Presentation is loading. Please wait.

Presentation is loading. Please wait.

A Game Loop Architecture for the Modeling and Simulation of Mission Threads Tom Tanner 540 663-4867 (Office) 540 903-6713 (Cell)

Similar presentations


Presentation on theme: "A Game Loop Architecture for the Modeling and Simulation of Mission Threads Tom Tanner 540 663-4867 (Office) 540 903-6713 (Cell)"— Presentation transcript:

1 A Game Loop Architecture for the Modeling and Simulation of Mission Threads Tom Tanner thomas.l.tanner@saic.com 540 663-4867 (Office) 540 903-6713 (Cell) thomas.l.tanner@saic.com

2 Outline Objective –To promote the viability of the game loop architecture as a means of modeling mission threads –Required elements of proof Organizational relevance –Impetus for promoting acceptance, assimilation and socialization –Preferably a key role within a current strategy, initiative or movement Added value Achievable and affordable Presentation –Alignment with Systems 2020 –Mission Threads - Mission Thread Models Brief (localized) description Their role in the mission systems engineering process (the analyst’s perspective) –The Game Loop Architecture Mission centric vs. system centric approach Components Game loops – Example “Swarm Raid” mission thread from CGX program Detailing the game loop –Transition from wireframe to system design –Summary and Recommendations

3 Mission Threads in Support of Systems 2020 Closing The “Churchillian Delta” –Looking back at the lessons of WWII Sir Winston Churchill wrote: “It's not enough that we do our best; sometimes we have to do what's required.” –Initiatives that matter target the delta between “our best” and “what’s required” –In a plenary address at the NDIA Systems Engineering Symposium Kristen Baldwin framed the Churchillian Delta that is driving Systems 2020 as “DoD structures and business practices are not structured for adaptability, need or opportunity” Situational Dependence –Adaptation is a response to situation –Need and opportunity are both created by situation –Mission Threads provide a useful mechanism for establishing situational context

4 Setting Situational Context With Mission Threads Nation Interests Projected Capabilities Opposing Interests Plausible Capabilities Technology Innovation Situations are likely to be “crafted” by the opposing interests Mission Scenarios must be an integral part of the process. Ideally they should span the limits of plausibility. Collectively they must demand a broad range of capabilities. Mission scenarios must be played out at the enterprise level. “Free play” is a seminal precept of valid scenario modeling. Adaptability Need Instruments of National Power Opportunity DoD Structures and Business Practices Systems 2020 Targeted restructuring The Game Loop approach brings economy of scale to the modeling, simulation and analysis of mission threads Situation Driven Attributes Can Not Be Situation Limited Mission Threads provide a natural interface Mission Threads provide a natural context Affordable scenario based modeling is key

5 Localizing Systems 2020 Goals Rapid Capability Toolbox Study Final Report, March 2010 (Cited as Key Study in Systems 2020)

6 Mission Threads (Quick Look) Analytical Context –In this discussion “analysis” refers to the assimilation of information into a body of knowledge that will serve as the basis for decisions –Analytical Context is the basis and the principle mechanism of the assimilation process Scenario Based Context Provides: –A roadmap for mutually insightful discussions between operators and designers –The basis for assessing the mission system causation chain performances attributes capabilities effects objectives Mission Threads (as a type of scenario based context) –Qualitative, sequential representation of a mission scenario –A rapid means of framing the mission from end to end –Typically have prescribed branching decisions to ensure a productive trajectory –Ideal infusion point for “what ifs” –Well suited to requirements definition and design guidance –Useful as precursors of test and analysis plans –Current efforts are aimed at making them more quantitative –Agility and rapidity of development do not transfer to M&S

7 Analytic Life Cycle Premise 1. There is nothing to model Premise 2. Pre design models are “throwaways” RigorousFranticAd Hoc Supporting M&S Perfunctory Pre Design Construction In-service Support Design Concept Exploration & Engineering Design Verification Design Guidance Proactive Adaptation Reactive Adaptation Analytical Phases Analytical Context VagueDeterminedDynamic Supporting M&S “Greatest leverage in the ‘front’ end of the life cycle”

