / Uncertainty

Quantification

Human Factors Engineering

Tools & Technologies Validated Through Broad Use

Technologies and Customer Base in Supportability

31

Optimization ModelingExample Output

SystemModel

SystemModel Optimization

Module

OptimizationModule

Our Optimization Modeling Supports all Aspects of the Life Cycle

Our Optimization Modeling Supports all Aspects of the Life Cycle

Modeling ToolsModeling Tools

•Fault Trees/Block Diagrams•Discrete Event Simulation•State Space Modeling•Agent-Based/Object Oriented•Finite Element

.

.

.

•Fault Trees/Block Diagrams•Discrete Event Simulation•State Space Modeling•Agent-Based/Object Oriented•Finite Element

.

.

.