Using adaptive agents in Java to simulate U.S. air force pilot retention

The retention of skilled pilots is a problem that plagues the United States Air Force. After spending millions of dollars on training and education, it is disheartening to see the mass exodus of experienced aviators from the Air Force that has been occurring in the past decade. Many blame the economy, others the Air Force itself, but few are able to accurately predict how or why they are all leaving. Complex adaptive systems theory might provide some insight. By modeling the system at the pilot's level, allowing each pilot to be represented as an autonomous, independent agent continually adapting to its environment and the other agents in it, an alternate model can be built; one that accounts for the interactions among the pilots, not just their interactions with their environment. PICAS (Pilot Inventory Complex Adaptive System) is just such a model. Constructed in the Java language, the PICAS model exploits the notions of complex adaptive systems theory and employs dynamic user controls to discern retention rates over a pilot career time period. Pilots 'evolve', for lack of a better word, to a greater fitness within their environment, and in the process, the model user can better determine what kind of environment needs to be created and maintained in order to ensure that trained and experienced pilots are in fact retained for their services beyond their initial service commitments.