8 DeterminedDynamic Mission Threads? Mission Threads in the Analytic Life Cycle RigorousFranticAd Hoc Supporting M&S Pre Design Construction In-service Support Design Concept Exploration & Engineering Design Verification Design Guidance Proactive Adaptation Reactive Adaptation Analytical Phases Analytical Context Mission Thread M&S Insightful Evolvable Codified Mission Threads Test/Analysis Plans

9 Provides objectives and requirements for early model Pre design models are wireframes for later M&S Dynamics of Mission Thread Modeling Mission Thread M&S Mission Threads Test/Analysis Plans Actors Scenario Based Contexts CONOPs - Requirements Behavioral Models Storyline 1 Mission threads establish the trajectory for the development of M&S 1 Performance/ Capability Driven Emulation Deterministic Branching Enterprise/ Physics Driven Simulation ; Dynamic Branching Free Play Model Maturation Requirements Allocation/ Assessment Design Guidance & Performance Leveling Design Assessment Verification Intended Use Progression 2 Data Test Articles Supplements 3 Usage and capability of the M&S co-evolve 2 Evolution of mission thread M&S parallels the needs of testing and analysis 3

10 Mission Thread Modeling Payoffs –Requirements bridge between analytical needs and M&S capabilities –Scoping mechanism for self contained, piecewise analysis of a complex mission space –Continuity of analysis, design and test cases –Provides a basis for confidence based on an understanding what the M&S is doing Making it Work –Experience suggests incremental development starting with a wire frame of essential features using place holders if needed –Simulation must be modular, agile, expandable and readily evolvable –Host structure must support large scale composability of an eclectic assortment of component models –Must be functional at the front end of a project The Game Loop Architecture

11 System Centric Two Approaches to Mission System Modeling Blue Enterprise Red Enterprise Environmental Entities Environmental Effects Adjudication Omniscient Omnipotent System of Interest Model the system and wrap the war around it Model the war and put the system in it Logical appeal in that it leverages the representation of the system of interest Ideally suited to design assessment Simulation lags design – limiting usefulness for design guidance Invites bias in free play scenarios Decentralized (inconsistent?) adjudication Simulation can run prior to system design (using place holder if required) Consistent with the charter of the suggested “Concept Engineering Center” Mitigates ripple effect induced by changes to the representation of the system of interest Localizes branching decisions to adjudication Mission Centric Blue Enterprise Red Enterprise Environmental Entities Environmental Effects Adjudication System of Interest Adjudication Context for Game Loops

12 Game Loop Architecture Players –Mutually incompatible objectives –Capabilities that can influence the end point of the “game” –Represented by quasi autonomous objects Rule Set and Adjudication Mechanism –Rule Set can be combination of convention and “physics” based –Centralized adjudication component Fields adjudication requests from players Promulgates results to effected players Venue –Common Operational Environment –Optional “game board” Game Driver –Consistent with quasi autonomous objects –Centralized – Event Queue enforces temporal causality (event drive with time driven and agent based tendencies) –Decentralized – Individual Game Loop Legend Problem Domain Core Architecture This combination is a key contributor to “agile, expandable and readily evolvable” The underlying goal is to design mechanisms that make the most of one time investments in both the core and in the problem domain.

13 Basic Game Loop Model The basic game loop model can be applied any game Sport, Card, Board, Wargame, Combat … One loop per player (players of similar type can share the loop implementation) Situations are based on a shared Common Operation Environment (COE) Can be implemented as an augmented OODA loop Assess Situation Select/Modify Strategy Execute Strategy Update Situation Game Loop Perception may be omniscient or limited / distorted Compare perceived state and trend to near term and long term objectives Identify desired effects and formulate/select means of realizing them Identify changes that are beneficial and feasible Activation and vectored application of power Determination of aggregated effects Determination of updated state parameters Determination of aggregated effects Determination of updated state parameters Assigned to Player/Entity Assigned to Adjudicator

14 The Modified Game Loop Assess Situation Select/Modify Strategy Execute Strategy Update Situation Game Loop Perception may be omniscient or limited / distorted Compare perceived state and trend to near term and long term objectives Identify desired effects and formulate/select means of realizing them Identify changes that are beneficial and feasible Activation and vectored application of power Determination of aggregated effects Determination of updated state parameters Kinematic Loop Adaptation- Influence Loop In simulations of mission systems the speed of the reaction, adaptation process has tactical significance. The game loop can be implemented as a kinematic loop synchronized with the “game clock” and an adaptation influence loop operating a pace driven by the scenario. Inter-player synchronization

