A Protocol for the Use of Simulationbased Analysis by Decision Support Systems on the Global Informa

1
A Protocol for the Use of Simulation-based
Analysis by Decision Support Systems on the
Global Information Grid
05 April 2006 Wayne Smith, ERDC-CERL
2
Problem

• Simulations frequently used to support decisions
• Global Information Grid (GIG) Service Oriented
  Architecture supports assembling services
• No known protocol for communication between
  Decision Support Services (DSS) and Simulation or
  Analysis tools
• Without a protocol, transfer of criteria
  information to a DSS is a manual and error prone
  process

3
For Example - Installation Analysis Problem
Range Risk and Encroachment Tools
Power Projection Planner
What if? Study options in the Virtual Installation
Installation Planning
Force Protection Tools
4
Hypothesis

• There exists a common and abstract control and
  data schema that can be used by a DSS and
  simulations to communicate, given a suitable
  protocol.

5
Approach

• Define system endpoints for GIG
• Determine data requirements for common DSS
• Create abstract control and data schema
• Create communication protocol
• Conduct proof-of-concept test of protocol and
  schema using a simple DSS with multiple
  simulation systems.

6
System Endpoints

• Networked
• GIG-like system
• SOAP-based communication

7
Abstract Concepts

• Study
• Stakeholder
• Scenario
• Event
• Dataset
• Alternative
• Action
• Plugin
• Goal
• Criterion
• Criterion Value

8
Simulation System Meta-data

• Simulations register with DSS using XML
• DSS can call PluginAdapter methods
• PluginAdapter has access to study data
• PluginAdapter formats input for sim, controls
  execution, and retrieves results using SOAP XML
• DSS compares criteria and alternatives

9
PluginAdapter Methods and Protocol

• For each remote plugin and study, DSS uses SOAP
  call to
• Initialize service to run a job for all
  alternatives
• Create inputs
• Start jobs
• Get status
• Get results
• Tell service to delete data
• Optional
• Abort all or each job

10
Fort Future Deployment Example
11
Creating A Study
12
Setting The Study Scope
13
Setting Goals and Criteria
14
Events
15
Setting Up Alternatives
16
Deployment Alternatives Editor
17
Managing Simulations
18
Setting Simulation Service Parameters
19
Running Simulations
20
Viewing Simulation Results
21
Protocol Supports Simulation-Specific Results
22
Other simulations were tested
23
Summary and Future Work

• Proof of concept for a DSS-Simulation protocol
  was successful
• XML-based registration could be made more dynamic
  using WSDL
• May be possible to eliminate plugin adapter by
  requiring services to provide editing capability
• Protocol could be incorporated into MSDL as
  meta-control language
• With Fort Futures pluggable architecture, new
  simulations can be added at any time, and
  different simulations can be run within the same
  study

24
Questions?

• Development is ongoing stop by periodically for
  updatesFort Future Home Page
  http//ff.cecer.army.mil/ff/home.do
• Contacts Dr. Michael Case Michael.P.Case_at_us.arm
  y.milWayne Smith Wayne.R.Smith_at_erdc.usace.army.m
  il