What if ground attack air fact could choose targets based on market dynamics?

JFACC AND PHASING

• Source of grand strategy for U.S. air operations is the Joint Force Air Component Commander JFACC, reflecting Joint Force Commander’s intent, establishes goals for each phase of the campaign Air apportionment: scheduling of resources and target classes for priority attack Air allocation: scheduling of individual platforms against individual targets

LEARN FROM COMPLEXITY

• "Pre-planning" for targets in ODS was augmented as targets moved
• 10% of A-10s sat on "trip alert"
• Global Information Grid was only an emerging concept

DIGITIZING THE BATTLEFIELD (2002)

• Several initiatives were underway in 2002 to provide a “common picture” of a combat situationGlobal Information Grid DoD C4ISR Architecture Framework Defense Information Infrastructure Common Operating Environment (DII COE) If the battle space were ever truly digitized, perhaps equally “smart” air platforms could determine optimal targeting as the situation evolves, as contrasted with scripted air planning.

CONNECT WEAPONS TO TARGET DATA?

• Connect via algorithms
• Aircraft operate based on rules (Boids, Craig Reynolds 1986)
• Would this be a more efficient use of resources?
• Is this feasible?

HOW WOULD WE MODEL THIS

• Equation-based modeling cannot address complex systems
• Agent-based simulation allows for emergence
• No common modeling platform to compare status quo with self-organized approach

DEVELOPING ABACUS

• Agent-Based Air Campaign Using Swarm Swarm modeling environment Objective C Allowed for construct of rule-based agents and environments, along with parameterization Sugarscape application - build within Swarm libraries Used because it combined autonomous, mobile agents with cellular automata ABACUS uses two types of agents: Aircraft and targets

ISSUES

• Why would a non-hierarchical campaign not be the answer? Strategic effects (Warden) are best achieved through degradation of key targets in a timely sequence, to encourage the collapse of interlocked systems Why would the Air Force resist? History is not kind to decentralized air tasking (Vietnam) Where might it work? In a target-rich, resource-constrained environment with extremely good target intelligence available to all platforms Where might it not work? Anywhere else

MOTIVATION

• During campaign, inside the planning cycle Air attack assets will become available unexpectedly Targets will emerge A completely scripted air campaign by definition does not address these contingencies “Revolution in Military Affairs” was made up of several elements, the increase of which should exert pressures on existing doctrine Speed, agility, lethality, and information Construct a nonhierarchical model for the tasking of air assets in order to test an agent-based approach to the servicing of targets in an air campaign

Structure of the Paper

Lessons learned

Problems for air planners

Script

Flock of birds

Agent-based modeling

Computer modeling

Decentralized execution

Testing the model

Caveat

Policy implications

Summary

BIOGRAPHY