15 Conducting the Game Set Up Each player’s starting state (position, various load-outs …) is defined in a file. At least one player or group of players will be assigned an attack profile. Periodic checks of objectives and conditions are part of each players game loop An attack profile may have several phases with conditions and objectives. A player or group may advance to the next phase when the objectives of the current phase and the conditions of the next phase are satisfied Attack Phase 1 … Attack Phase N Objectives Attack Phase 2 ObjectivesConditions Attack Phase 3 ObjectivesConditions Objectives A given attack phase can be implemented once and used in may different mission threads Each attack phase knows what parameters define its conditions and objectives Threshold values for conditions and objectives are specified in the setup of each mission thread Run Each player acts as directed by its own game loop, pursuant to its assigned role in the mission (may be prescribed by attack profile, may be to loiter on station). Adjudicator has a list of all players and can access all pertinent information. Players assess the situation by telling the adjudicator where and how they are looking. The adjudicator then responds telling them what they can see (possibly with random error). Engagements are adjudicated in similar fashion. Modeling a mission thread is a matter of initializing each of the players, including roles and objectives to varying degrees; and hitting run. (May require the creation of new player types or attack phases)

16 Example (Pilot Project) Swarm Raid Mission Thread from CGX Check for close CPAs Monitor position and velocity relative to station Calculate vectorgram of solutions to the relative motion problem Select course and speed for quickest stationing subject to collision avoidance Calculate engine and rudder orders Execute kinematic model Assess Situation Select/Modify Strategy Execute Strategy Update Situation Game Loop Determine readiness to activate next phase of attack Monitor status of attack plan( time position cover) Promulgate current phase to group Game Loop for attack boats Game Loop for Leader Attack Boat Game Loop This scenario pits CGX against 15 small, agile, missile capable surface craft in 3 groups of 5 There is a vibrant fishing industry and a shipping lane off shore that will provide cover for the attack. This demonstration is tuned to finding exploitable indicators of a pending attack. This single scenario can be used to explore and assess all options in the detect to engagement sequence. Each entity (CGX, small boat, merchant, fishing boat, helo, UAV ) executes its own game loop. The attack profile consists of 6 phases for each of the groups. Each group has a designated leader that will decided when to proceed to the next phase based on achieving the objectives of the current phase and meeting the prerequisites of the next

17 Swarm Raid Mission Phases Transit PIM Get to specified point Formation PIM Get to specified point Random Loiter PIM Zero out Slack Time Ambient Firing Target Attack Target Close Distance to Tgt Stream Dash Initial condition - None Initial condition – Target & cover assigned Initial condition – Time & posit Initial condition – At launch range (time opt.) Terminal condition - PositTerminal condition - Time & posit Terminal condition - None Terminal condition - Time or posit Terminal condition - Posit Target Cover Ship Form Up – Mask Intent Covered

18 Swarm Raid Scenario Setup CL FT RT AL SDFD Sea Lane Hostile Port CGX Line of fire from ship likely to include shipping traffic Air assets can attack off axis Open range seaward could add options Difficult to recognize attack craft among fishing boats When do we have sufficient confidence? Launch surveillance helo/UAV? Launch attack helo/UAV? Open distance from sea lane? Call in SUW configured LCS? Can we effectively vector air surveillance assets to effectively ID suspicious water craft? What is the right load out of air assets? How many UAVs do we need to support simultaneously What does a medium range surface to surface missile system buy us? ASUW problem becomes AAW What are we willing to do to operate in this environment? What are the design options for CGX? Other options? Mission Thread Modeling promotes valuable questions. 100+ fishing boats in this area CL FT RT AL SD FD Formation Transit Random Transit Ambient Loiter Covered Loiter Stream Dash Firing Dash Can we recognize patterns that suggest hostile intent? Attack Phase Transition Point

