Cooperative rule-based systems for aircraft control

This paper examines the application of artificial intelligence theory in the flight domain of a military aircraft. Nine rule-based systems were implemented to demonstrate complex system cooperation in combat aircraft operations. The organization of tasks within each rule-based system is described, and details of knowledge-base development and implementation are given. An interactive simulation testbed was developed to provide a realistic view of intersystem cooperation and pilot-system interaction. Single-processor emulations demonstrate cooperation of parallel rule-based systems. Software tools developed to aid in fast prototyping of rule-based systems are described. Search effort metrics were used to quantify and compare light- and heavy-workload phases of a combat mission.