SimNavy - An Enterprise Model of the US Navy

Transcription

1 SimNavy - An Enterprise Model of the US Navy Naval Postgraduate School & Thinking Tools Inc, Monterey Technology Initiative Game (TIG) September 1998 Fort Belvoir,, Virginia Outline Enterprise modeling & Agent-Based Adaptive Simulation Existing Enterprise Models Demo Building an Enterprise Model for the Navy SimNavy - A Proposal SimNavy Mock-up 1

2 Enterprise Modeling & Agent-Based Adaptive Simulation Michael Zyda Naval Postgraduate School Interactive Simulation Interactive simulations have at their base actors and interactions. The actors can be atoms, fish, people, organizations, nations... Their interactions may consist of attraction, combat, mating, communication, trade, partnership, rivalry... 2

3 Interactive Simulation Because the study of large numbers of actors with changing patterns of interactions often gets too difficult for a mathematical solution, a primary tool is often interactive simulation. The trick is to specify how the agents interact, and then observe properties that occur at the level of the whole. Interactive Simulation For example, with given rules about actors and their interactions, Do the actors tend to align into two competing groups? Do particular strategies dominate the population? Do clear patterns of behavior develop? (i.e., emergent behavior) 3

4 Interactive Simulation The simulation of agents and their interactions is known by several names but we will use agent-based, adaptive simulation Whatever the name used, the purpose of agent- based simulation is to understand properties of complex systems through the analysis of interactive simulations. Agent-Based Adaptive Simulation Agents are independent software objects that adapt to changes in the environment by altering their behaviors. Many agents interacting with each other produce what is called emergent behavior. Emergent behavior has many of the characteristics of natural complex systems - including unpredictability. 4

5 Some definitions Object - A way to organize the code and data of a software program. An object is a self-contained capsule of data and logic that is an independent piece of the program. Agent - An object designed to take action on its own. An agent selects its action using information it chooses from a range of available actions. Some definitions Autonomous Agent - An agent that selects its behavior with the goal of achieving a better fit to the requirements of its environment. Adaptive Agent - An agent that creates new behaviors in a non-random way with the goal of achieving a progressively better fit to the requirements of its environment. Both adaptive and autonomous agents can be used to produce adaptive simulations. 5

6 Adaptive simulations Adaptive simulations customize the learning experience by allowing the simulation to discern each user's vulnerabilities and then confront that user with simulated events that focus on his or her weaknesses. This hostile environment introduces the challenges that a particular user needs most to understand. Adaptive simulations Training based on adaptive simulations emphasizes diagnosis, planning, and problem-solving skills rather than rote learning or routine/automatic responses. A technology bibliography for agent-based adaptive simulation is included in the paper submitted for the TIG. 6

7 Existing Enterprise Model Demos John Hiles Thinking Tools, Inc. Historical Interactive Simulations SimRefinery TeleSim Project Challenge... 7

8 Demonstrations of agents at work MarketScape MalionShield Transport Building an Enterprise Model for the US Navy Dr. Allen Zeman N7 8

9 Wouldn t it be nice... To have an enterprise model of the US Navy that reflected the IWAR concepts... IWAR STRUCTURE Navy Integrated Warfare Architecture = TOA $ Maritime Dominance USW Superiority - ASW - Mine Warfare SUW Superiority Deterrence Strategic deterrence Counter WMD Forward Presence and Engagement Sensors & Information Superiority C4 ISR Information PED IW/IM Power Projection Strike Warfare Littoral/ Expeditionary Warfare - Naval Fire Support - STOM Air Dominance Air Superiority Missile Defense - TBMD - Cruise Missile Defense Sustainment Infrastructure Manpower & Personnel Readiness Training & Education Technology Force Structure CLF MSC Service-wide Transportation BOS Force Protection MILCON, FHN, Environment BRAC Admin Spt MPN/RPN QOL Recruiting Manpower Spt Health Affairs OPTEMPO FHP Maintenance Spares Operational& Tng Spt Inacts Individual Tng Ranges/Targets NPS RDTE (6.1/6.2) SCN APN 1-4 Assoc RDTE 9

