Living Transformation: ISAF-SEAS Case Study for Urgent Computing

Transcription

1 Living Transformation: ISAF-SEAS Case Study for Urgent Computing Alok Chaturvedi Purdue University West Lafayette, IN (765)

2 Topics What is ISAF? SEAS role in ISAF SEAS in theater Integration with external systems - ONA Challenges Organizational Process Technology Future Direction Towards Sentient World 1

3 SEAS-ISAF ISAF - International Security Assistance Force Setup by United Nation in 2001 to provide security assistance to the people of Afghanistan SEAS (Synthetic Environment For Analysis and Simulation) is deployed in the theater for courses of actions analysis Research funded in part by National Science Foundation since 1998 Transitioned to Simulex Inc. Currently in use at JFCOM, SOUTHCOM, OSD PA&E, USMC School, a Fortune 50 Company 2

4 SEAS current state SEAS VIS is a representation of 62 countries with validated models (well accepted, published in peer-reviewed reviewed journals) at varying degrees of detail Represents Political Nodes Military Nodes Economic Nodes Social Nodes Information Infrastructure 450+ Named Organizations 300+ Named Leaders 12,000+ Named Infrastructure Nodes 500+ Named media nodes 10+ Million Active Agents 3

5 N-sided Analysis Human-in-Loop - Adversary Multiple Human-in-the- Loop Capabilities Human-in-Loop - Coalition Intelligent Agent Country X s Intelligent Agent Intelligent Agents SEAS simulations model N-sided interactions where each side may be represented by HITL players or intelligent artificial agents.

6 Topics What is ISAF? SEAS role in ISAF SEAS in theater Integration with external systems - ONA Challenges Organizational Process Technology Future Direction Towards Sentient World 5

7 SEAS in the theater First persistent Simulation-based Decision Support Environment for Field Commander To be able to generate prediction i of instability in the region as real world events unfold To provide a Wind-tunnel for Courses of Action Simultaneous research, development, and deployment Recent events/intel constantly fed into the xna Multiple Courses of actions simulated in updated synthetic environments Decode adversaries playbooks System redundancies to ensure high availability and reach-back capability Runtime of single digit minutes to provide near instant feedback of proposed strategies 6

8 SEAS On-demand Scaling indicate on demand hotspots ous asynchron layers PMESII IED Supply Chain Recruiting GLOBAL week ticks Insurgency Orgs Communities (Refugee Camps) HumInt CITY hour ticks SECTOR IED Attack JSAF CBRN minute ticks finer detail Workflow Modeling TacSim

9 Dynamic Data Acquisition, Parameterization, and Calibration in SEAS Continuous & Concurrent Development of the Passive Environment Active Environment Input & Analysis COA Planners Automated Data Discovery Semantic Match/Filter xna Selective Persistence ONA Semantic Match/Filter Ontological Repository Calendar Day Initialization SEAS Engine Other Intel Systems Semantic Match/Filter SME User Interface Visualizations & Data Services Public Opinion PmESII Metrics Trends & Explanations Possible Alternative Futures

10 Data in extensible Net Assessment Snapshot at a specific time So ource Legend (Example) Points of View Blue Green Red 9 Sources ONA Simulex Researcher CNN Simulation COA Al Jazeera Every piece of data is stored with an effective date range, its source, and its point of view. Snapshots can be taken at any point in time, with any subset of sources and points of view.

11 Semantic Mining for XNA Transliteration Dictionary xna Repository Web Data Sources: Major News Sources Blogs of Interest Social Networking Sites Government Statistics Web Crawler Transliteration Keyword Expansion and search weighting Search Index Social network search Ontological Expansion and Search weighting Chronological Index Chronological Search Returns Search Results are Returned with Associated Timelines that Allow Analysts to View Data at Various Points in Time Analysts & Researchers xna Data Repository 10

12 System Redundancies 11

13 Service Oriented Experiment Management 1. ExMan Main Portal 2. Simulation Browsing and Selection 3. Nodes in the selected simulation 5. Execution Environments Management Calendar-based Simulation Initialization

14 Multiple Excursions Real World SEAS Reference Synthetic World Inject 1 Inject 2 Inject 3 Inject 4 Track Real World Organization 1 Organization 2 Organization 3 SEAS Deployment 1 T E1 E2 E3 SEAS Deployment 2 T E1 E2 E3 SEAS Deployment 3 T E1 E2 E3 Analysis Analysis Analysis Future 1 Future 2 Future 3

15 Topics What is ISAF? SEAS implmentation in ISAF SEAS in theater Integration with external systems - ONA and data entry tools Challenges Organizational Process Technology Future Direction Towards Sentient World 14

16 Tools for Updating xna ONA-XNA Integration Tool and SESEN Analyst Workbench High-level tools that allow for easy addition and manipulation of Node, Action, and Event information. SESEN allows for the addition and editing of node information within xna. The ONA-XNA Integration Tool allows for nodes to be exported from ONAto xna and augmented with additional information needed by SEAS. 15 Currently Supported: Leaders Organizations Infrastructure xna Low-level tool for viewing and manipulating the ontological structure and other data within xna. Currently a prototype, but under active development to become part of the suite of tools available for SMEs to use. Under Development: Media Citizens Institutions Economic Sectors Actions and Events

17 Topics What is ISAF? SEAS implementation in ISAF SEAS in theater Integration with external systems - ONA and data entry tools Challenges Organizational Process Technology Future Direction Towards Sentient World 16

18 Organizational Challenges The project marks several firsts, including: First operational system deployment First theater deployment First deployment outside the country First system with the requirement of near perfect uptime Living Transformation Simultaneous research, development, and deployment Simultaneous CONOP/TTP development and training Rapid accreditation, field testing, and deployment Legal, business and acquisition challenges 17

19 Process Challenges Security System The need to ready the system for NSWAN security compliance The flavor of Linux that SEAS prefers has to be replaced by Red Hat Linux Server Java Development Kit usage has to be replaced by Java Runtime Environment Data Constant research and update of the simulation with the latest data ISAF and J9 analysts are updating the xna with data with secret classification Merging and conflict detection is a challenge 18

20 Technological Challenges Low bandwidth and reliability environment The IT infrastructure in the theater is heterogeneous. Simulation clients could be accessed via satellite internet, which could be very slow and unreliable The machines available are often lesser than the required spec Guaranteed uptime and backward compatibility Constant service availability is crucial in the theater Simulation has to be running even during analysts update data Courses of actions and simulation results has to be saved and retrievable even if produced by previous versions of the software 19

21 Topics What is ISAF? SEAS implementation in ISAF SEAS in theater Integration with external systems - ONA and data entry tools Challenges Organizational Process Technology Future Direction Towards Sentient World 20

22 Towards Sentient World 21

23 SWS Deployment Framework Public Domain Open source data Developed in Collaboration Classified Worlds Hosted by Specific Intelligence Communities Augmented with User- Provided d Data Sources Reference World Regulated Development Pure Reference World Hosted at J9 Basis for Instantiating Classified & Proprietary Worlds Proprietary Worlds Inter Agency Communities