20050531ZGT11462

RAIDReal-time Adversarial Intelligence & Decision-making: 

RAID Real-time Adversarial Intelligence & Decision-making Briefing to Industry BAA 04-16 17 March 2004 Arlington VA Sponsored by DARPA/IXO

Slide2: 

Army – Predictive Analysis ... used in the US Army community to denote a process and tools for predicting future enemy actions. When the operation is executed, the commander and staff make the necessary changes ... while continually visualizing the possible future actions and reactions of the enemy. Air Force – Predictive Battlespace Awareness ... refers to future techniques and technologies that would help the commander and staff to characterize and predict likely enemy courses of action, to relate the history of the enemy’s performance to its current and future actions, and to associate these predictions with opportunities for friendly actions and effects. Unsolved Service Requirements

Today’s Technologies Fail to Address the Problem: 

Today’s Technologies Fail to Address the Problem Today’s Automation: Detailed optimization of resource allocation, times, & routes No attempt to infer or to influence the actions of the Red Humans: Focused on impacting actions of the Red, by deception and exploiting Red errors Recent experimentation compared the task planning and resource allocations produced by advanced automation to those produced by humans. Lack of consideration for the adversary’s moves and countermoves was identified as a key functional shortfall in the automation.

RAID Objective: 

RAID Objective A tool for: real-time anticipation of enemy actions in tactical ground operations 17 March 2004 RAID Program (BAA 04-16) Briefing to Industry 4

The Problem DomainUrban Operations: 

The Problem Domain Urban Operations Full 3-D Dynamic Environment Lines Open/Close Barriers Close/Open Limited Line-of-Sight Unlimited Hide Sites Weapons Constraints Combat ID Non-combatants Protected Sites

Slide6: 

The Problem Domain Urban Operations Consideration of weapon capabilities and limitations. For example, tanks have multiple dead zones and trajectory limits, which can make them vulnerable in an urban environment.

Slide7: 

The Problem Domain Urban Operations Consideration of urban tactics. For example, crossing an intersection requires multiple, coordinated moves. Step 1 – Isolate and Suppress Step 2 – Cross Over Step 3 – Cover and Proceed

Vision of the RAID Program: 

Vision of the RAID Program Develop a near real-time adversarial predictive analysis tool that operates as an automated Red predictor providing a continuously updated picture of probable Red actions in tactical ground operations

Program Structure: 

Program Structure 17 March 2004 RAID Program (BAA 04-16) Briefing to Industry 9

RAID System Concept: 

RAID System Concept

RAID System ConceptInformation Flows: 

RAID System Concept Information Flows

Program Schedule: 

Program Schedule Experimentation and Evaluation Integration Core Technologies Adversarial Reasoning, Deception Reasoning Theme of the Phase Anticipate and counteract Feints and attacks Concealment and Deception RAID performs action-reaction-counteraction predictive analysis Contracts Start Combatant models Human preferences Breadth and Robustness w/ sim system RAID detects/anti-cipates adversary’s concealment and deception Adversarial anticipation and counteraction Adversarial reasoning about concealment and deception Integration and Transition Q3 Q2 Q1 Q4 Q3 Q2 Q1 Q4 Q3 Q2 Q1 Q4 Q3 Q2 Q1 Phase III Phase II Phase I FY07 FY06 FY05 FY04 Virtual Transition-driven extensions Partial information w/ DCGS-A system RAID is used with operational C2 and ISR systems Q4 w/ sim system

Slide13: 

The RAID Program will be CLASSIFIED in accordance with the DD Form 254 associated with the PIP. The prime contractor in each of the four task areas must comply with US National Security Requirements and Export Control Laws. However, the basic research in the two technology areas (Adversarial Reasoning and Deception Reasoning) can be performed by an UNCLASSIFIED subcontractor, such as a university. Proposals submitted in response to this BAA shall be UNCLASSIFIED. Security

Technical Objectives and Experimental Approach: 

Technical Objectives and Experimental Approach 17 March 2004 RAID Program (BAA 04-16) Briefing to Industry 14

Slide15: 

Tight interdependence, coupling of blue and red actions. Blue knowledge of red assets and actions is inevitably limited. Observations as well as interpretations of the observations are subject to a significant degree of errors and latency. In addition to partial, delayed and often erroneous observations, the battlefield knowledge is limited by a purposeful, continuous, aggressive, intelligent concealment and deception. Cultural, doctrinal, psychological effects. It is not enough to consider the most dangerous course of action. The most likely course of action can be significantly different from the theoretically most advantageous one. Complex urban terrain offers a high density of threats and opportunities for forces. Further, the terrain itself is dynamic because it is modified by human actions. The presence of non-combatants on the battlefield must be explicitly considered. Fire and maneuver of forces are not the only actions that must be carefully considered. Intelligence gathering, communications, and logistics (including casualty evacuation) are tightly coupled with fire and maneuver. The scale of the computational problem is immense and yet solutions must be generated in near real-time. To be of practical value, a successful technical approach must allow for easy modification and extension of the coverage. Overarching Technical Challenges