10 SimNavy - Phase 0 Michael Zyda Naval Postgraduate School 10

11 SimNavy Concept - Phase 0 Goal Design a software program called SimNavy to simulate the operation of the Navy. This program will operate in a manner similar to Maxis SimCity,, with a dramatic computer-game interface, but with much higher resolution and Navy- relevant scenarios. SimNavy Concept Simulate the OPNAV/N-code operation of the Navy, including: Resource allocation The psychology of decision making The zero sum economy Multilateral decisions and constraints. The conflict between political process and military requirements. The dynamic nature of decision making. 11

12 SimNavy Concept All assumptions observable and changeable. CNA and IDA provided analytical equations and models will be utilized, as available. Simulate negotiation, compromises, tradeoffs, frustrations and anxieties of top-level policies. Expose motives and goals of the decision making process. SimNavy Concept Examine what-if scenarios and alternate outcomes. Include the Navy s interactions with the other services and political organizations. Provide a high-spark, controversial platform for understanding problem solving in the Navy hierarchy. Use NPS MOVES students and faculty to develop and fill in the underlying models from CNA/IDA data. 12

13 SimNavy s Value To provide an interactive experience through which Naval officers can learn how to make decisions at the various levels of the Navy s hierarchy. It should be understood by all that this is not a simulation for Navy war-fighting or tactics but rather a tool to examine the relationship between budget allocation and operational possibilities. Proposed Technologies for the SimNavy Project Agent-based adaptive simulation (underlying basis for SimCity-like simulations) Other technologies as identified and appropriate. 13

14 14

15 Enterprise Model of the Navy Treat the enterprise as a system. Help users develop insight into the enterprise system. Highlight through hands-on learning the interrelationships between pieces. Provide a fast way to explore key trade-offs. Provide a clear view of dynamic behavior over extended periods. SimNavy Mock-up Interface Design Scenario Editor Assumption Editor Global Mission Display Budget Allocation Operations Control 15

16 Interface Design Scenario Editor Create libraries of possible simulations Multi-year outlook with multiple mission events Manual event creation Random staging of events Manual & Random Scheduling of Events Design of Mission Dimensions Missions Supporting Joint Operations 16

17 17

18 Assumption Editor Assumption designers create relationships between key variables i.e.: Capability - mission accomplishment Meta-assumptions Benign Realistic Malevolent 18

19 19

20 20

21 Global Mission Display Primary feedback and interaction mechanism World map Theater lines Events status Shown on situation map Clickable icons for levels of information 21

22 22

23 Budget Allocation Review current budget, and plan future expenditures Current and future budget choices Categories of resource allocation 23

24 24

25 Operations Control Tailor force structure to meet perceived requirements Administrative Control Existing Units or New Inventory screen for assignment of assets 25

26 26

27 SimNavy - Development Phases, Partners... Michael Zyda Naval Postgraduate School Recommended development phases Phase 0 (now) Develop vocabulary, architecture, and mock-up of SimNavy enterprise model. Phase 1 Single-role, single PC version, 12 to 24 months, depending on level of modeling. Phase 2 Multi-role version, 12 to 24 months, depending on level of modeling. Phase 3 Multi-role, multi-station version, runs on the Internet, 12 to 24 months, depending on level of modeling and networking desired. Variants to support special planning or analytical needs. 27

28 Partners for Phase 0 NPS MOVES Research Center Thinking Tools, Inc. thinkingtools.com Center for Naval Analyses NPS MOVES Research Center The MOVES Research Center is comprised of faculty of the NPS MOVES Academic Group, a multidisciplinary faculty collection focused around providing support for N6 on its mid-term and long-term issues in modeling, virtual environments and simulation. 28

29 NPS MOVES Research Center The MOVES Research Center has evolved from the NPSNET Research Group, a research group of faculty and officer students with more than twelve years experience in research and development on the large-scale, networked virtual environment. npsnet.nps.navy.mil 29