19 Analytical Assessment Point Assess Situation Select/Modify Strategy Execute Strategy Combat Power Emulator Placeholders are useful for “closing the game loop” providing a complete but rudimentary set of capabilities player must have to participate in the game Integrated Combat Power Direct Indirect Unintended Stochastic influence Change of state Situational awareness Posture Readiness Situational awareness Combat potential Environment Adjudicated Effects Red directed effects Environmental factors (deterministic) End State Composite Operational Environment Measure of Mission Success Hosting the System of Interest (Closing the Game Loop – integration with the host) A player must be able to field sensor inputs from the adjudication component In response to adjudication request for active sensors Initiated by the adjudicator for passive sensors A player must have the ability to select a course of action A player must have a means of assessing the perceived trajectory of the game relative to a strategic objective A player must have the ability to activate or modify a course of action A player must have the ability to update its status (position, velocity, resources …) and forward prescribe data to the adjudication component (not shown on this diagram)

20 Analytical Assessment Point Assess Situation Select/Modify Strategy Execute Strategy Integrated Combat Power Direct Indirect Unintended Stochastic influence Change of state Situational awareness Posture Readiness Situational awareness Combat potential Environment Adjudicated Effects Red directed effects Environmental factors (deterministic) End State Composite Operational Environment Measure of Mission Success Hosting the System of Interest (Interim Hosting – The Apprenticeship Phase) Sensors Suite Resource Manager Track Manager TEWA Weapon Delivery Engagement Support C and D Engagement Control Placeholders can be modified incrementally to incorporate prototypes or equivalent representations of proposed system elements

21 Integrated Combat Power Direct Indirect Unintended Stochastic influence Change of state Situational awareness Posture Readiness Situational awareness Combat potential Environment Adjudicated Effects Red directed effects Environmental factors (deterministic) End State Composite Operational Environment Measure of Mission Success Hosting the System of Interest (Functional Hosting) Sensors Suite Resource Manager Track Manager TEWA Weapon Delivery Engagement Support C and D Engagement Control When the system is sufficiently mature and complete to satisfies the basic game loop requirements the original wireframe can be retired and the system refined incrementally

22 Summary and Recommendations Summary –Mission threads are effective and affordable analysis drivers in support of mission system engineering –Their utility as simulation drivers is an appealing but elusive goal –Mission threads and mission modeling are inline with and are potential enablers of the vitally important “Systems 2020” initiative –The game loop architecture is well suited to the needs and the objectives of mission thread modeling and simulation –The effectiveness of game loop approach grows with the breadth of the reuse library of entities, mission phases and associated constructs Opinion –We need the “Concept Engineering Center” recommended by the “Rapid Capability Toolbox Study” –Mission threads and mission thread modeling can provide structure and guidance during the Concept Engineering phase Recommendation –Develop a small nucleolus of expertise in the area of mission thread modeling by way of a suitable pilot program –Refine the concept of mission thread modeling –Establish a basis of understanding to generate a productive set of standards for the components and structures with which to assemble mission thread models

23 Questions Tom Tanner thomas.l.tanner@saic.com 540 663 4867 (Office) 540 903 6713 (Cell) thomas.l.tanner@saic.com

24 2011 System of Systems Engineering Collaborators Information Exchange Webinars April 12 th : A Game Loop Architecture for the Modeling and Simulation of Mission Threats, Thomas Tanner, SAIC May 3 rd : Mission Engineering for Warfighting Integration of Net-Centric Systems, Eileen Bjorkman and Timothy Menke, USAF May 10 th : The Role of Enterprise Architecture Updates in Guiding Decentralized Organizations, John Schatz, SPEC Innovations May 24 th : Test and Evaluation Issues for Systems of Systems: Sleepless Nights to Sominex, Dr. Beth Wilson, Raytheon & Dr. Judith Dahmann, MITRE June 14 th : Evaluating the Readiness of Federations-of-Models for Use in Simulation-Based Concept Development of Advanced Warfighting Capabilities, Bryan Herdlick, JPU/APL For information, email dasd-se@osd.mil or visit our website: http://www.acq.osd.mil/se/outreach/sosecollab.htmldasd-se@osd.mil http://www.acq.osd.mil/se/outreach/sosecollab.html


Download ppt "A Game Loop Architecture for the Modeling and Simulation of Mission Threads Tom Tanner 540 663-4867 (Office) 540 903-6713 (Cell)"

Similar presentations


Ads by Google