Slide16: 

Any technology or a combination of technologies that addresses the problem and the challenges discussed in relation to the Adversarial Reasoning and Deception Reasoning, and which can be developed in a manner consistent with the RAID program concept, would be of interest to this BAA. Potential candidate groups of technologies include but are not limited to: Game-theoretic and game-playing approaches: Devising sequences of actions for both red and blue forces in a manner that assumes both sides strive to maximize the achievement of their respective objectives. Adversarial planning: Forming plans for both red and blue actions that lead to the achievement of the respective desired goals while preventing the attainment of the goals of the other side. Deception discovery: Analyzing the information state from risk-sensitive perspective to determine which of alternative hypotheses would benefit the enemy the most if accepted by the friendly forces. Pattern recognition: Identifying patterns and anomalies in spatial and temporal locations, movements and other actions of the red force that could indicate concealment, deception and future intended course of action. Candidate Technical Approaches

Challenge Problem and Scenarios: 

Challenge Problem and Scenarios Three (3) Scenarios Seize and control several key intersections in a city. Search, clear and secure built-up area near an airfield. Arrest urban guerilla leaders; avoid collateral damage. Blue Force – Company Sized Unit Several light armored vehicles, helicopters or fixed wing CAS Supported by joint, close supporting fires, including Netfires Red Force – experienced and motivated urban fighters Up to 200 fighters in teams of 3-7 personnel Small arms, RPGs, mortars, heavy machine guns, MANPADS, Strong knowledge of terrain Blends with and assisted by civilians Urban or built-up terrain; space ~ 5x5 km; time ~ 8 hrs Role of RAID (e.g., application within future DCGS-A): Watch the rapidly evolving COP; Deliver to the Company Commander (via FBCB2 or handheld CDA) real-time predictions on upcoming RED actions (movements, ambushes, etc.) and suggested BLUE counteractions. Perform the analysis unobtrusively, on demand, in less than 30 seconds and look forward 15-300 minutes as requested. Developed by Thomas Garrett, MG USA (Ret).

Challenge ProblemScenario 1 – Control the Bridges: 

Challenge Problem Scenario 1 – Control the Bridges Mission: Attack to seize key intersections controlling approaches to key bridges that lead to the heart of the city. Continue the attack to seize and secure 2 key bridges. Situation: BLUFOR is attacking into the heart of large urban capital city. A RAID supported company is operating as an advanced element The weather is clear. Red Force: Remnants of regular forces, irregular forces, and small extremist elements from outside the country defend the bridges with anti-tank weapons, heavy machine guns, mortars, rockets, and MANPADS. The approaches are mined, and the mines are covered by ambushing fire. Time: This is a night approach. The broader attack to this point has been extremely fast paced and relentless. The concept is to keep the momentum of the assault going. A follow-on force will cross the bridges, pass through our company, and enter the heart of the city.

Challenge ProblemScenario 2 – Secure Area West Of Airfield: 

Challenge Problem Scenario 2 – Secure Area West Of Airfield Mission: Search, clear and secure area west of airfield. Situation: Aircraft using the airfield have been fired upon from the surrounding neighborhood. Weather is day and there is a dust storm. Red Force: Small insurgent teams armed with SA-7, SA-18, and RPGs, as well as small arms are believed to be moving into and out of the area under the cover of darkness and adverse weather, and engaging approaching and departing aircraft. Time: Need to swiftly cordon off the area, then systematically search and clear. Establish a sustained security of the area to prevent it’s use by insurgents.

Overview of Experimental Approach: 

Overview of Experimental Approach Blue Cell 5-7 personnel w/o RAID 1-2 personnel w/ RAID Red Cell 5-7 personnel Commands agile and aggressive Red Force Data collection and analysis cell (1-2 personnel) computes scores and predictive accuracy w/ and w/o RAID Control Cell (2-3 personnel) enforces realism and integrity of the wargame Commands Situation Commands Situation RAID System Predictions of Red Suggestions for Blue Combat Simulation System

Overview of Experimental Approach(cont.): 

Overview of Experimental Approach (cont.) Blue Cell 5-7 personnel w/o RAID 1-2 personnel w/ RAID Red Cell 5-7 personnel Commands agile and aggressive Red Force Data collection and analysis cell (1-2 personnel) computes scores and predictive accuracy w/ and w/o RAID Control Cell (2-3 personnel) enforces realism and integrity of the wargame Commands Situation Commands Situation RAID System Predictions of Red Suggestions for Blue Each series is 5-10 benchmark games (without RAID) and 5-10 test games (with RAID), duration 2-6 hours Simulation software Hardware: up to 20 PCs Interfaces: Phase 1,2 – RAID/Simulation; Phase 3 – RAID/DCGS-A/ASAS-L/FBCB2/CDA Game scoring scheme: points for mission completion within time allotted; enemy captured and killed; penalties for friendly losses and collateral casualties. Complexity of the problem varies by number and granularity of red/blue units; allowed set of actions; weapons

Experimental Plan: 

Experimental Plan